OpenJPA User's Guide


Table of Contents

I. Introduction
1. OpenJPA
1. About This Document
II. Java Persistence API
1. Introduction
1. Intended Audience
2. Lightweight Persistence
2. Why JPA?
3. Java Persistence API Architecture
1. JPA Exceptions
4. Entity
1. Restrictions on Persistent Classes
1.1. Default or No-Arg Constructor
1.2. Final
1.3. Identity Fields
1.4. Version Field
1.5. Inheritance
1.6. Persistent Fields
1.7. Conclusions
2. Entity Identity
2.1. Identity Class
2.1.1. Identity Hierarchies
3. Lifecycle Callbacks
3.1. Callback Methods
3.2. Using Callback Methods
3.3. Using Entity Listeners
3.4. Entity Listeners Hierarchy
4. Conclusions
5. Metadata
1. Class Metadata
1.1. Entity
1.2. Id Class
1.3. Mapped Superclass
1.4. Embeddable
1.5. EntityListeners
1.6. Example
2. Field and Property Metadata
2.1. Transient
2.2. Id
2.3. Generated Value
2.4. Embedded Id
2.5. Version
2.6. Basic
2.6.1. Fetch Type
2.7. Embedded
2.8. Many To One
2.8.1. Cascade Type
2.9. One To Many
2.9.1. Bidirectional Relations
2.10. One To One
2.11. Many To Many
2.12. Order By
2.13. Map Key
2.14. Persistent Field Defaults
3. XML Schema
4. Conclusion
6. Persistence
1. persistence.xml
2. Non-EE Use
7. EntityManagerFactory
1. Obtaining an EntityManagerFactory
2. Obtaining EntityManagers
3. Persistence Context
3.1. Transaction Persistence Context
3.2. Extended Persistence Context
4. Closing the EntityManagerFactory
8. EntityManager
1. Transaction Association
2. Entity Lifecycle Management
3. Lifecycle Examples
4. Entity Identity Management
5. Cache Management
6. Query Factory
7. Closing
9. Transaction
1. Transaction Types
2. The EntityTransaction Interface
10. JPA Query
1. JPQL API
1.1. Query Basics
1.2. Relation Traversal
1.3. Fetch Joins
1.4. JPQL Functions
1.5. Polymorphic Queries
1.6. Query Parameters
1.7. Ordering
1.8. Aggregates
1.9. Named Queries
1.10. Delete By Query
1.11. Update By Query
2. JPQL Language Reference
2.1. JPQL Statement Types
2.1.1. JPQL Select Statement
2.1.2. JPQL Update and Delete Statements
2.2. JPQL Abstract Schema Types and Query Domains
2.2.1. JPQL Entity Naming
2.2.2. JPQL Schema Example
2.3. JPQL FROM Clause and Navigational Declarations
2.3.1. JPQL FROM Identifiers
2.3.2. JPQL Identification Variables
2.3.3. JPQL Range Declarations
2.3.4. JPQL Path Expressions
2.3.5. JPQL Joins
2.3.5.1. JPQL Inner Joins (Relationship Joins)
2.3.5.2. JPQL Outer Joins
2.3.5.3. JPQL Fetch Joins
2.3.6. JPQL Collection Member Declarations
2.3.7. JPQL Polymorphism
2.4. JPQL WHERE Clause
2.5. JPQL Conditional Expressions
2.5.1. JPQL Literals
2.5.2. JPQL Identification Variables
2.5.3. JPQL Path Expressions
2.5.4. JPQL Input Parameters
2.5.4.1. JPQL Positional Parameters
2.5.4.2. JPQL Named Parameters
2.5.5. JPQL Conditional Expression Composition
2.5.6. JPQL Operators and Operator Precedence
2.5.7. JPQL Between Expressions
2.5.8. JPQL In Expressions
2.5.9. JPQL Like Expressions
2.5.10. JPQL Null Comparison Expressions
2.5.11. JPQL Empty Collection Comparison Expressions
2.5.12. JPQL Collection Member Expressions
2.5.13. JPQL Exists Expressions
2.5.14. JPQL All or Any Expressions
2.5.15. JPQL Subqueries
2.5.16. JPQL Functional Expressions
2.5.16.1. JPQL String Functions
2.5.16.2. JPQL Arithmetic Functions
2.5.16.3. JPQL Datetime Functions
2.6. JPQL GROUP BY, HAVING
2.7. JPQL SELECT Clause
2.7.1. JPQL Result Type of the SELECT Clause
2.7.2. JPQL Constructor Expressions
2.7.3. JPQL Null Values in the Query Result
2.7.4. JPQL Aggregate Functions
2.7.4.1. JPQL Aggregate Examples
2.8. JPQL ORDER BY Clause
2.9. JPQL Bulk Update and Delete
2.10. JPQL Null Values
2.11. JPQL Equality and Comparison Semantics
2.12. JPQL BNF
11. SQL Queries
1. Creating SQL Queries
2. Retrieving Persistent Objects with SQL
12. Mapping Metadata
1. Table
2. Unique Constraints
3. Column
4. Identity Mapping
5. Generators
5.1. Sequence Generator
5.2. TableGenerator
5.3. Example
6. Inheritance
6.1. Single Table
6.1.1. Advantages
6.1.2. Disadvantages
6.2. Joined
6.2.1. Advantages
6.2.2. Disadvantages
6.3. Table Per Class
6.3.1. Advantages
6.3.2. Disadvantages
6.4. Putting it All Together
7. Discriminator
8. Field Mapping
8.1. Basic Mapping
8.1.1. LOBs
8.1.2. Enumerated
8.1.3. Temporal Types
8.1.4. The Updated Mappings
8.2. Secondary Tables
8.3. Embedded Mapping
8.4. Direct Relations
8.5. Join Table
8.6. Bidirectional Mapping
8.7. Map Mapping
9. The Complete Mappings
13. Conclusion
III. Reference Guide
1. Introduction
1. Intended Audience
2. Configuration
1. Introduction
2. Runtime Configuration
3. Command Line Configuration
3.1. Code Formatting
4. Plugin Configuration
5. OpenJPA Properties
5.1. openjpa.AutoClear
5.2. openjpa.AutoDetach
5.3. openjpa.BrokerFactory
5.4. openjpa.BrokerImpl
5.5. openjpa.ClassResolver
5.6. openjpa.Compatibility
5.7. openjpa.ConnectionDriverName
5.8. openjpa.Connection2DriverName
5.9. openjpa.ConnectionFactory
5.10. openjpa.ConnectionFactory2
5.11. openjpa.ConnectionFactoryName
5.12. openjpa.ConnectionFactory2Name
5.13. openjpa.ConnectionFactoryMode
5.14. openjpa.ConnectionFactoryProperties
5.15. openjpa.ConnectionFactory2Properties
5.16. openjpa.ConnectionPassword
5.17. openjpa.Connection2Password
5.18. openjpa.ConnectionProperties
5.19. openjpa.Connection2Properties
5.20. openjpa.ConnectionURL
5.21. openjpa.Connection2URL
5.22. openjpa.ConnectionUserName
5.23. openjpa.Connection2UserName
5.24. openjpa.ConnectionRetainMode
5.25. openjpa.DataCache
5.26. openjpa.DataCacheManager
5.27. openjpa.DataCacheTimeout
5.28. openjpa.DetachState
5.29. openjpa.DynamicDataStructs
5.30. openjpa.FetchBatchSize
5.31. openjpa.FetchGroups
5.32. openjpa.FlushBeforeQueries
5.33. openjpa.IgnoreChanges
5.34. openjpa.Id
5.35. openjpa.InverseManager
5.36. openjpa.LockManager
5.37. openjpa.LockTimeout
5.38. openjpa.Log
5.39. openjpa.ManagedRuntime
5.40. openjpa.Mapping
5.41. openjpa.MaxFetchDepth
5.42. openjpa.MetaDataFactory
5.43. openjpa.Multithreaded
5.44. openjpa.Optimistic
5.45. openjpa.OrphanedKeyAction
5.46. openjpa.NontransactionalRead
5.47. openjpa.NontransactionalWrite
5.48. openjpa.ProxyManager
5.49. openjpa.QueryCache
5.50. openjpa.QueryCompilationCache
5.51. openjpa.ReadLockLevel
5.52. openjpa.RemoteCommitProvider
5.53. openjpa.RestoreState
5.54. openjpa.RetainState
5.55. openjpa.RetryClassRegistration
5.56. openjpa.SavepointManager
5.57. openjpa.Sequence
5.58. openjpa.TransactionMode
5.59. openjpa.WriteLockLevel
6. OpenJPA JDBC Properties
6.1. openjpa.jdbc.ConnectionDecorators
6.2. openjpa.jdbc.DBDictionary
6.3. openjpa.jdbc.DriverDataSource
6.4. openjpa.jdbc.EagerFetchMode
6.5. openjpa.jdbc.FetchDirection
6.6. openjpa.jdbc.JDBCListeners
6.7. openjpa.jdbc.LRSSize
6.8. openjpa.jdbc.MappingDefaults
6.9. openjpa.jdbc.MappingFactory
6.10. openjpa.jdbc.ResultSetType
6.11. openjpa.jdbc.Schema
6.12. openjpa.jdbc.SchemaFactory
6.13. openjpa.jdbc.Schemas
6.14. openjpa.jdbc.SQLFactory
6.15. openjpa.jdbc.SubclassFetchMode
6.16. openjpa.jdbc.SynchronizeMappings
6.17. openjpa.jdbc.TransactionIsolation
6.18. openjpa.jdbc.UpdateManager
3. Logging
1. Logging Channels
2. OpenJPA Logging
3. Disabling Logging
4. Log4J
5. Apache Commons Logging
5.1. JDK 1.4 java.util.logging
6. Custom Log
4. JDBC
1. Using the OpenJPA DataSource
2. Using a Third-Party DataSource
2.1. Managed and XA DataSources
3. Runtime Access to DataSource
4. Database Support
4.1. DBDictionary Properties
4.2. MySQLDictionary Properties
4.3. OracleDictionary Properties
5. Setting the Transaction Isolation
6. Setting the SQL Join Syntax
7. Accessing Multiple Databases
8. Configuring the Use of JDBC Connections
9. Large Result Sets
10. Default Schema
11. Schema Reflection
11.1. Schemas List
11.2. Schema Factory
12. Schema Tool
13. XML Schema Format
5. Persistent Classes
1. Persistent Class List
2. Enhancement
2.1. Enhancing at Build Time
2.2. Enhancing JPA Entities on Deployment
2.3. Enhancing at Runtime
2.4. Serializing Enhanced Types
3. Object Identity
3.1. Datastore Identity
3.2. Entities as Identity Fields
3.3. Application Identity Tool
3.4. Autoassign / Identity Strategy Caveats
4. Managed Inverses
5. Persistent Fields
5.1. Restoring State
5.2. Typing and Ordering
5.3. Calendar Fields and TimeZones
5.4. Proxies
5.4.1. Smart Proxies
5.4.2. Large Result Set Proxies
5.4.3. Custom Proxies
5.5. Externalization
5.5.1. External Values
6. Fetch Groups
6.1. Custom Fetch Groups
6.2. Custom Fetch Group Configuration
6.3. Per-field Fetch Configuration
6.4. Implementation Notes
7. Eager Fetching
7.1. Configuring Eager Fetching
7.2. Eager Fetching Considerations and Limitations
6. Metadata
1. Metadata Factory
2. Additional JPA Metadata
2.1. Datastore Identity
2.2. Surrogate Version
2.3. Persistent Field Values
2.4. Persistent Collection Fields
2.5. Persistent Map Fields
3. Metadata Extensions
3.1. Class Extensions
3.1.1. Fetch Groups
3.1.2. Data Cache
3.1.3. Detached State
3.2. Field Extensions
3.2.1. Dependent
3.2.2. Load Fetch Group
3.2.3. LRS
3.2.4. Inverse-Logical
3.2.5. Read-Only
3.2.6. Type
3.2.7. Externalizer
3.2.8. Factory
3.2.9. External Values
3.3. Example
7. Mapping
1. Forward Mapping
1.1. Using the Mapping Tool
1.2. Generating DDL SQL
1.3. Runtime Forward Mapping
2. Reverse Mapping
2.1. Customizing Reverse Mapping
3. Meet-in-the-Middle Mapping
4. Mapping Defaults
5. Mapping Factory
6. Non-Standard Joins
7. Additional JPA Mappings
7.1. Datastore Identity Mapping
7.2. Surrogate Version Mapping
7.3. Multi-Column Mappings
7.4. Join Column Attribute Targets
7.5. Embedded Mapping
7.6. Collections
7.6.1. Container Table
7.6.2. Element Join Columns
7.6.3. Order Column
7.7. One-Sided One-Many Mapping
7.8. Maps
7.9. Indexes and Constraints
7.9.1. Indexes
7.9.2. Foreign Keys
7.9.3. Unique Constraints
8. Mapping Limitations
8.1. Table Per Class
9. Mapping Extensions
9.1. Class Extensions
9.1.1. Subclass Fetch Mode
9.1.2. Strategy
9.1.3. Discriminator Strategy
9.1.4. Version Strategy
9.2. Field Extensions
9.2.1. Eager Fetch Mode
9.2.2. Nonpolymorphic
9.2.3. Class Criteria
9.2.4. Strategy
10. Custom Mappings
10.1. Custom Class Mapping
10.2. Custom Discriminator and Version Strategies
10.3. Custom Field Mapping
10.3.1. Value Handlers
10.3.2. Field Strategies
10.3.3. Configuration
11. Orphaned Keys
8. Deployment
1. Factory Deployment
1.1. Standalone Deployment
1.2. EntityManager Injection
2. Integrating with the Transaction Manager
3. XA Transactions
3.1. Using OpenJPA with XA Transactions
9. Runtime Extensions
1. Architecture
1.1. Broker Finalization
1.2. Broker Customization and Finalization
2. JPA Extensions
2.1. OpenJPAEntityManagerFactory
2.2. OpenJPAEntityManager
2.3. OpenJPAQuery
2.4. Extent
2.5. StoreCache
2.6. QueryResultCache
2.7. FetchPlan
2.8. OpenJPAPersistence
3. Object Locking
3.1. Configuring Default Locking
3.2. Configuring Lock Levels at Runtime
3.3. Object Locking APIs
3.4. Lock Manager
3.5. Rules for Locking Behavior
3.6. Known Issues and Limitations
4. Savepoints
4.1. Using Savepoints
4.2. Configuring Savepoints
5. MethodQL
6. Generators
6.1. Runtime Access
7. Transaction Events
8. Non-Relational Stores
10. Caching
1. Data Cache
1.1. Data Cache Configuration
1.2. Data Cache Usage
1.3. Query Cache
1.4. Cache Extension
1.5. Important Notes
1.6. Known Issues and Limitations
2. Query Compilation Cache
11. Remote and Offline Operation
1. Detach and Attach
1.1. Detach Behavior
1.2. Attach Behavior
1.3. Defining the Detached Object Graph
1.3.1. Detached State Field
2. Remote Event Notification Framework
2.1. Remote Commit Provider Configuration
2.1.1. JMS
2.1.2. TCP
2.1.3. Common Properties
2.2. Customization
12. Third Party Integration
1. Apache Ant
1.1. Common Ant Configuration Options
1.2. Enhancer Ant Task
1.3. Application Identity Tool Ant Task
1.4. Mapping Tool Ant Task
1.5. Reverse Mapping Tool Ant Task
1.6. Schema Tool Ant Task
13. Optimization Guidelines
A. JPA Resources
B. Supported Databases
1. Apache Derby
2. Borland Interbase
2.1. Known issues with Interbase
3. JDataStore
4. IBM DB2
4.1. Known issues with DB2
5. Empress
5.1. Known issues with Empress
6. H2 Database Engine
6.1. Known issues with H2 Database Engine
7. Hypersonic
7.1. Known issues with Hypersonic
8. Firebird
8.1. Known issues with Firebird
9. Informix
9.1. Known issues with Informix
10. InterSystems Cache
10.1. Known issues with InterSystems Cache
11. Microsoft Access
11.1. Known issues with Microsoft Access
12. Microsoft SQL Server
12.1. Known issues with SQL Server
13. Microsoft FoxPro
13.1. Known issues with Microsoft FoxPro
14. MySQL
14.1. Known issues with MySQL
15. Oracle
15.1. Using Query Hints with Oracle
15.2. Known issues with Oracle
16. Pointbase
16.1. Known issues with Pointbase
17. PostgreSQL
17.1. Known issues with PostgreSQL
18. Sybase Adaptive Server
18.1. Known issues with Sybase

List of Tables

2.1. Persistence Mechanisms
4.1. OpenJPA Automatic Flush Behavior
5.1. Externalizer Options
5.2. Factory Options
10.1. Data access methods
10.2. Pre-defined aliases
13.1. Optimization Guidelines
B.1. Supported Databases and JDBC Drivers

List of Examples

3.1. Interaction of Interfaces Outside Container
3.2. Interaction of Interfaces Inside Container
4.1. Persistent Class
4.2. Identity Class
5.1. Class Metadata
5.2. Complete Metadata
6.1. persistence.xml
6.2. Obtaining an EntityManagerFactory
7.1. Behavior of Transaction Persistence Context
7.2. Behavior of Extended Persistence Context
8.1. Persisting Objects
8.2. Updating Objects
8.3. Removing Objects
8.4. Detaching and Merging
9.1. Grouping Operations with Transactions
10.1. Delete by Query
10.2. Update by Query
11.1. Creating a SQL Query
11.2. Retrieving Persistent Objects
11.3. SQL Query Parameters
12.1. Mapping Classes
12.2. Defining a Unique Constraint
12.3. Identity Mapping
12.4. Generator Mapping
12.5. Single Table Mapping
12.6. Joined Subclass Tables
12.7. Table Per Class Mapping
12.8. Inheritance Mapping
12.9. Discriminator Mapping
12.10. Basic Field Mapping
12.11. Secondary Table Field Mapping
12.12. Embedded Field Mapping
12.13. Mapping Mapped Superclass Field
12.14. Direct Relation Field Mapping
12.15. Join Table Mapping
12.16. Join Table Map Mapping
12.17. Full Entity Mappings
2.1. Code Formatting with the Application Id Tool
3.1. Standard OpenJPA Log Configuration
3.2. Standard OpenJPA Log Configuration + All SQL Statements
3.3. Logging to a File
3.4. Standard Log4J Logging
3.5. JDK 1.4 Log Properties
3.6. Custom Logging Class
4.1. Properties for the OpenJPA DataSource
4.2. Properties File for a Third-Party DataSource
4.3. Managed DataSource Configuration
4.4. Using the EntityManager's Connection
4.5. Using the EntityManagerFactory's DataSource
4.6. Specifying a DBDictionary
4.7. Specifying a Transaction Isolation
4.8. Specifying the Join Syntax Default
4.9. Specifying the Join Syntax at Runtime
4.10. Specifying Connection Usage Defaults
4.11. Specifying Connection Usage at Runtime
4.12. Specifying Result Set Defaults
4.13. Specifying Result Set Behavior at Runtime
4.14. Schema Creation
4.15. SQL Scripting
4.16. Table Cleanup
4.17. Schema Drop
4.18. Schema Reflection
4.19. Basic Schema
4.20. Full Schema
5.1. Using the OpenJPA Enhancer
5.2. Using the OpenJPA Agent for Runtime Enhancement
5.3. Passing Options to the OpenJPA Agent
5.4. JPA Datastore Identity Metadata
5.5. Finding an Entity with an Entity Identity Field
5.6. Id Class for Entity Identity Fields
5.7. Using the Application Identity Tool
5.8. Specifying Logical Inverses
5.9. Enabling Managed Inverses
5.10. Log Inconsistencies
5.11. Using Initial Field Values
5.12. Using a Large Result Set Iterator
5.13. Marking a Large Result Set Field
5.14. Configuring the Proxy Manager
5.15. Using Externalization
5.16. Querying Externalization Fields
5.17. Using External Values
5.18. Custom Fetch Group Metadata
5.19. Load Fetch Group Metadata
5.20. Using the FetchPlan
5.21. Adding an Eager Field
5.22. Setting the Default Eager Fetch Mode
5.23. Setting the Eager Fetch Mode at Runtime
6.1. Setting a Standard Metadata Factory
6.2. Setting a Custom Metadata Factory
6.3. OpenJPA Metadata Extensions
7.1. Using the Mapping Tool
7.2. Creating the Relational Schema from Mappings
7.3. Refreshing entire schema and cleaning out tables
7.4. Dropping Mappings and Association Schema
7.5. Create DDL for Current Mappings
7.6. Create DDL to Update Database for Current Mappings
7.7. Configuring Runtime Forward Mapping
7.8. Reflection with the Schema Tool
7.9. Using the Reverse Mapping Tool
7.10. Customizing Reverse Mapping with Properties
7.11. Validating Mappings
7.12. Configuring Mapping Defaults
7.13. Standard JPA Configuration
7.14. Datastore Identity Mapping
7.15. Overriding Complex Mappings
7.16. One-Sided One-Many Mapping
7.17. Custom Logging Orphaned Keys
8.1. Configuring Transaction Manager Integration
9.1. Evict from Data Cache
9.2. Using a JPA Extent
9.3. Setting Default Lock Levels
9.4. Setting Runtime Lock Levels
9.5. Locking APIs
9.6. Disabling Locking
9.7. Using Savepoints
9.8. Named Seq Sequence
9.9. System Sequence Configuration
10.1. Single-JVM Data Cache
10.2. Data Cache Size
10.3. Data Cache Timeout
10.4. Accessing the StoreCache
10.5. StoreCache Usage
10.6. Automatic Data Cache Eviction
10.7. Accessing the QueryResultCache
10.8. Query Cache Size
10.9. Disabling the Query Cache
10.10. Evicting Queries
10.11. Pinning, and Unpinning Query Results
10.12. Disabling and Enabling Query Caching
10.13. Query Replaces Extent
11.1. Configuring Detached State
11.2. TCP Remote Commit Provider Configuration
12.1. Using the <config> Ant Tag
12.2. Using the Properties Attribute of the <config> Tag
12.3. Using the PropertiesFile Attribute of the <config> Tag
12.4. Using the <classpath> Ant Tag
12.5. Using the <codeformat> Ant Tag
12.6. Invoking the Enhancer from Ant
12.7. Invoking the Application Identity Tool from Ant
12.8. Invoking the Mapping Tool from Ant
12.9. Invoking the Reverse Mapping Tool from Ant
12.10. Invoking the Schema Tool from Ant
B.1. Example properties for Derby
B.2. Example properties for Interbase
B.3. Example properties for JDataStore
B.4. Example properties for IBM DB2
B.5. Example properties for Empress
B.6. Example properties for H2 Database Engine
B.7. Example properties for Hypersonic
B.8. Example properties for Firebird
B.9. Example properties for Informix Dynamic Server
B.10. Example properties for InterSystems Cache
B.11. Example properties for Microsoft Access
B.12. Example properties for Microsoft SQLServer
B.13. Example properties for Microsoft FoxPro
B.14. Example properties for MySQL
B.15. Example properties for Oracle
B.16. Using Oracle Hints
B.17. Example properties for Pointbase
B.18. Example properties for PostgreSQL
B.19. Example properties for Sybase

Introduction

Chapter 1.  OpenJPA

Table of Contents

1. About This Document

OpenJPA is Apache's implementation of Sun's Java Persistence API (JPA) specification for the transparent persistence of Java objects. This document provides an overview of the JPA standard and technical details on the use of OpenJPA.

1.  About This Document

This document is intended for OpenJPA users. It is divided into several parts:

  • The JPA Overview describes the fundamentals of the JPA specification.

  • The OpenJPA Reference Guide contains detailed documentation on all aspects of OpenJPA. Browse through this guide to familiarize yourself with the many advanced features and customization opportunities OpenJPA provides. Later, you can use the guide when you need details on a specific aspect of OpenJPA.

Java Persistence API

Table of Contents

1. Introduction
1. Intended Audience
2. Lightweight Persistence
2. Why JPA?
3. Java Persistence API Architecture
1. JPA Exceptions
4. Entity
1. Restrictions on Persistent Classes
1.1. Default or No-Arg Constructor
1.2. Final
1.3. Identity Fields
1.4. Version Field
1.5. Inheritance
1.6. Persistent Fields
1.7. Conclusions
2. Entity Identity
2.1. Identity Class
2.1.1. Identity Hierarchies
3. Lifecycle Callbacks
3.1. Callback Methods
3.2. Using Callback Methods
3.3. Using Entity Listeners
3.4. Entity Listeners Hierarchy
4. Conclusions
5. Metadata
1. Class Metadata
1.1. Entity
1.2. Id Class
1.3. Mapped Superclass
1.4. Embeddable
1.5. EntityListeners
1.6. Example
2. Field and Property Metadata
2.1. Transient
2.2. Id
2.3. Generated Value
2.4. Embedded Id
2.5. Version
2.6. Basic
2.6.1. Fetch Type
2.7. Embedded
2.8. Many To One
2.8.1. Cascade Type
2.9. One To Many
2.9.1. Bidirectional Relations
2.10. One To One
2.11. Many To Many
2.12. Order By
2.13. Map Key
2.14. Persistent Field Defaults
3. XML Schema
4. Conclusion
6. Persistence
1. persistence.xml
2. Non-EE Use
7. EntityManagerFactory
1. Obtaining an EntityManagerFactory
2. Obtaining EntityManagers
3. Persistence Context
3.1. Transaction Persistence Context
3.2. Extended Persistence Context
4. Closing the EntityManagerFactory
8. EntityManager
1. Transaction Association
2. Entity Lifecycle Management
3. Lifecycle Examples
4. Entity Identity Management
5. Cache Management
6. Query Factory
7. Closing
9. Transaction
1. Transaction Types
2. The EntityTransaction Interface
10. JPA Query
1. JPQL API
1.1. Query Basics
1.2. Relation Traversal
1.3. Fetch Joins
1.4. JPQL Functions
1.5. Polymorphic Queries
1.6. Query Parameters
1.7. Ordering
1.8. Aggregates
1.9. Named Queries
1.10. Delete By Query
1.11. Update By Query
2. JPQL Language Reference
2.1. JPQL Statement Types
2.1.1. JPQL Select Statement
2.1.2. JPQL Update and Delete Statements
2.2. JPQL Abstract Schema Types and Query Domains
2.2.1. JPQL Entity Naming
2.2.2. JPQL Schema Example
2.3. JPQL FROM Clause and Navigational Declarations
2.3.1. JPQL FROM Identifiers
2.3.2. JPQL Identification Variables
2.3.3. JPQL Range Declarations
2.3.4. JPQL Path Expressions
2.3.5. JPQL Joins
2.3.5.1. JPQL Inner Joins (Relationship Joins)
2.3.5.2. JPQL Outer Joins
2.3.5.3. JPQL Fetch Joins
2.3.6. JPQL Collection Member Declarations
2.3.7. JPQL Polymorphism
2.4. JPQL WHERE Clause
2.5. JPQL Conditional Expressions
2.5.1. JPQL Literals
2.5.2. JPQL Identification Variables
2.5.3. JPQL Path Expressions
2.5.4. JPQL Input Parameters
2.5.4.1. JPQL Positional Parameters
2.5.4.2. JPQL Named Parameters
2.5.5. JPQL Conditional Expression Composition
2.5.6. JPQL Operators and Operator Precedence
2.5.7. JPQL Between Expressions
2.5.8. JPQL In Expressions
2.5.9. JPQL Like Expressions
2.5.10. JPQL Null Comparison Expressions
2.5.11. JPQL Empty Collection Comparison Expressions
2.5.12. JPQL Collection Member Expressions
2.5.13. JPQL Exists Expressions
2.5.14. JPQL All or Any Expressions
2.5.15. JPQL Subqueries
2.5.16. JPQL Functional Expressions
2.5.16.1. JPQL String Functions
2.5.16.2. JPQL Arithmetic Functions
2.5.16.3. JPQL Datetime Functions
2.6. JPQL GROUP BY, HAVING
2.7. JPQL SELECT Clause
2.7.1. JPQL Result Type of the SELECT Clause
2.7.2. JPQL Constructor Expressions
2.7.3. JPQL Null Values in the Query Result
2.7.4. JPQL Aggregate Functions
2.7.4.1. JPQL Aggregate Examples
2.8. JPQL ORDER BY Clause
2.9. JPQL Bulk Update and Delete
2.10. JPQL Null Values
2.11. JPQL Equality and Comparison Semantics
2.12. JPQL BNF
11. SQL Queries
1. Creating SQL Queries
2. Retrieving Persistent Objects with SQL
12. Mapping Metadata
1. Table
2. Unique Constraints
3. Column
4. Identity Mapping
5. Generators
5.1. Sequence Generator
5.2. TableGenerator
5.3. Example
6. Inheritance
6.1. Single Table
6.1.1. Advantages
6.1.2. Disadvantages
6.2. Joined
6.2.1. Advantages
6.2.2. Disadvantages
6.3. Table Per Class
6.3.1. Advantages
6.3.2. Disadvantages
6.4. Putting it All Together
7. Discriminator
8. Field Mapping
8.1. Basic Mapping
8.1.1. LOBs
8.1.2. Enumerated
8.1.3. Temporal Types
8.1.4. The Updated Mappings
8.2. Secondary Tables
8.3. Embedded Mapping
8.4. Direct Relations
8.5. Join Table
8.6. Bidirectional Mapping
8.7. Map Mapping
9. The Complete Mappings
13. Conclusion

Chapter 1.  Introduction

The Java Persistence API (JPA) is a specification from Sun Microsystems for the persistence of Java objects to any relational datastore. JPA requires J2SE 1.5 (also referred to as "Java 5") or higher, as it makes heavy use of new Java language features such as annotations and generics. This document provides an overview of JPA. Unless otherwise noted, the information presented applies to all JPA implementations.

Note

For coverage of OpenJPA's many extensions to the JPA specification, see the Reference Guide.

1.  Intended Audience

This document is intended for developers who want to learn about JPA in order to use it in their applications. It assumes that you have a strong knowledge of object-oriented concepts and Java, including Java 5 annotations and generics. It also assumes some experience with relational databases and the Structured Query Language (SQL).

2.  Lightweight Persistence

Persistent data is information that can outlive the program that creates it. The majority of complex programs use persistent data: GUI applications need to store user preferences across program invocations, web applications track user movements and orders over long periods of time, etc.

Lightweight persistence is the storage and retrieval of persistent data with little or no work from you, the developer. For example, Java serialization is a form of lightweight persistence because it can be used to persist Java objects directly to a file with very little effort. Serialization's capabilities as a lightweight persistence mechanism pale in comparison to those provided by JPA, however. The next chapter compares JPA to serialization and other available persistence mechanisms.

Chapter 2.  Why JPA?

Java developers who need to store and retrieve persistent data already have several options available to them: serialization, JDBC, JDO, proprietary object-relational mapping tools, object databases, and EJB 2 entity beans. Why introduce yet another persistence framework? The answer to this question is that with the exception of JDO, each of the aforementioned persistence solutions has severe limitations. JPA attempts to overcome these limitations, as illustrated by the table below.

Table 2.1.  Persistence Mechanisms

Supports: Serialization JDBC ORM ODB EJB 2 JDO JPA
Java Objects Yes No Yes Yes Yes Yes Yes
Advanced OO Concepts Yes No Yes Yes No Yes Yes
Transactional Integrity No Yes Yes Yes Yes Yes Yes
Concurrency No Yes Yes Yes Yes Yes Yes
Large Data Sets No Yes Yes Yes Yes Yes Yes
Existing Schema No Yes Yes No Yes Yes Yes
Relational and Non-Relational Stores No No No No Yes Yes No
Queries No Yes Yes Yes Yes Yes Yes
Strict Standards / Portability Yes No No No Yes Yes Yes
Simplicity Yes Yes Yes Yes No Yes Yes
  • Serialization is Java's built-in mechanism for transforming an object graph into a series of bytes, which can then be sent over the network or stored in a file. Serialization is very easy to use, but it is also very limited. It must store and retrieve the entire object graph at once, making it unsuitable for dealing with large amounts of data. It cannot undo changes that are made to objects if an error occurs while updating information, making it unsuitable for applications that require strict data integrity. Multiple threads or programs cannot read and write the same serialized data concurrently without conflicting with each other. It provides no query capabilities. All these factors make serialization useless for all but the most trivial persistence needs.

  • Many developers use the Java Database Connectivity (JDBC) APIs to manipulate persistent data in relational databases. JDBC overcomes most of the shortcomings of serialization: it can handle large amounts of data, has mechanisms to ensure data integrity, supports concurrent access to information, and has a sophisticated query language in SQL. Unfortunately, JDBC does not duplicate serialization's ease of use. The relational paradigm used by JDBC was not designed for storing objects, and therefore forces you to either abandon object-oriented programming for the portions of your code that deal with persistent data, or to find a way of mapping object-oriented concepts like inheritance to relational databases yourself.

  • There are many proprietary software products that can perform the mapping between objects and relational database tables for you. These object-relational mapping (ORM) frameworks allow you to focus on the object model and not concern yourself with the mismatch between the object-oriented and relational paradigms. Unfortunately, each of these product has its own set of APIs. Your code becomes tied to the proprietary interfaces of a single vendor. If the vendor raises prices, fails to fix show-stopping bugs, or falls behind in features, you cannot switch to another product without rewriting all of your persistence code. This is referred to as vendor lock-in.

  • Rather than map objects to relational databases, some software companies have developed a form of database designed specifically to store objects. These object databases (ODBs) are often much easier to use than object-relational mapping software. The Object Database Management Group (ODMG) was formed to create a standard API for accessing object databases; few object database vendors, however, comply with the ODMG's recommendations. Thus, vendor lock-in plagues object databases as well. Many companies are also hesitant to switch from tried-and-true relational systems to the relatively unknown object database technology. Fewer data-analysis tools are available for object database systems, and there are vast quantities of data already stored in older relational databases. For all of these reasons and more, object databases have not caught on as well as their creators hoped.

  • The Enterprise Edition of the Java platform introduced entity Enterprise Java Beans (EJBs). EJB 2.x entities are components that represent persistent information in a datastore. Like object-relational mapping solutions, EJB 2.x entities provide an object-oriented view of persistent data. Unlike object-relational software, however, EJB 2.x entities are not limited to relational databases; the persistent information they represent may come from an Enterprise Information System (EIS) or other storage device. Also, EJB 2.x entities use a strict standard, making them portable across vendors. Unfortunately, the EJB 2.x standard is somewhat limited in the object-oriented concepts it can represent. Advanced features like inheritance, polymorphism, and complex relations are absent. Additionally, EBJ 2.x entities are difficult to code, and they require heavyweight and often expensive application servers to run.

  • The JDO specification uses an API that is strikingly similar to JPA. JDO, however, supports non-relational databases, a feature that some argue dilutes the specification.

JPA combines the best features from each of the persistence mechanisms listed above. Creating entities under JPA is as simple as creating serializable classes. JPA supports the large data sets, data consistency, concurrent use, and query capabilities of JDBC. Like object-relational software and object databases, JPA allows the use of advanced object-oriented concepts such as inheritance. JPA avoids vendor lock-in by relying on a strict specification like JDO and EJB 2.x entities. JPA focuses on relational databases. And like JDO, JPA is extremely easy to use.

Note

OpenJPA typically stores data in relational databases, but can be customized for use with non-relational datastores as well.

JPA is not ideal for every application. For many applications, though, it provides an exciting alternative to other persistence mechanisms.

Chapter 3.  Java Persistence API Architecture

Table of Contents

1. JPA Exceptions

The diagram below illustrates the relationships between the primary components of the JPA architecture.

Note

A number of the depicted interfaces are only required outside of an EJB3-compliant application server. In an application server, EntityManager instances are typically injected, rendering the EntityManagerFactory unnecessary. Also, transactions within an application server are handled using standard application server transaction controls. Thus, the EntityTransaction also goes unused.

  • Persistence: The javax.persistence.Persistence class contains static helper methods to obtain EntityManagerFactory instances in a vendor-neutral fashion.

  • EntityManagerFactory: The javax.persistence.EntityManagerFactory class is a factory for EntityManager s.

  • EntityManager : The javax.persistence.EntityManager is the primary JPA interface used by applications. Each EntityManager manages a set of persistent objects, and has APIs to insert new objects and delete existing ones. When used outside the container, there is a one-to-one relationship between an EntityManager and an EntityTransaction. EntityManagers also act as factories for Query instances.

  • Entity : Entites are persistent objects that represent datastore records.

  • EntityTransaction: Each EntityManager has a one-to-one relation with a single javax.persistence.EntityTransaction. EntityTransactions allow operations on persistent data to be grouped into units of work that either completely succeed or completely fail, leaving the datastore in its original state. These all-or-nothing operations are important for maintaining data integrity.

  • Query : The javax.persistence.Query interface is implemented by each JPA vendor to find persistent objects that meet certain criteria. JPA standardizes support for queries using both the Java Persistence Query Language (JPQL) and the Structured Query Language (SQL). You obtain Query instances from an EntityManager.

The example below illustrates how the JPA interfaces interact to execute a JPQL query and update persistent objects. The example assumes execution outside a container.

Example 3.1.  Interaction of Interfaces Outside Container

// get an EntityManagerFactory using the Persistence class; typically 
// the factory is cached for easy repeated use
EntityManagerFactory factory = Persistence.createEntityManagerFactory(null);

// get an EntityManager from the factory
EntityManager em = factory.createEntityManager(PersistenceContextType.EXTENDED);

// updates take place within transactions
EntityTransaction tx = em.getTransaction();
tx.begin();

// query for all employees who work in our research division
// and put in over 40 hours a week average
Query query = em.createQuery("select e from Employee e where "
    + "e.division.name = 'Research' AND e.avgHours > 40");
List results = query.getResultList ();

// give all those hard-working employees a raise
for (Object res : results) {
    Employee emp = (Employee) res;
    emp.setSalary(emp.getSalary() * 1.1);
}

// commit the updates and free resources
tx.commit();
em.close();
factory.close();

Within a container, the EntityManager will be injected and transactions will be handled declaratively. Thus, the in-container version of the example consists entirely of business logic:

Example 3.2.  Interaction of Interfaces Inside Container

// query for all employees who work in our research division
// and put in over 40 hours a week average - note that the EntityManager em
// is injected using a @Resource annotation
Query query = em.createQuery("select e from Employee e where "
    + "e.division.name = 'Research' and e.avgHours > 40");
List results = query.getResultList();

// give all those hard-working employees a raise
for (Object res : results) {
    emp = (Employee) res;
    emp.setSalary(emp.getSalary() * 1.1);
} 

The remainder of this document explores the JPA interfaces in detail. We present them in roughly the order that you will use them as you develop your application.

1.  JPA Exceptions

The diagram above depicts the JPA exception architecture. All exceptions are unchecked. JPA uses standard exceptions where appropriate, most notably IllegalArgumentExceptions and IllegalStateExceptions. The specification also provides a few JPA-specific exceptions in the javax.persistence package. These exceptions should be self-explanatory. See the Javadoc for additional details on JPA exceptions.

Note

All exceptions thrown by OpenJPA implement org.apache.openjpa.util.ExceptionInfo to provide you with additional error information.

Chapter 4.  Entity

JPA recognizes two types of persistent classes: entity classes and embeddable classes. Each persistent instance of an entity class - each entity - represents a unique datastore record. You can use the EntityManager to find an entity by its persistent identity (covered later in this chapter), or use a Query to find entities matching certain criteria.

An instance of an embeddable class, on the other hand, is only stored as part of a separate entity. Embeddable instances have no persistent identity, and are never returned directly from the EntityManager or from a Query.

Despite these differences, there are few distinctions between entity classes and embeddable classes. In fact, writing either type of persistent class is a lot like writing any other class. There are no special parent classes to extend from, field types to use, or methods to write. This is one important way in which JPA makes persistence transparent to you, the developer.

Note

JPA supports both fields and JavaBean properties as persistent state. For simplicity, however, we will refer to all persistent state as persistent fields, unless we want to note a unique aspect of persistent properties.

Example 4.1.  Persistent Class

package org.mag;

/**
 * Example persistent class.  Notice that it looks exactly like any other
 * class.  JPA makes writing persistent classes completely transparent.
 */
public class Magazine {

    private String isbn;
    private String title;
    private Set articles = new HashSet();
    private Article coverArticle;
    private int copiesSold;
    private double price;
    private Company publisher;
    private int version;

    protected Magazine() {
    }

    public Magazine(String title, String isbn) {
        this.title = title;
        this.isbn = isbn;
    }

    public void publish(Company publisher, double price) {
        this.publisher = publisher;
        publisher.addMagazine(this);
        this.price = price;
    }
    
    public void sell() {
        copiesSold++;
        publisher.addRevenue(price);
    }

    public void addArticle(Article article) {
        articles.add(article);
    }

    // rest of methods omitted
}

1.  Restrictions on Persistent Classes

There are very few restrictions placed on persistent classes. Still, it never hurts to familiarize yourself with exactly what JPA does and does not support.

1.1.  Default or No-Arg Constructor

The JPA specification requires that all persistent classes have a no-arg constructor. This constructor may be public or protected. Because the compiler automatically creates a default no-arg constructor when no other constructor is defined, only classes that define constructors must also include a no-arg constructor.

Note

OpenJPA's enhancer will automatically add a protected no-arg constructor to your class when required. Therefore, this restriction does not apply when using OpenJPA. See Section 2, “ Enhancement ” of the Reference Guide for details.

1.2.  Final

Entity classes may not be final. No method of an entity class can be final.

Note

OpenJPA supports final classes and final methods.

1.3.  Identity Fields

All entity classes must declare one or more fields which together form the persistent identity of an instance. These are called identity or primary key fields. In our Magazine class, isbn and title are identity fields, because no two magazine records in the datastore can have the same isbn and title values. Section 2.2, “ Id ” will show you how to denote your identity fields in JPA metadata. Section 2, “ Entity Identity ” below examines persistent identity.

Note

OpenJPA fully supports identity fields, but does not require them. See Section 3, “ Object Identity ” of the Reference Guide for details.

1.4.  Version Field

The version field in our Magazine class may seem out of place. JPA uses a version field in your entities to detect concurrent modifications to the same datastore record. When the JPA runtime detects an attempt to concurrently modify the same record, it throws an exception to the transaction attempting to commit last. This prevents overwriting the previous commit with stale data.

A version field is not required, but without one concurrent threads or processes might succeed in making conflicting changes to the same record at the same time. This is unacceptable to most applications. Section 2.5, “ Version ” shows you how to designate a version field in JPA metadata.

The version field must be an integral type ( int, Long, etc) or a java.sql.Timestamp. You should consider version fields immutable. Changing the field value has undefined results.

Note

OpenJPA fully supports version fields, but does not require them for concurrency detection. OpenJPA can maintain surrogate version values or use state comparisons to detect concurrent modifications. See Section 7, “ Additional JPA Mappings ” in the Reference Guide.

1.5.  Inheritance

JPA fully supports inheritance in persistent classes. It allows persistent classes to inherit from non-persistent classes, persistent classes to inherit from other persistent classes, and non-persistent classes to inherit from persistent classes. It is even possible to form inheritance hierarchies in which persistence skips generations. There are, however, a few important limitations:

  • Persistent classes cannot inherit from certain natively-implemented system classes such as java.net.Socket and java.lang.Thread.

  • If a persistent class inherits from a non-persistent class, the fields of the non-persistent superclass cannot be persisted.

  • All classes in an inheritance tree must use the same identity type. We cover entity identity in Section 2, “ Entity Identity ”.

1.6.  Persistent Fields

JPA manages the state of all persistent fields. Before you access persistent state, the JPA runtime makes sure that it has been loaded from the datastore. When you set a field, the runtime records that it has changed so that the new value will be persisted. This allows you to treat the field in exactly the same way you treat any other field - another aspect of JPA's transparency.

JPA does not support static or final fields. It does, however, include built-in support for most common field types. These types can be roughly divided into three categories: immutable types, mutable types, and relations.

Immutable types, once created, cannot be changed. The only way to alter a persistent field of an immutable type is to assign a new value to the field. JPA supports the following immutable types:

  • All primitives (int, float, byte, etc)

  • All primitive wrappers (java.lang.Integer, java.lang.Float, java.lang.Byte, etc)

  • java.lang.String

  • java.math.BigInteger

  • java.math.BigDecimal

JPA also supports byte[], Byte[], char[], and Character[] as immutable types. That is, you can persist fields of these types, but you should not manipulate individual array indexes without resetting the array into the persistent field.

Persistent fields of mutable types can be altered without assigning the field a new value. Mutable types can be modified directly through their own methods. The JPA specification requires that implementations support the following mutable field types:

  • java.util.Date

  • java.util.Calendar

  • java.sql.Date

  • java.sql.Timestamp

  • java.sql.Time

  • Enums

  • Entity types (relations between entities)

  • Embeddable types

  • java.util.Collections of entities

  • java.util.Sets of entities

  • java.util.Lists of entities

  • java.util.Maps in which each entry maps the value of one of a related entity's fields to that entity.

Collection and map types may be parameterized.

Most JPA implementations also have support for persisting serializable values as binary data in the datastore. Chapter 5, Metadata has more information on persisting serializable types.

Note

OpenJPA also supports arrays, java.lang.Number, java.util.Locale, all JDK 1.2 Set, List, and Map types, and many other mutable and immutable field types. OpenJPA also allows you to plug in support for custom types.

1.7.  Conclusions

This section detailed all of the restrictions JPA places on persistent classes. While it may seem like we presented a lot of information, you will seldom find yourself hindered by these restrictions in practice. Additionally, there are often ways of using JPA's other features to circumvent any limitations you run into.

2.  Entity Identity

Java recognizes two forms of object identity: numeric identity and qualitative identity. If two references are numerically identical, then they refer to the same JVM instance in memory. You can test for this using the == operator. Qualitative identity, on the other hand, relies on some user-defined criteria to determine whether two objects are "equal". You test for qualitative identity using the equals method. By default, this method simply relies on numeric identity.

JPA introduces another form of object identity, called entity identity or persistent identity. Entity identity tests whether two persistent objects represent the same state in the datastore.

The entity identity of each persistent instance is encapsulated in its identity field(s). If two entities of the same type have the same identity field values, then the two entities represent the same state in the datastore. Each entity's identity field values must be unique among all other entites of the same type.

Identity fields must be primitives, primitive wrappers, Strings, Dates, Timestamps, or embeddable types.

Note

OpenJPA supports entities as identity fields, as the Reference Guide discusses in Section 3.2, “ Entities as Identity Fields ”. For legacy schemas with binary primary key columns, OpenJPA also supports using identity fields of type byte[]. When you use a byte[] identity field, you must create an identity class. Identity classes are covered below.

Warning

Changing the fields of an embeddable instance while it is assigned to an identity field has undefined results. Always treat embeddable identity instances as immutable objects in your applications.

If you are dealing with a single persistence context (see Section 3, “ Persistence Context ”), then you do not have to compare identity fields to test whether two entity references represent the same state in the datastore. There is a much easier way: the == operator. JPA requires that each persistence context maintain only one JVM object to represent each unique datastore record. Thus, entity identity is equivalent to numeric identity within a persistence context. This is referred to as the uniqueness requirement.

The uniqueness requirement is extremely important - without it, it would be impossible to maintain data integrity. Think of what could happen if two different objects in the same transaction were allowed to represent the same persistent data. If you made different modifications to each of these objects, which set of changes should be written to the datastore? How would your application logic handle seeing two different "versions" of the same data? Thanks to the uniqueness requirement, these questions do not have to be answered.

2.1.  Identity Class

If your entity has only one identity field, you can use the value of that field as the entity's identity object in all EntityManager APIs. Otherwise, you must supply an identity class to use for identity objects. Your identity class must meet the following criteria:

  • The class must be public.

  • The class must be serializable.

  • The class must have a public no-args constructor.

  • The names of the non-static fields or properties of the class must be the same as the names of the identity fields or properties of the corresponding entity class, and the types must be identical.

  • The equals and hashCode methods of the class must use the values of all fields or properties corresponding to identity fields or properties in the entity class.

  • If the class is an inner class, it must be static.

  • All entity classes related by inheritance must use the same identity class, or else each entity class must have its own identity class whose inheritance hierarchy mirrors the inheritance hierarchy of the owning entity classes (see Section 2.1.1, “ Identity Hierarchies ”).

Note

Though you may still create identity classes by hand, OpenJPA provides the appidtool to automatically generate proper identity classes based on your identity fields. See Section 3.3, “ Application Identity Tool ” of the Reference Guide.

Example 4.2.  Identity Class

This example illustrates a proper identity class for an entity with multiple identity fields.

/**
 * Persistent class using application identity.
 */
public class Magazine {

    private String isbn;    // identity field
    private String title;   // identity field

    // rest of fields and methods omitted


    /**
     * Application identity class for Magazine.
     */
    public static class MagazineId {

        // each identity field in the Magazine class must have a
        // corresponding field in the identity class
        public String isbn;
        public String title;

        /**
         * Equality must be implemented in terms of identity field
         * equality, and must use instanceof rather than comparing 
         * classes directly (some JPA implementations may subclass the
         * identity class).
         */
        public boolean equals(Object other) {
            if (other == this)
                return true;
            if (!(other instanceof MagazineId))
                return false;
    
            MagazineId mi = (MagazineId) other;
            return (isbn == mi.isbn
                || (isbn != null && isbn.equals(mi.isbn)))
                && (title == mi.title
                || (title != null && title.equals(mi.title)));
        }
     
        /**
         * Hashcode must also depend on identity values.
         */
        public int hashCode() {
            return ((isbn == null) ? 0 : isbn.hashCode())
                ^ ((title == null) ? 0 : title.hashCode());
        } 

        public String toString() {
            return isbn + ":" + title;
        }
    }
}

2.1.1.  Identity Hierarchies

An alternative to having a single identity class for an entire inheritance hierarchy is to have one identity class per level in the inheritance hierarchy. The requirements for using a hierarchy of identity classes are as follows:

  • The inheritance hierarchy of identity classes must exactly mirror the hierarchy of the persistent classes that they identify. In the example pictured above, abstract class Person is extended by abstract class Employee, which is extended by non-abstract class FullTimeEmployee, which is extended by non-abstract class Manager. The corresponding identity classes, then, are an abstract PersonId class, extended by an abstract EmployeeId class, extended by a non-abstract FullTimeEmployeeId class, extended by a non-abstract ManagerId class.

  • Subclasses in the identity hierarchy may define additional identity fields until the hierarchy becomes non-abstract. In the aforementioned example, Person defines an identity field ssn, Employee defines additional identity field userName , and FullTimeEmployee adds a final identity field, empId. However, Manager may not define any additional identity fields, since it is a subclass of a non-abstract class. The hierarchy of identity classes, of course, must match the identity field definitions of the persistent class hierarchy.

  • It is not necessary for each abstract class to declare identity fields. In the previous example, the abstract Person and Employee classes could declare no identity fields, and the first concrete subclass FullTimeEmployee could define one or more identity fields.

  • All subclasses of a concrete identity class must be equals and hashCode-compatible with the concrete superclass. This means that in our example, a ManagerId instance and a FullTimeEmployeeId instance with the same identity field values should have the same hash code, and should compare equal to each other using the equals method of either one. In practice, this requirement reduces to the following coding practices:

    1. Use instanceof instead of comparing Class objects in the equals methods of your identity classes.

    2. An identity class that extends another non-abstract identity class should not override equals or hashCode.

3.  Lifecycle Callbacks

It is often necessary to perform various actions at different stages of a persistent object's lifecycle. JPA includes a variety of callbacks methods for monitoring changes in the lifecycle of your persistent objects. These callbacks can be defined on the persistent classes themselves and on non-persistent listener classes.

3.1.  Callback Methods

Every persistence event has a corresponding callback method marker. These markers are shared between persistent classes and their listeners. You can use these markers to designate a method for callback either by annotating that method or by listing the method in the XML mapping file for a given class. The lifecycle events and their corresponding method markers are:

  • PrePersist: Methods marked with this annotation will be invoked before an object is persisted. This could be used for assigning primary key values to persistent objects. This is equivalent to the XML element tag pre-persist.

  • PostPersist: Methods marked with this annotation will be invoked after an object has transitioned to the persistent state. You might want to use such methods to update a screen after a new row is added. This is equivalent to the XML element tag post-persist.

  • PostLoad: Methods marked with this annotation will be invoked after all eagerly fetched fields of your class have been loaded from the datastore. No other persistent fields can be accessed in this method. This is equivalent to the XML element tag post-load.

    PostLoad is often used to initialize non-persistent fields whose values depend on the values of persistent fields, such as a complex datastructure.

  • PreUpdate: Methods marked with this annotation will be invoked just the persistent values in your objects are flushed to the datastore. This is equivalent to the XML element tag pre-update.

    PreUpdate is the complement to PostLoad . While methods marked with PostLoad are most often used to initialize non-persistent values from persistent data, methods annotated with PreUpdate is normally used to set persistent fields with information cached in non-persistent data.

  • PostUpdate: Methods marked with this annotation will be invoked after changes to a given instance have been stored to the datastore. This is useful for clearing stale data cached at the application layer. This is equivalent to the XML element tag post-update.

  • PreRemove: Methods marked with this annotation will be invoked before an object transactions to the deleted state. Access to persistent fields is valid within this method. You might use this method to cascade the deletion to related objects based on complex criteria, or to perform other cleanup. This is equivalent to the XML element tag pre-remove.

  • PostRemove: Methods marked with this annotation will be invoked after an object has been marked as to be deleted. This is equivalent to the XML element tag post-remove.

3.2.  Using Callback Methods

When declaring callback methods on a persistent class, any method may be used which takes no arguments and is not shared with any property access fields. Multiple events can be assigned to a single method as well.

Below is an example of how to declare callback methods on persistent classes:

/**
 * Example persistent class declaring our entity listener.
 */
@Entity
public class Magazine {

    @Transient 
    private byte[][] data;

    @ManyToMany
    private List<Photo> photos;

    @PostLoad
    public void convertPhotos() {
        data = new byte[photos.size()][];
        for (int i = 0; i < photos.size(); i++)
            data[i] = photos.get(i).toByteArray();
    }

    @PreDelete
    public void logMagazineDeletion() {
        getLog().debug("deleting magazine containing" + photos.size() 
            + " photos.");
    }
}

In an XML mapping file, we can define the same methods without annotations:

<entity class="Magazine">
    <pre-remove>logMagazineDeletion</pre-remove>
    <post-load>convertPhotos</post-load>
</entity>

Note

We fully explore persistence metadata annotations and XML in Chapter 5, Metadata .

3.3.  Using Entity Listeners

Mixing lifecycle event code into your persistent classes is not always ideal. It is often more elegant to handle cross-cutting lifecycle events in a non-persistent listener class. JPA allows for this, requiring only that listener classes have a public no-arg constructor. Like persistent classes, your listener classes can consume any number of callbacks. The callback methods must take in a single java.lang.Object argument which represents the persistent object that triggered the event.

Entities can enumerate listeners using the EntityListeners annotation. This annotation takes an array of listener classes as its value.

Below is an example of how to declare an entity and its corresponding listener classes.

/**
 * Example persistent class declaring our entity listener.
 */
@Entity
@EntityListeners({ MagazineLogger.class, ... })
public class Magazine {

    // ... //
}


/**
 * Example entity listener.
 */
public class MagazineLogger {

    @PostPersist
    public void logAddition(Object pc) {
        getLog ().debug ("Added new magazine:" + ((Magazine) pc).getTitle ());
    }


    @PreRemove
    public void logDeletion(Object pc) {
        getLog().debug("Removing from circulation:" + 
            ((Magazine) pc).getTitle());
    }
}

In XML, we define both the listeners and their callback methods as so:

<entity class="Magazine">
    <entity-listeners>
        <entity-listener class="MagazineLogger">
            <post-persist>logAddition</post-persist>
            <pre-remove>logDeletion</pre-remove>
        </entity-listener>
    </entity-listeners>
</entity>

3.4.  Entity Listeners Hierarchy

Entity listener methods are invoked in a specific order when a given event is fired. So-called default listeners are invoked first: these are listeners which have been defined in a package annotation or in the root element of XML mapping files. Next, entity listeners are invoked in the order of the inheritance hierarchy, with superclass listeners being invoked before subclass listeners. Finally, if an entity has multiple listeners for the same event, the listeners are invoked in declaration order.

You can exclude default listeners and listeners defined in superclasses from the invocation chain through the use of two class-level annotations:

  • ExcludeDefaultListeners: This annotation indicates that no default listeners will be invoked for this class, or any of its subclasses. The XML equivalent is the empty exclude-default-listeners element.

  • ExcludeSuperclassListeners: This annotation will cause OpenJPA to skip invoking any listeners declared in superclasses. The XML equivalent is empty the exclude-superclass-listeners element.

4.  Conclusions

This chapter covered everything you need to know to write persistent class definitions in JPA. JPA cannot use your persistent classes, however, until you complete one additional step: you must define the persistence metadata. The next chapter explores metadata in detail.

Chapter 5.  Metadata

JPA requires that you accompany each persistent class with persistence metadata. This metadata serves three primary purposes:

  1. To identify persistent classes.

  2. To override default JPA behavior.

  3. To provide the JPA implementation with information that it cannot glean from simply reflecting on the persistent class.

Persistence metadata is specified using either the Java 5 annotations defined in the javax.persistence package, XML mapping files, or a mixture of both. In the latter case, XML declarations override conflicting annotations. If you choose to use XML metadata, the XML files must be available at development and runtime, and must be discoverable via either of two strategies:

  1. In a resource named orm.xml placed in a META-INF directory within a directory in your classpath or within a jar archive containing your persistent classes.

  2. Declared in your persistence.xml configuration file. In this case, each XML metadata file must be listed in a mapping-file element whose content is either a path to the given file or a resource location available to the class' class loader.

We describe the standard metadata annotations and XML equivalents throughout this chapter. The full schema for XML mapping files is available in Section 3, “ XML Schema ”. JPA also standardizes relational mapping metadata and named query metadata, which we discuss in Chapter 12, Mapping Metadata and Section 1.9, “ Named Queries ” respectively.

Note

OpenJPA defines many useful annotations beyond the standard set. See Section 2, “ Additional JPA Metadata ” and Section 3, “ Metadata Extensions ” in the Reference Guide for details. There are currently no XML equivalents for these extension annotations.

Through the course of this chapter, we will create the persistent object model above.

1.  Class Metadata

The following metadata annotations and XML elements apply to persistent class declarations.

1.1.  Entity

The Entity annotation denotes an entity class. All entity classes must have this annotation. The Entity annotation takes one optional property:

  • String name: Name used to refer to the entity in queries. Must not be a reserved literal in JPQL. Defaults to the unqualified name of the entity class.

The equivalent XML element is entity. It has the following attributes:

  • class: The entity class. This attribute is required.

  • name: Named used to refer to the class in queries. See the name property above.

  • access: The access type to use for the class. Must either be FIELD or PROPERTY. For details on access types, see Section 2, “ Field and Property Metadata ”.

Note

OpenJPA uses a process called enhancement to modify the bytecode of entities for transparent lazy loading and immediate dirty tracking. See Section 2, “ Enhancement ” in the Reference Guide for details on enhancement.

1.2.  Id Class

As we discussed in Section 2.1, “ Identity Class ”, entities with multiple identity fields must use an identity class to encapsulate their persistent identity. The IdClass annotation specifies this class. It accepts a single java.lang.Class value.

The equivalent XML element is id-class, which has a single attribute:

  • class: Set this required attribute to the name of the identity class.

1.3.  Mapped Superclass

A mapped superclass is a non-entity class that can define persistent state and mapping information for entity subclasses. Mapped superclasses are usually abstract. Unlike true entities, you cannot query a mapped superclass, pass a mapped superclass instance to any EntityManager or Query methods, or declare a persistent relation with a mapped superclass target. You denote a mapped superclass with the MappedSuperclass marker annotation.

The equivalent XML element is mapped-superclass. It expects the following attributes:

Note

OpenJPA allows you to query on mapped superclasses. A query on a mapped superclass will return all matching subclass instances. OpenJPA also allows you to declare relations to mapped superclass types; however, you cannot query across these relations.

1.4.  Embeddable

The Embeddable annotation designates an embeddable persistent class. Embeddable instances are stored as part of the record of their owning instance. All embeddable classes must have this annotation.

A persistent class can either be an entity or an embeddable class, but not both.

The equivalent XML element is embeddable. It understands the following attributes:

Note

OpenJPA allows a persistent class to be both an entity and an embeddable class. Instances of the class will act as entites when persisted explicitly or assigned to non-embedded fields of entities. Instances will act as embedded values when assigned to embedded fields of entities.

To signal that a class is both an entity and an embeddable class in OpenJPA, simply add both the @Entity and the @Embeddable annotations to the class.

1.5.  EntityListeners

An entity may list its lifecycle event listeners in the EntityListeners annotation. This value of this annotation is an array of the listener Class es for the entity. The equivalent XML element is entity-listeners. For more details on entity listeners, see Section 3, “ Lifecycle Callbacks ”.

1.6.  Example

Here are the class declarations for our persistent object model, annotated with the appropriate persistence metadata. Note that Magazine declares an identity class, and that Document and Address are a mapped superclass and an embeddable class, respectively. LifetimeSubscription and TrialSubscription override the default entity name to supply a shorter alias for use in queries.

Example 5.1.  Class Metadata

package org.mag;

@Entity
@IdClass(Magazine.MagazineId.class)
public class Magazine {
    ...

    public static class MagazineId {
        ...
    }
}

@Entity
public class Article {
    ...
}


package org.mag.pub;

@Entity
public class Company {
    ...
}

@Entity
public class Author {
    ...
}

@Embeddable
public class Address {
    ...
}


package org.mag.subscribe;

@MappedSuperclass
public abstract class Document {
    ...
}

@Entity
public class Contract
    extends Document {
    ...
}

@Entity
public class Subscription {
    ...

    @Entity
    public static class LineItem
        extends Contract {
        ...
    }
}

@Entity(name="Lifetime")
public class LifetimeSubscription
    extends Subscription {
    ...
}

@Entity(name="Trial")
public class TrialSubscription
    extends Subscription {
    ...
}

The equivalent declarations in XML:

<entity-mappings xmlns="http://java.sun.com/xml/ns/persistence/orm" 
    xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" 
    xsi:schemaLocation="http://java.sun.com/xml/ns/persistence/orm orm_1_0.xsd"
    version="1.0">
    <mapped-superclass class="org.mag.subscribe.Document">
        ...
    </mapped-superclass>
    <entity class="org.mag.Magazine">
        <id-class class="org.mag.Magazine$MagazineId"/>
        ...
    </entity>
    <entity class="org.mag.Article">
        ...
    </entity>
    <entity class="org.mag.pub.Company">
        ...
    </entity>
    <entity class="org.mag.pub.Author">
        ...
    </entity>
    <entity class="org.mag.subscribe.Contract">
        ...
    </entity>
    <entity class="org.mag.subscribe.LineItem">
        ...
    </entity>
    <entity class="org.mag.subscribe.LifetimeSubscription" name="Lifetime">
        ...
    </entity>
    <entity class="org.mag.subscribe.TrialSubscription" name="Trial">
        ...
    </entity>
    <embeddable class="org.mag.pub.Address">
        ...
    </embeddable>
</entity-mappings>

2.  Field and Property Metadata

The persistence implementation must be able to retrieve and set the persistent state of your entities, mapped superclasses, and embeddable types. JPA offers two modes of persistent state access: field access, and property access. Under field access, the implementation injects state directly into your persistent fields, and retrieves changed state from your fields as well. To declare field access on an entity with XML metadata, set the access attribute of your entity XML element to FIELD. To use field access for an entity using annotation metadata, simply place your metadata and mapping annotations on your field declarations:

@ManyToOne
private Company publisher;

Property access, on the other hand, retrieves and loads state through JavaBean "getter" and "setter" methods. For a property p of type T, you must define the following getter method:

T getP();

For boolean properties, this is also acceptable:

boolean isP();

You must also define the following setter method:

void setP(T value);

To use property access, set your entity element's access attribute to PROPERTY, or place your metadata and mapping annotations on the getter method:

@ManyToOne
private Company getPublisher() { ... }

private void setPublisher(Company publisher) { ... }

Warning

When using property access, only the getter and setter method for a property should ever access the underlying persistent field directly. Other methods, including internal business methods in the persistent class, should go through the getter and setter methods when manipulating persistent state.

Also, take care when adding business logic to your getter and setter methods. Consider that they are invoked by the persistence implementation to load and retrieve all persistent state; other side effects might not be desirable.

Each class must use either field access or property access for all state; you cannot use both access types within the same class. Additionally, a subclass must use the same access type as its superclass.

The remainder of this document uses the term "persistent field" to refer to either a persistent field or a persistent property.

2.1.  Transient

The Transient annotation specifies that a field is non-persistent. Use it to exclude fields from management that would otherwise be persistent. Transient is a marker annotation only; it has no properties.

The equivalent XML element is transient. It has a single attribute:

  • name: The transient field or property name. This attribute is required.

2.2.  Id

Annotate your simple identity fields with Id. This annotation has no properties. We explore entity identity and identity fields in Section 1.3, “ Identity Fields ”.

The equivalent XML element is id. It has one required attribute:

  • name: The name of the identity field or property.

2.3.  Generated Value

The previous section showed you how to declare your identity fields with the Id annotation. It is often convenient to allow the persistence implementation to assign a unique value to your identity fields automatically. JPA includes the the GeneratedValue annotation for this purpose. It has the following properties:

  • GenerationType strategy: Enum value specifying how to auto-generate the field value. The GenerationType enum has the following values:

    • GeneratorType.AUTO: The default. Assign the field a generated value, leaving the details to the JPA vendor.

    • GenerationType.IDENTITY: The database will assign an identity value on insert.

    • GenerationType.SEQUENCE: Use a datastore sequence to generate a field value.

    • GenerationType.TABLE: Use a sequence table to generate a field value.

  • String generator: The name of a generator defined in mapping metadata. We show you how to define named generators in Section 5, “ Generators ”. If the GenerationType is set but this property is unset, the JPA implementation uses appropriate defaults for the selected generation type.

The equivalent XML element is generated-value, which includes the following attributes:

  • strategy: One of TABLE, SEQUENCE, IDENTITY, or AUTO, defaulting to AUTO.

  • generator: Equivalent to the generator property listed above.

Note

OpenJPA allows you to use the GeneratedValue annotation on any field, not just identity fields. Before using the IDENTITY generation strategy, however, read Section 3.4, “ Autoassign / Identity Strategy Caveats ” in the Reference Guide.

OpenJPA also offers two additional generator strategies for non-numeric fields, which you can access by setting strategy to AUTO (the default), and setting the generator string to:

  • uuid-string: OpenJPA will generate a 128-bit UUID unique within the network, represented as a 16-character string. For more information on UUIDs, see the IETF UUID draft specification at: http://www1.ics.uci.edu/~ejw/authoring/uuid-guid/

  • uuid-hex: Same as uuid-string, but represents the UUID as a 32-character hexadecimal string.

These string constants are defined in org.apache.openjpa.persistence.Generator.

2.4.  Embedded Id

If your entity has multiple identity values, you may declare multiple @Id fields, or you may declare a single @EmbeddedId field. The type of a field annotated with EmbeddedId must be an embeddable entity class. The fields of this embeddable class are considered the identity values of the owning entity. We explore entity identity and identity fields in Section 1.3, “ Identity Fields ”.

The EmbeddedId annotation has no properties.

The equivalent XML element is embedded-id. It has one required attribute:

  • name: The name of the identity field or property.

2.5.  Version

Use the Version annotation to designate a version field. Section 1.4, “ Version Field ” explained the importance of version fields to JPA. This is a marker annotation; it has no properties.

The equivalent XML element is version, which has a single attribute:

  • name: The name of the version field or property. This attribute is required.

2.6.  Basic

Basic signifies a standard value persisted as-is to the datastore. You can use the Basic annotation on persistent fields of the following types: primitives, primitive wrappers, java.lang.String, byte[], Byte[], char[], Character[], java.math.BigDecimal, java.math.BigInteger, java.util.Date, java.util.Calendar, java.sql.Date, java.sql.Timestamp, Enums, and Serializable types.

Basic declares these properties:

  • FetchType fetch: Whether to load the field eagerly (FetchType.EAGER) or lazily ( FetchType.LAZY). Defaults to FetchType.EAGER.

  • boolean optional: Whether the datastore allows null values. Defaults to true.

The equivalent XML element is basic. It has the following attributes:

  • name: The name of the field or property. This attribute is required.

  • fetch: One of EAGER or LAZY .

  • optional: Boolean indicating whether the field value may be null.

2.6.1.  Fetch Type

Many metadata annotations in JPA have a fetch property. This property can take on one of two values: FetchType.EAGER or FetchType.LAZY. FetchType.EAGER means that the field is loaded by the JPA implementation before it returns the persistent object to you. Whenever you retrieve an entity from a query or from the EntityManager, you are guaranteed that all of its eager fields are populated with datastore data.

FetchType.LAZY is a hint to the JPA runtime that you want to defer loading of the field until you access it. This is called lazy loading. Lazy loading is completely transparent; when you attempt to read the field for the first time, the JPA runtime will load the value from the datastore and populate the field automatically. Lazy loading is only a hint and not a directive because some JPA implementations cannot lazy-load certain field types.

With a mix of eager and lazily-loaded fields, you can ensure that commonly-used fields load efficiently, and that other state loads transparently when accessed. As you will see in Section 3, “ Persistence Context ”, you can also use eager fetching to ensure that entites have all needed data loaded before they become detached at the end of a persistence context.

Note

OpenJPA can lazy-load any field type. OpenJPA also allows you to dynamically change which fields are eagerly or lazily loaded at runtime. See Section 6, “ Fetch Groups ” in the Reference Guide for details.

The Reference Guide details OpenJPA's eager fetching behavior in Section 7, “ Eager Fetching ”.

2.7.  Embedded

Use the Embedded marker annotation on embeddable field types. Embedded fields are mapped as part of the datastore record of the declaring entity. In our sample model, Author and Company each embed their Address, rather than forming a relation to an Address as a separate entity.

The equivalent XML element is embedded, which expects a single attribute:

  • name: The name of the field or property. This attribute is required.

2.8.  Many To One

When an entity A references a single entity B, and other As might also reference the same B, we say there is a many to one relation from A to B. In our sample model, for example, each magazine has a reference to its publisher. Multiple magazines might have the same publisher. We say, then, that the Magazine.publisher field is a many to one relation from magazines to publishers.

JPA indicates many to one relations between entities with the ManyToOne annotation. This annotation has the following properties:

  • Class targetEntity: The class of the related entity type.

  • CascadeType[] cascade: Array of enum values defining cascade behavior for this field. We explore cascades below. Defaults to an empty array.

  • FetchType fetch: Whether to load the field eagerly (FetchType.EAGER) or lazily (FetchType.LAZY). Defaults to FetchType.EAGER. See Section 2.6.1, “ Fetch Type ” above for details on fetch types.

  • boolean optional: Whether the related object must exist. If false, this field cannot be null. Defaults to true.

The equivalent XML element is many-to-one. It accepts the following attributes:

  • name: The name of the field or property. This attribute is required.

  • target-entity: The class of the related type.

  • fetch: One of EAGER or LAZY.

  • optional: Boolean indicating whether the field value may be null.

2.8.1.  Cascade Type

We introduce the JPA EntityManager in Chapter 8, EntityManager . The EntityManager has APIs to persist new entities, remove (delete) existing entities, refresh entity state from the datastore, and merge detached entity state back into the persistence context. We explore all of these APIs in detail later in the overview.

When the EntityManager is performing the above operations, you can instruct it to automatically cascade the operation to the entities held in a persistent field with the cascade property of your metadata annotation. This process is recursive. The cascade property accepts an array of CascadeType enum values.

  • CascadeType.PERSIST: When persisting an entity, also persist the entities held in this field. We suggest liberal application of this cascade rule, because if the EntityManager finds a field that references a new entity during flush, and the field does not use CascadeType.PERSIST, it is an error.

  • CascadeType.REMOVE: When deleting an entity, also delete the entities held in this field.

  • CascadeType.REFRESH: When refreshing an entity, also refresh the entities held in this field.

  • CascadeType.MERGE: When merging entity state, also merge the entities held in this field.

Note

OpenJPA offers enhancements to JPA's CascadeType.REMOVE functionality, including additional annotations to control how and when dependent fields will be removed. See Section 3.2.1, “ Dependent ” for more details.

CascadeType defines one additional value, CascadeType.ALL, that acts as a shortcut for all of the values above. The following annotations are equivalent:

@ManyToOne(cascade={CascadeType.PERSIST,CascadeType.REMOVE,
    CascadeType.REFRESH,CascadeType.MERGE})
private Company publisher;
@ManyToOne(cascade=CascadeType.ALL)
private Company publisher;

In XML, these enumeration constants are available as child elements of the cascade element. The cascade element is itself a child of many-to-one. The following examples are equivalent:

<many-to-one name="publisher">
    <cascade>
        <cascade-persist/>
        <cascade-merge/>
        <cascade-remove/>
        <cascade-refresh/>
    </cascade>
</many-to-one>
<many-to-one name="publisher">
    <cascade>
        <cascade-all/>
    </cascade>
</many-to-one>

2.9.  One To Many

When an entity A references multiple B entities, and no two As reference the same B, we say there is a one to many relation from A to B.

One to many relations are the exact inverse of the many to one relations we detailed in the preceding section. In that section, we said that the Magazine.publisher field is a many to one relation from magazines to publishers. Now, we see that the Company.mags field is the inverse - a one to many relation from publishers to magazines. Each company may publish multiple magazines, but each magazine can have only one publisher.

JPA indicates one to many relations between entities with the OneToMany annotation. This annotation has the following properties:

  • Class targetEntity: The class of the related entity type. This information is usually taken from the parameterized collection or map element type. You must supply it explicitly, however, if your field isn't a parameterized type.

  • String mappedBy: Names the many to one field in the related entity that maps this bidirectional relation. We explain bidirectional relations below. Leaving this property unset signals that this is a standard unidirectional relation.

  • CascadeType[] cascade: Array of enum values defining cascade behavior for the collection elements. We explore cascades above in Section 2.8.1, “ Cascade Type ”. Defaults to an empty array.

  • FetchType fetch: Whether to load the field eagerly (FetchType.EAGER) or lazily (FetchType.LAZY). Defaults to FetchType.LAZY. See Section 2.6.1, “ Fetch Type ” above for details on fetch types.

The equivalent XML element is one-to-many, which includes the following attributes:

  • name: The name of the field or property. This attribute is required.

  • target-entity: The class of the related type.

  • fetch: One of EAGER or LAZY.

  • mapped-by: The name of the field or property that owns the relation. See Section 2, “ Field and Property Metadata ”.

You may also nest the cascade element within a one-to-many element.

2.9.1.  Bidirectional Relations

When two fields are logical inverses of each other, they form a bidirectional relation. Our model contains two bidirectional relations: Magazine.publisher and Company.mags form one bidirectional relation, and Article.authors and Author.articles form the other. In both cases, there is a clear link between the two fields that form the relationship. A magazine refers to its publisher while the publisher refers to all its published magazines. An article refers to its authors while each author refers to her written articles.

When the two fields of a bidirectional relation share the same datastore mapping, JPA formalizes the connection with the mappedBy property. Marking Company.mags as mappedBy Magazine.publisher means two things:

  1. Company.mags uses the datastore mapping for Magazine.publisher, but inverses it. In fact, it is illegal to specify any additional mapping information when you use the mappedBy property. All mapping information is read from the referenced field. We explore mapping in depth in Chapter 12, Mapping Metadata .

  2. Magazine.publisher is the "owner" of the relation. The field that specifies the mapping data is always the owner. This means that changes to the Magazine.publisher field are reflected in the datastore, while changes to the Company.mags field alone are not. Changes to Company.mags may still affect the JPA implementation's cache, however. Thus, it is very important that you keep your object model consistent by properly maintaining both sides of your bidirectional relations at all times.

You should always take advantage of the mappedBy property rather than mapping each field of a bidirectional relation independently. Failing to do so may result in the JPA implementation trying to update the database with conflicting data. Be careful to only mark one side of the relation as mappedBy, however. One side has to actually do the mapping!

Note

You can configure OpenJPA to automatically synchronize both sides of a bidirectional relation, or to perform various actions when it detects inconsistent relations. See Section 4, “ Managed Inverses ” in the Reference Guide for details.

2.10.  One To One

When an entity A references a single entity B, and no other As can reference the same B, we say there is a one to one relation between A and B. In our sample model, Magazine has a one to one relation to Article through the Magazine.coverArticle field. No two magazines can have the same cover article.

JPA indicates one to one relations between entities with the OneToOne annotation. This annotation has the following properties:

  • Class targetEntity: The class of the related entity type. This information is usually taken from the field type.

  • String mappedBy: Names the field in the related entity that maps this bidirectional relation. We explain bidirectional relations in Section 2.9.1, “ Bidirectional Relations ” above. Leaving this property unset signals that this is a standard unidirectional relation.

  • CascadeType[] cascade: Array of enum values defining cascade behavior for this field. We explore cascades in Section 2.8.1, “ Cascade Type ” above. Defaults to an empty array.

  • FetchType fetch: Whether to load the field eagerly (FetchType.EAGER) or lazily (FetchType.LAZY). Defaults to FetchType.EAGER. See Section 2.6.1, “ Fetch Type ” above for details on fetch types.

  • boolean optional: Whether the related object must exist. If false, this field cannot be null. Defaults to true.

The equivalent XML element is one-to-one which understands the following attributes:

  • name: The name of the field or property. This attribute is required.

  • target-entity: The class of the related type.

  • fetch: One of EAGER or LAZY.

  • mapped-by: The field that owns the relation. See Section 2, “ Field and Property Metadata ”.

You may also nest the cascade element within a one-to-one element.

2.11.  Many To Many

When an entity A references multiple B entities, and other As might reference some of the same Bs, we say there is a many to many relation between A and B. In our sample model, for example, each article has a reference to all the authors that contributed to the article. Other articles might have some of the same authors. We say, then, that Article and Author have a many to many relation through the Article.authors field.

JPA indicates many to many relations between entities with the ManyToMany annotation. This annotation has the following properties:

  • Class targetEntity: The class of the related entity type. This information is usually taken from the parameterized collection or map element type. You must supply it explicitly, however, if your field isn't a parameterized type.

  • String mappedBy: Names the many to many field in the related entity that maps this bidirectional relation. We explain bidirectional relations in Section 2.9.1, “ Bidirectional Relations ” above. Leaving this property unset signals that this is a standard unidirectional relation.

  • CascadeType[] cascade: Array of enum values defining cascade behavior for the collection elements. We explore cascades above in Section 2.8.1, “ Cascade Type ”. Defaults to an empty array.

  • FetchType fetch: Whether to load the field eagerly (FetchType.EAGER) or lazily (FetchType.LAZY). Defaults to FetchType.LAZY. See Section 2.6.1, “ Fetch Type ” above for details on fetch types.

The equivalent XML element is many-to-many. It accepts the following attributes:

  • name: The name of the field or property. This attribute is required.

  • target-entity: The class of the related type.

  • fetch: One of EAGER or LAZY.

  • mapped-by: The field that owns the relation. See Section 2, “ Field and Property Metadata ”.

You may also nest the cascade element within a many-to-many element.

2.12.  Order By

Datastores such as relational databases do not preserve the order of records. Your persistent List fields might be ordered one way the first time you retrieve an object from the datastore, and a completely different way the next. To ensure consistent ordering of collection fields, you must use the OrderBy annotation. The OrderBy annotation's value is a string defining the order of the collection elements. An empty value means to sort on the identity value(s) of the elements in ascending order. Any other value must be of the form:

<field name>[ ASC|DESC][, ...]

Each <field name> is the name of a persistent field in the collection's element type. You can optionally follow each field by the keyword ASC for ascending order, or DESC for descending order. If the direction is omitted, it defaults to ascending.

The equivalent XML element is order-by which can be listed as a sub-element of the one-to-many or many-to-many elements. The text within this element is parsed as the order by string.

Note

OpenJPA expands the available ordering syntax. See ??? in the Reference Guide for details.

2.13.  Map Key

JPA supports persistent Map fields through either a OneToMany or ManyToMany association. The related entities form the map values. JPA derives the map keys by extracting a field from each entity value. The MapKey annotation designates the field that is used as the key. It has the following properties:

  • String name: The name of a field in the related entity class to use as the map key. If no name is given, defaults to the identity field of the related entity class.

The equivalent XML element is map-key which can be listed as a sub-element of the one-to-many or many-to-many elements. The map-key element has the following attributes:

  • name: The name of the field in the related entity class to use as the map key.

2.14.  Persistent Field Defaults

In the absence of any of the annotations above, JPA defines the following default behavior for declared fields:

  1. Fields declared static, transient, or final default to non-persistent.

  2. Fields of any primitive type, primitive wrapper type, java.lang.String, byte[], Byte[], char[], Character[], java.math.BigDecimal, java.math.BigInteger, java.util.Date, java.util.Calendar, java.sql.Date, java.sql.Timestamp, or any Serializable type default to persistent, as if annotated with @Basic.

  3. Fields of an embeddable type default to persistent, as if annotated with @Embedded.

  4. All other fields default to non-persistent.

Note that according to these defaults, all relations between entities must be annotated explicitly. Without an annotation, a relation field will default to serialized storage if the related entity type is serializable, or will default to being non-persistent if not.

3.  XML Schema

We present the complete XML schema below. Many of the elements relate to object/relational mapping rather than metadata; these elements are discussed in Chapter 12, Mapping Metadata .

<?xml version="1.0" encoding="UTF-8"?>
<xsd:schema targetNamespace="http://java.sun.com/xml/ns/persistence/orm" 
  xmlns:orm="http://java.sun.com/xml/ns/persistence/orm" 
  xmlns:xsd="http://www.w3.org/2001/XMLSchema" 
  elementFormDefault="qualified" 
  attributeFormDefault="unqualified" 
  version="1.0">

  <xsd:annotation>
    <xsd:documentation>
      @(#)orm_1_0.xsd 1.0  Feb 14 2006
    </xsd:documentation>
  </xsd:annotation>
  <xsd:annotation>
     <xsd:documentation>

       This is the XML Schema for the persistence object-relational 
       mapping file.
       The file may be named "META-INF/orm.xml" in the persistence 
       archive or it may be named some other name which would be 
       used to locate the file as resource on the classpath.

     </xsd:documentation>
  </xsd:annotation>

  <xsd:complexType name="emptyType"/>

  <xsd:simpleType name="versionType">
    <xsd:restriction base="xsd:token">
      <xsd:pattern value="[0-9]+(\.[0-9]+)*"/>
    </xsd:restriction>
  </xsd:simpleType>

  <!-- **************************************************** -->

  <xsd:element name="entity-mappings">
    <xsd:complexType>
      <xsd:annotation>
        <xsd:documentation>

        The entity-mappings element is the root element of an mapping
        file. It contains the following four types of elements:

        1. The persistence-unit-metadata element contains metadata
        for the entire persistence unit. It is undefined if this element
        occurs in multiple mapping files within the same persistence unit.
        
        2. The package, schema, catalog and access elements apply to all of
        the entity, mapped-superclass and embeddable elements defined in
        the same file in which they occur.

        3. The sequence-generator, table-generator, named-query,
        named-native-query and sql-result-set-mapping elements are global
        to the persistence unit. It is undefined to have more than one
        sequence-generator or table-generator of the same name in the same
        or different mapping files in a persistence unit. It is also 
        undefined to have more than one named-query or named-native-query
        of the same name in the same or different mapping files in a 
        persistence unit.

        4. The entity, mapped-superclass and embeddable elements each define
        the mapping information for a managed persistent class. The mapping
        information contained in these elements may be complete or it may
        be partial.

        </xsd:documentation>
      </xsd:annotation>
      <xsd:sequence>
        <xsd:element name="description" type="xsd:string"
                     minOccurs="0"/>
        <xsd:element name="persistence-unit-metadata" 
                     type="orm:persistence-unit-metadata"
                     minOccurs="0"/>
        <xsd:element name="package" type="xsd:string"
                     minOccurs="0"/>
        <xsd:element name="schema" type="xsd:string"
                     minOccurs="0"/>
        <xsd:element name="catalog" type="xsd:string"
                     minOccurs="0"/>
        <xsd:element name="access" type="orm:access-type"
                     minOccurs="0"/>
        <xsd:element name="sequence-generator" type="orm:sequence-generator"
                     minOccurs="0" maxOccurs="unbounded"/>
        <xsd:element name="table-generator" type="orm:table-generator" 
                     minOccurs="0" maxOccurs="unbounded"/>
        <xsd:element name="named-query" type="orm:named-query" 
                     minOccurs="0" maxOccurs="unbounded"/>
        <xsd:element name="named-native-query" type="orm:named-native-query"
                     minOccurs="0" maxOccurs="unbounded"/>
        <xsd:element name="sql-result-set-mapping" 
                     type="orm:sql-result-set-mapping" 
                     minOccurs="0" maxOccurs="unbounded"/>
        <xsd:element name="mapped-superclass" type="orm:mapped-superclass" 
                     minOccurs="0" maxOccurs="unbounded"/>
        <xsd:element name="entity" type="orm:entity" 
                     minOccurs="0" maxOccurs="unbounded"/>
        <xsd:element name="embeddable" type="orm:embeddable" 
                     minOccurs="0" maxOccurs="unbounded"/>
      </xsd:sequence>
      <xsd:attribute name="version" type="orm:versionType" 
                     fixed="1.0" use="required"/>
    </xsd:complexType>
  </xsd:element>

  <!-- **************************************************** -->

  <xsd:complexType name="persistence-unit-metadata">
    <xsd:annotation>
      <xsd:documentation>

        Metadata that applies to the persistence unit and not just to 
        the mapping file in which it is contained. 

        If the xml-mapping-metadata-complete element is specified then 
        the complete set of mapping metadata for the persistence unit 
        is contained in the XML mapping files for the persistence unit.

      </xsd:documentation>
    </xsd:annotation>
    <xsd:sequence>
      <xsd:element name="xml-mapping-metadata-complete" type="orm:emptyType"
                   minOccurs="0"/>
      <xsd:element name="persistence-unit-defaults" 
                   type="orm:persistence-unit-defaults"
                   minOccurs="0"/>
    </xsd:sequence>
  </xsd:complexType>

  <!-- **************************************************** -->

  <xsd:complexType name="persistence-unit-defaults">
    <xsd:annotation>
      <xsd:documentation>

        These defaults are applied to the persistence unit as a whole 
        unless they are overridden by local annotation or XML 
        element settings. 
        
        schema - Used as the schema for all tables or secondary tables
            that apply to the persistence unit
        catalog - Used as the catalog for all tables or secondary tables
            that apply to the persistence unit
        access - Used as the access type for all managed classes in
            the persistence unit
        cascade-persist - Adds cascade-persist to the set of cascade options
            in entity relationships of the persistence unit
        entity-listeners - List of default entity listeners to be invoked 
            on each entity in the persistence unit. 

      </xsd:documentation>
    </xsd:annotation>
    <xsd:sequence>
        <xsd:element name="schema" type="xsd:string"
                     minOccurs="0"/>
        <xsd:element name="catalog" type="xsd:string"
                     minOccurs="0"/>
        <xsd:element name="access" type="orm:access-type"
                     minOccurs="0"/>
        <xsd:element name="cascade-persist" type="orm:emptyType" 
                     minOccurs="0"/>
        <xsd:element name="entity-listeners" type="orm:entity-listeners"
                     minOccurs="0"/>
    </xsd:sequence>
  </xsd:complexType>

  <!-- **************************************************** -->

  <xsd:complexType name="entity">
    <xsd:annotation>
      <xsd:documentation>

        Defines the settings and mappings for an entity. Is allowed to be
        sparsely populated and used in conjunction with the annotations.
        Alternatively, the metadata-complete attribute can be used to 
        indicate that no annotations on the entity class (and its fields
        or properties) are to be processed. If this is the case then 
        the defaulting rules for the entity and its subelements will 
        be recursively applied.

        @Target(TYPE) @Retention(RUNTIME)
          public @interface Entity {
          String name() default "";
        }

      </xsd:documentation>
    </xsd:annotation>
    <xsd:sequence>
      <xsd:element name="description" type="xsd:string" minOccurs="0"/>
      <xsd:element name="table" type="orm:table" 
                   minOccurs="0"/>
      <xsd:element name="secondary-table" type="orm:secondary-table" 
                   minOccurs="0" maxOccurs="unbounded"/>
      <xsd:element name="primary-key-join-column" 
                   type="orm:primary-key-join-column" 
                   minOccurs="0" maxOccurs="unbounded"/>
      <xsd:element name="id-class" type="orm:id-class" minOccurs="0"/>
      <xsd:element name="inheritance" type="orm:inheritance" minOccurs="0"/>
      <xsd:element name="discriminator-value" type="orm:discriminator-value" 
                   minOccurs="0"/>
      <xsd:element name="discriminator-column" 
                   type="orm:discriminator-column" 
                   minOccurs="0"/>
      <xsd:element name="sequence-generator" type="orm:sequence-generator" 
                   minOccurs="0"/>
      <xsd:element name="table-generator" type="orm:table-generator" 
                   minOccurs="0"/>
      <xsd:element name="named-query" type="orm:named-query" 
                   minOccurs="0" maxOccurs="unbounded"/>
      <xsd:element name="named-native-query" type="orm:named-native-query" 
                   minOccurs="0" maxOccurs="unbounded"/>
      <xsd:element name="sql-result-set-mapping" 
                   type="orm:sql-result-set-mapping" 
                   minOccurs="0" maxOccurs="unbounded"/>
      <xsd:element name="exclude-default-listeners" type="orm:emptyType" 
                   minOccurs="0"/>
      <xsd:element name="exclude-superclass-listeners" type="orm:emptyType" 
                   minOccurs="0"/>
      <xsd:element name="entity-listeners" type="orm:entity-listeners" 
                   minOccurs="0"/>
      <xsd:element name="pre-persist" type="orm:pre-persist" minOccurs="0"/>
      <xsd:element name="post-persist" type="orm:post-persist" 
                   minOccurs="0"/>
      <xsd:element name="pre-remove" type="orm:pre-remove" minOccurs="0"/>
      <xsd:element name="post-remove" type="orm:post-remove" minOccurs="0"/>
      <xsd:element name="pre-update" type="orm:pre-update" minOccurs="0"/>
      <xsd:element name="post-update" type="orm:post-update" minOccurs="0"/>
      <xsd:element name="post-load" type="orm:post-load" minOccurs="0"/>
      <xsd:element name="attribute-override" type="orm:attribute-override" 
                   minOccurs="0" maxOccurs="unbounded"/>
      <xsd:element name="association-override" 
                   type="orm:association-override"
                   minOccurs="0" maxOccurs="unbounded"/>
      <xsd:element name="attributes" type="orm:attributes" minOccurs="0"/>
    </xsd:sequence>
    <xsd:attribute name="name" type="xsd:string"/>
    <xsd:attribute name="class" type="xsd:string" use="required"/>
    <xsd:attribute name="access" type="orm:access-type"/>
    <xsd:attribute name="metadata-complete" type="xsd:boolean"/>
  </xsd:complexType>

  <!-- **************************************************** -->

  <xsd:complexType name="attributes">
    <xsd:annotation>
      <xsd:documentation>

        This element contains the entity field or property mappings.
        It may be sparsely populated to include only a subset of the
        fields or properties. If metadata-complete for the entity is true
        then the remainder of the attributes will be defaulted according
        to the default rules.

      </xsd:documentation>
    </xsd:annotation>
    <xsd:sequence>
      <xsd:choice>
        <xsd:element name="id" type="orm:id" 
                     minOccurs="0" maxOccurs="unbounded"/>
        <xsd:element name="embedded-id" type="orm:embedded-id" 
                     minOccurs="0"/>
      </xsd:choice>
      <xsd:element name="basic" type="orm:basic"
                   minOccurs="0" maxOccurs="unbounded"/>
      <xsd:element name="version" type="orm:version"
                   minOccurs="0" maxOccurs="unbounded"/>
      <xsd:element name="many-to-one" type="orm:many-to-one"
                   minOccurs="0" maxOccurs="unbounded"/>
      <xsd:element name="one-to-many" type="orm:one-to-many"
                   minOccurs="0" maxOccurs="unbounded"/>
      <xsd:element name="one-to-one" type="orm:one-to-one"
                   minOccurs="0" maxOccurs="unbounded"/>
      <xsd:element name="many-to-many" type="orm:many-to-many" 
                   minOccurs="0" maxOccurs="unbounded"/>
      <xsd:element name="embedded" type="orm:embedded"
                   minOccurs="0" maxOccurs="unbounded"/>
      <xsd:element name="transient" type="orm:transient"
                   minOccurs="0" maxOccurs="unbounded"/>
    </xsd:sequence>
  </xsd:complexType>

  <!-- **************************************************** -->

  <xsd:simpleType name="access-type">
    <xsd:annotation>
      <xsd:documentation>

        This element determines how the persistence provider accesses the
        state of an entity or embedded object.

      </xsd:documentation>
    </xsd:annotation>
    <xsd:restriction base="xsd:token">
      <xsd:enumeration value="PROPERTY"/>
      <xsd:enumeration value="FIELD"/>
    </xsd:restriction>
  </xsd:simpleType>

  <!-- **************************************************** -->

  <xsd:complexType name="entity-listeners">
    <xsd:annotation>
      <xsd:documentation>

        @Target({TYPE}) @Retention(RUNTIME)
        public @interface EntityListeners {
          Class[] value();
        }

      </xsd:documentation>
    </xsd:annotation>
    <xsd:sequence>
      <xsd:element name="entity-listener" type="orm:entity-listener" 
                   minOccurs="0" maxOccurs="unbounded"/>
    </xsd:sequence>
  </xsd:complexType>

  <!-- **************************************************** -->

  <xsd:complexType name="entity-listener">
    <xsd:annotation>
      <xsd:documentation>

        Defines an entity listener to be invoked at lifecycle events
        for the entities that list this listener.

      </xsd:documentation>
    </xsd:annotation>
    <xsd:sequence>
      <xsd:element name="pre-persist" type="orm:pre-persist" minOccurs="0"/>
      <xsd:element name="post-persist" type="orm:post-persist" 
                   minOccurs="0"/>
      <xsd:element name="pre-remove" type="orm:pre-remove" minOccurs="0"/>
      <xsd:element name="post-remove" type="orm:post-remove" minOccurs="0"/>
      <xsd:element name="pre-update" type="orm:pre-update" minOccurs="0"/>
      <xsd:element name="post-update" type="orm:post-update" minOccurs="0"/>
      <xsd:element name="post-load" type="orm:post-load" minOccurs="0"/>
    </xsd:sequence>
    <xsd:attribute name="class" type="xsd:string" use="required"/>
  </xsd:complexType>

  <!-- **************************************************** -->

  <xsd:complexType name="pre-persist">
    <xsd:annotation>
      <xsd:documentation>

        @Target({METHOD}) @Retention(RUNTIME)
        public @interface PrePersist {}

      </xsd:documentation>
    </xsd:annotation>
    <xsd:attribute name="method-name" type="xsd:string" use="required"/>
  </xsd:complexType>

  <!-- **************************************************** -->

  <xsd:complexType name="post-persist">
    <xsd:annotation>
      <xsd:documentation>

        @Target({METHOD}) @Retention(RUNTIME)
        public @interface PostPersist {}

      </xsd:documentation>
    </xsd:annotation>
    <xsd:attribute name="method-name" type="xsd:string" use="required"/>
  </xsd:complexType>

  <!-- **************************************************** -->

  <xsd:complexType name="pre-remove">
    <xsd:annotation>
      <xsd:documentation>

        @Target({METHOD}) @Retention(RUNTIME)
        public @interface PreRemove {}

      </xsd:documentation>
    </xsd:annotation>
    <xsd:attribute name="method-name" type="xsd:string" use="required"/>
  </xsd:complexType>

  <!-- **************************************************** -->

  <xsd:complexType name="post-remove">
    <xsd:annotation>
      <xsd:documentation>

        @Target({METHOD}) @Retention(RUNTIME)
        public @interface PostRemove {}

      </xsd:documentation>
    </xsd:annotation>
    <xsd:attribute name="method-name" type="xsd:string" use="required"/>
  </xsd:complexType>

  <!-- **************************************************** -->

  <xsd:complexType name="pre-update">
    <xsd:annotation>
      <xsd:documentation>

        @Target({METHOD}) @Retention(RUNTIME)
        public @interface PreUpdate {}

      </xsd:documentation>
    </xsd:annotation>
    <xsd:attribute name="method-name" type="xsd:string" use="required"/>
  </xsd:complexType>

  <!-- **************************************************** -->

  <xsd:complexType name="post-update">
    <xsd:annotation>
      <xsd:documentation>

        @Target({METHOD}) @Retention(RUNTIME)
        public @interface PostUpdate {}

      </xsd:documentation>
    </xsd:annotation>
    <xsd:attribute name="method-name" type="xsd:string" use="required"/>
  </xsd:complexType>

  <!-- **************************************************** -->

  <xsd:complexType name="post-load">
    <xsd:annotation>
      <xsd:documentation>

        @Target({METHOD}) @Retention(RUNTIME)
        public @interface PostLoad {}

      </xsd:documentation>
    </xsd:annotation>
    <xsd:attribute name="method-name" type="xsd:string" use="required"/>
  </xsd:complexType>

  <!-- **************************************************** -->

  <xsd:complexType name="query-hint">
    <xsd:annotation>
      <xsd:documentation>

        @Target({}) @Retention(RUNTIME) 
        public @interface QueryHint {
          String name();
          String value();
        }

      </xsd:documentation>
    </xsd:annotation>
    <xsd:attribute name="name" type="xsd:string" use="required"/>
    <xsd:attribute name="value" type="xsd:string" use="required"/>
  </xsd:complexType>

  <!-- **************************************************** -->

  <xsd:complexType name="named-query">
    <xsd:annotation>
      <xsd:documentation>

        @Target({TYPE}) @Retention(RUNTIME)
        public @interface NamedQuery {
          String name();
          String query();
          QueryHint[] hints() default {};
        }

      </xsd:documentation>
    </xsd:annotation>
    <xsd:sequence>
      <xsd:element name="query" type="xsd:string"/>
      <xsd:element name="hint" type="orm:query-hint" 
                   minOccurs="0" maxOccurs="unbounded"/>
    </xsd:sequence>
    <xsd:attribute name="name" type="xsd:string" use="required"/>
  </xsd:complexType>

  <!-- **************************************************** -->

  <xsd:complexType name="named-native-query">
    <xsd:annotation>
      <xsd:documentation>

        @Target({TYPE}) @Retention(RUNTIME)
        public @interface NamedNativeQuery {
          String name();
          String query();
          QueryHint[] hints() default {};
          Class resultClass() default void.class;
          String resultSetMapping() default ""; //named SqlResultSetMapping
        }

      </xsd:documentation>
    </xsd:annotation>
    <xsd:sequence>
      <xsd:element name="query" type="xsd:string"/>
      <xsd:element name="hint" type="orm:query-hint" 
                   minOccurs="0" maxOccurs="unbounded"/>
    </xsd:sequence>
    <xsd:attribute name="name" type="xsd:string" use="required"/>
    <xsd:attribute name="result-class" type="xsd:string"/>
    <xsd:attribute name="result-set-mapping" type="xsd:string"/>
  </xsd:complexType>

  <!-- **************************************************** -->

  <xsd:complexType name="sql-result-set-mapping">
    <xsd:annotation>
      <xsd:documentation>

        @Target({TYPE}) @Retention(RUNTIME)
        public @interface SqlResultSetMapping {
          String name();
          EntityResult[] entities() default {};
          ColumnResult[] columns() default {};
        }

      </xsd:documentation>
    </xsd:annotation>
    <xsd:sequence>
      <xsd:element name="entity-result" type="orm:entity-result" 
                   minOccurs="0" maxOccurs="unbounded"/>
      <xsd:element name="column-result" type="orm:column-result" 
                   minOccurs="0" maxOccurs="unbounded"/>
    </xsd:sequence>
    <xsd:attribute name="name" type="xsd:string" use="required"/>
  </xsd:complexType>

  <!-- **************************************************** -->

  <xsd:complexType name="entity-result">
    <xsd:annotation>
      <xsd:documentation>

        @Target({}) @Retention(RUNTIME)
        public @interface EntityResult {
          Class entityClass();
          FieldResult[] fields() default {};
          String discriminatorColumn() default "";
        }

      </xsd:documentation>
    </xsd:annotation>
    <xsd:sequence>
      <xsd:element name="field-result" type="orm:field-result" 
                   minOccurs="0" maxOccurs="unbounded"/>
    </xsd:sequence>
    <xsd:attribute name="entity-class" type="xsd:string" use="required"/>
    <xsd:attribute name="discriminator-column" type="xsd:string"/>
  </xsd:complexType>

  <!-- **************************************************** -->

  <xsd:complexType name="field-result">
    <xsd:annotation>
      <xsd:documentation>

        @Target({}) @Retention(RUNTIME)
        public @interface FieldResult {
          String name();
          String column();
        }

      </xsd:documentation>
    </xsd:annotation>
    <xsd:attribute name="name" type="xsd:string" use="required"/>
    <xsd:attribute name="column" type="xsd:string" use="required"/>
  </xsd:complexType>

  <!-- **************************************************** -->

  <xsd:complexType name="column-result">
    <xsd:annotation>
      <xsd:documentation>

        @Target({}) @Retention(RUNTIME)
        public @interface ColumnResult {
          String name();
        }

      </xsd:documentation>
    </xsd:annotation>
    <xsd:attribute name="name" type="xsd:string" use="required"/>
  </xsd:complexType>

  <!-- **************************************************** -->

  <xsd:complexType name="table">
    <xsd:annotation>
      <xsd:documentation>

        @Target({TYPE}) @Retention(RUNTIME)
        public @interface Table {
          String name() default "";
          String catalog() default "";
          String schema() default "";
          UniqueConstraint[] uniqueConstraints() default {};
        }

      </xsd:documentation>
    </xsd:annotation>
    <xsd:sequence>
      <xsd:element name="unique-constraint" type="orm:unique-constraint" 
                   minOccurs="0" maxOccurs="unbounded"/>
    </xsd:sequence>
    <xsd:attribute name="name" type="xsd:string"/>
    <xsd:attribute name="catalog" type="xsd:string"/>
    <xsd:attribute name="schema" type="xsd:string"/>
  </xsd:complexType>

  <!-- **************************************************** -->

  <xsd:complexType name="secondary-table">
    <xsd:annotation>
      <xsd:documentation>

        @Target({TYPE}) @Retention(RUNTIME)
        public @interface SecondaryTable {
          String name();
          String catalog() default "";
          String schema() default "";
          PrimaryKeyJoinColumn[] pkJoinColumns() default {};
          UniqueConstraint[] uniqueConstraints() default {};
         }

       </xsd:documentation>
     </xsd:annotation>
     <xsd:sequence>
       <xsd:element name="primary-key-join-column" 
                    type="orm:primary-key-join-column" 
                    minOccurs="0" maxOccurs="unbounded"/>
       <xsd:element name="unique-constraint" type="orm:unique-constraint" 
                    minOccurs="0" maxOccurs="unbounded"/>
    </xsd:sequence>
    <xsd:attribute name="name" type="xsd:string" use="required"/>
    <xsd:attribute name="catalog" type="xsd:string"/>
    <xsd:attribute name="schema" type="xsd:string"/>
  </xsd:complexType>

  <!-- **************************************************** -->

  <xsd:complexType name="unique-constraint">
    <xsd:annotation>
      <xsd:documentation>

        @Target({}) @Retention(RUNTIME)
        public @interface UniqueConstraint {
          String[] columnNames();
        }

      </xsd:documentation>
    </xsd:annotation>
    <xsd:sequence>
      <xsd:element name="column-name" type="xsd:string" 
                   maxOccurs="unbounded"/>
    </xsd:sequence>
  </xsd:complexType>

  <!-- **************************************************** -->

  <xsd:complexType name="column">
    <xsd:annotation>
      <xsd:documentation>

        @Target({METHOD, FIELD}) @Retention(RUNTIME)
        public @interface Column {
          String name() default "";
          boolean unique() default false;
          boolean nullable() default true;
          boolean insertable() default true;
          boolean updatable() default true;
          String columnDefinition() default "";
          String table() default "";
          int length() default 255;
          int precision() default 0; // decimal precision
          int scale() default 0; // decimal scale
        }

      </xsd:documentation>
    </xsd:annotation>
    <xsd:attribute name="name" type="xsd:string"/>
    <xsd:attribute name="unique" type="xsd:boolean"/>
    <xsd:attribute name="nullable" type="xsd:boolean"/>
    <xsd:attribute name="insertable" type="xsd:boolean"/>
    <xsd:attribute name="updatable" type="xsd:boolean"/>
    <xsd:attribute name="column-definition" type="xsd:string"/>
    <xsd:attribute name="table" type="xsd:string"/>
    <xsd:attribute name="length" type="xsd:int"/>
    <xsd:attribute name="precision" type="xsd:int"/>
    <xsd:attribute name="scale" type="xsd:int"/>
  </xsd:complexType>

  <!-- **************************************************** -->

  <xsd:complexType name="join-column">
    <xsd:annotation>
      <xsd:documentation>

        @Target({METHOD, FIELD}) @Retention(RUNTIME)
        public @interface JoinColumn {
          String name() default "";
          String referencedColumnName() default "";
          boolean unique() default false;
          boolean nullable() default true;
          boolean insertable() default true;
          boolean updatable() default true;
          String columnDefinition() default "";
          String table() default "";
        }

      </xsd:documentation>
    </xsd:annotation>
    <xsd:attribute name="name" type="xsd:string"/>
    <xsd:attribute name="referenced-column-name" type="xsd:string"/>
    <xsd:attribute name="unique" type="xsd:boolean"/>
    <xsd:attribute name="nullable" type="xsd:boolean"/>
    <xsd:attribute name="insertable" type="xsd:boolean"/>
    <xsd:attribute name="updatable" type="xsd:boolean"/>
    <xsd:attribute name="column-definition" type="xsd:string"/>
    <xsd:attribute name="table" type="xsd:string"/>
  </xsd:complexType>

  <!-- **************************************************** -->

  <xsd:simpleType name="generation-type">
    <xsd:annotation>
      <xsd:documentation>

        public enum GenerationType { TABLE, SEQUENCE, IDENTITY, AUTO };

      </xsd:documentation>
    </xsd:annotation>
    <xsd:restriction base="xsd:token">
      <xsd:enumeration value="TABLE"/>
      <xsd:enumeration value="SEQUENCE"/>
      <xsd:enumeration value="IDENTITY"/>
      <xsd:enumeration value="AUTO"/>
    </xsd:restriction>
  </xsd:simpleType>

  <!-- **************************************************** -->

  <xsd:complexType name="attribute-override">
    <xsd:annotation>
      <xsd:documentation>

        @Target({TYPE, METHOD, FIELD}) @Retention(RUNTIME)
        public @interface AttributeOverride {
          String name();
          Column column();
        }

      </xsd:documentation>
    </xsd:annotation>
    <xsd:sequence>
      <xsd:element name="column" type="orm:column"/>
    </xsd:sequence>
    <xsd:attribute name="name" type="xsd:string" use="required"/>
  </xsd:complexType>

  <!-- **************************************************** -->

  <xsd:complexType name="association-override">
    <xsd:annotation>
      <xsd:documentation>

        @Target({TYPE, METHOD, FIELD}) @Retention(RUNTIME)
        public @interface AssociationOverride {
          String name();
          JoinColumn[] joinColumns();
        }

      </xsd:documentation>
    </xsd:annotation>
    <xsd:sequence>
      <xsd:element name="join-column" type="orm:join-column"
                   maxOccurs="unbounded"/>
    </xsd:sequence>
    <xsd:attribute name="name" type="xsd:string" use="required"/>
  </xsd:complexType>

  <!-- **************************************************** -->

  <xsd:complexType name="id-class">
    <xsd:annotation>
      <xsd:documentation>

        @Target({TYPE}) @Retention(RUNTIME)
        public @interface IdClass {
          Class value();
        }

      </xsd:documentation>
    </xsd:annotation>
    <xsd:attribute name="class" type="xsd:string" use="required"/>
  </xsd:complexType>

  <!-- **************************************************** -->

  <xsd:complexType name="id">
    <xsd:annotation>
      <xsd:documentation>

        @Target({METHOD, FIELD}) @Retention(RUNTIME)
        public @interface Id {}

      </xsd:documentation>
    </xsd:annotation>
    <xsd:sequence>
      <xsd:element name="column" type="orm:column" 
                   minOccurs="0"/>
      <xsd:element name="generated-value" type="orm:generated-value"
                   minOccurs="0"/>
      <xsd:element name="temporal" type="orm:temporal" 
                   minOccurs="0"/>
      <xsd:element name="table-generator" type="orm:table-generator" 
                   minOccurs="0"/>
      <xsd:element name="sequence-generator" type="orm:sequence-generator"
                   minOccurs="0"/>
    </xsd:sequence>
    <xsd:attribute name="name" type="xsd:string" use="required"/>
  </xsd:complexType>

  <!-- **************************************************** -->

  <xsd:complexType name="embedded-id">
    <xsd:annotation>
      <xsd:documentation>

        @Target({METHOD, FIELD}) @Retention(RUNTIME)
        public @interface EmbeddedId {}

      </xsd:documentation>
    </xsd:annotation>
    <xsd:sequence>
      <xsd:element name="attribute-override" type="orm:attribute-override" 
                   minOccurs="0" maxOccurs="unbounded"/>
    </xsd:sequence>
    <xsd:attribute name="name" type="xsd:string" use="required"/>
  </xsd:complexType>

  <!-- **************************************************** -->

  <xsd:complexType name="transient">
    <xsd:annotation>
      <xsd:documentation>

        @Target({METHOD, FIELD}) @Retention(RUNTIME)
        public @interface Transient {}

      </xsd:documentation>
    </xsd:annotation>
    <xsd:attribute name="name" type="xsd:string" use="required"/>
  </xsd:complexType>

  <!-- **************************************************** -->

  <xsd:complexType name="version">
    <xsd:annotation>
      <xsd:documentation>

        @Target({METHOD, FIELD}) @Retention(RUNTIME)
        public @interface Version {}

      </xsd:documentation>
    </xsd:annotation>
    <xsd:sequence>
      <xsd:element name="column" type="orm:column" minOccurs="0"/>
      <xsd:element name="temporal" type="orm:temporal" minOccurs="0"/>
    </xsd:sequence>
    <xsd:attribute name="name" type="xsd:string" use="required"/>
  </xsd:complexType>

  <!-- **************************************************** -->

  <xsd:complexType name="basic">
    <xsd:annotation>
      <xsd:documentation>

        @Target({METHOD, FIELD}) @Retention(RUNTIME)
        public @interface Basic {
          FetchType fetch() default EAGER;
          boolean optional() default true;
        }

      </xsd:documentation>
    </xsd:annotation>
    <xsd:sequence>
      <xsd:element name="column" type="orm:column" minOccurs="0"/>
      <xsd:choice>
        <xsd:element name="lob" type="orm:lob" minOccurs="0"/>
        <xsd:element name="temporal" type="orm:temporal" minOccurs="0"/>
        <xsd:element name="enumerated" type="orm:enumerated" minOccurs="0"/>
      </xsd:choice>
    </xsd:sequence>
    <xsd:attribute name="name" type="xsd:string" use="required"/>
    <xsd:attribute name="fetch" type="orm:fetch-type"/>
    <xsd:attribute name="optional" type="xsd:boolean"/>
  </xsd:complexType>

  <!-- **************************************************** -->

  <xsd:simpleType name="fetch-type">
    <xsd:annotation>
      <xsd:documentation>

        public enum FetchType { LAZY, EAGER };

      </xsd:documentation>
    </xsd:annotation>
    <xsd:restriction base="xsd:token">
      <xsd:enumeration value="LAZY"/>
      <xsd:enumeration value="EAGER"/>
    </xsd:restriction>
  </xsd:simpleType>

  <!-- **************************************************** -->

  <xsd:complexType name="lob">
    <xsd:annotation>
      <xsd:documentation>

        @Target({METHOD, FIELD}) @Retention(RUNTIME)
        public @interface Lob {}

      </xsd:documentation>
    </xsd:annotation>
  </xsd:complexType>

  <!-- **************************************************** -->

  <xsd:simpleType name="temporal">
    <xsd:annotation>
      <xsd:documentation>

        @Target({METHOD, FIELD}) @Retention(RUNTIME)
        public @interface Temporal {
          TemporalType value();
        }

      </xsd:documentation>
    </xsd:annotation>
    <xsd:restriction base="orm:temporal-type"/>
  </xsd:simpleType>

  <!-- **************************************************** -->

  <xsd:simpleType name="temporal-type">
    <xsd:annotation>
      <xsd:documentation>

        public enum TemporalType {
          DATE, // java.sql.Date
          TIME, // java.sql.Time
          TIMESTAMP // java.sql.Timestamp
        }

      </xsd:documentation>
    </xsd:annotation>
      <xsd:restriction base="xsd:token">
        <xsd:enumeration value="DATE"/>
        <xsd:enumeration value="TIME"/>
        <xsd:enumeration value="TIMESTAMP"/>
     </xsd:restriction>
  </xsd:simpleType>

  <!-- **************************************************** -->

  <xsd:simpleType name="enumerated">
    <xsd:annotation>
      <xsd:documentation>

        @Target({METHOD, FIELD}) @Retention(RUNTIME)
        public @interface Enumerated {
          EnumType value() default ORDINAL;
        }

      </xsd:documentation>
    </xsd:annotation>
    <xsd:restriction base="orm:enum-type"/>
  </xsd:simpleType>

  <!-- **************************************************** -->

  <xsd:simpleType name="enum-type">
    <xsd:annotation>
      <xsd:documentation>

        public enum EnumType {
          ORDINAL,
          STRING
        }

      </xsd:documentation>
    </xsd:annotation>
    <xsd:restriction base="xsd:token">
      <xsd:enumeration value="ORDINAL"/>
      <xsd:enumeration value="STRING"/>
    </xsd:restriction>
  </xsd:simpleType>

  <!-- **************************************************** -->

  <xsd:complexType name="many-to-one">
    <xsd:annotation>
      <xsd:documentation>

        @Target({METHOD, FIELD}) @Retention(RUNTIME)
        public @interface ManyToOne {
          Class targetEntity() default void.class;
          CascadeType[] cascade() default {};
          FetchType fetch() default EAGER;
          boolean optional() default true;
        }

      </xsd:documentation>
    </xsd:annotation>
    <xsd:sequence>
      <xsd:choice>       
        <xsd:element name="join-column" type="orm:join-column" 
                     minOccurs="0" maxOccurs="unbounded"/>
        <xsd:element name="join-table" type="orm:join-table" 
                     minOccurs="0"/>
      </xsd:choice>       
      <xsd:element name="cascade" type="orm:cascade-type" 
                   minOccurs="0"/>
    </xsd:sequence>
    <xsd:attribute name="name" type="xsd:string" use="required"/>
    <xsd:attribute name="target-entity" type="xsd:string"/>
    <xsd:attribute name="fetch" type="orm:fetch-type"/>
    <xsd:attribute name="optional" type="xsd:boolean"/>
  </xsd:complexType>

  <!-- **************************************************** -->

  <xsd:complexType name="cascade-type">
    <xsd:annotation>
      <xsd:documentation>

        public enum CascadeType { ALL, PERSIST, MERGE, REMOVE, REFRESH};

      </xsd:documentation>
    </xsd:annotation>
    <xsd:sequence>
      <xsd:element name="cascade-all" type="orm:emptyType"
                   minOccurs="0"/>
      <xsd:element name="cascade-persist" type="orm:emptyType"
                   minOccurs="0"/>
      <xsd:element name="cascade-merge" type="orm:emptyType"
                   minOccurs="0"/>
      <xsd:element name="cascade-remove" type="orm:emptyType"
                   minOccurs="0"/>
      <xsd:element name="cascade-refresh" type="orm:emptyType"
                   minOccurs="0"/>
    </xsd:sequence>
  </xsd:complexType>

  <!-- **************************************************** -->

  <xsd:complexType name="one-to-one">
    <xsd:annotation>
      <xsd:documentation>

        @Target({METHOD, FIELD}) @Retention(RUNTIME)
        public @interface OneToOne {
          Class targetEntity() default void.class;
          CascadeType[] cascade() default {};
          FetchType fetch() default EAGER;
          boolean optional() default true;
          String mappedBy() default "";
        }

      </xsd:documentation>
    </xsd:annotation>
    <xsd:sequence>
      <xsd:choice>
        <xsd:element name="primary-key-join-column" 
                     type="orm:primary-key-join-column" 
                     minOccurs="0" maxOccurs="unbounded"/>
        <xsd:element name="join-column" type="orm:join-column" 
                     minOccurs="0" maxOccurs="unbounded"/>
        <xsd:element name="join-table" type="orm:join-table" 
                     minOccurs="0"/>
      </xsd:choice>
      <xsd:element name="cascade" type="orm:cascade-type" 
                   minOccurs="0"/>
    </xsd:sequence>
    <xsd:attribute name="name" type="xsd:string" use="required"/>
    <xsd:attribute name="target-entity" type="xsd:string"/>
    <xsd:attribute name="fetch" type="orm:fetch-type"/>
    <xsd:attribute name="optional" type="xsd:boolean"/>
    <xsd:attribute name="mapped-by" type="xsd:string"/>
  </xsd:complexType>

  <!-- **************************************************** -->

  <xsd:complexType name="one-to-many">
    <xsd:annotation>
      <xsd:documentation>

        @Target({METHOD, FIELD}) @Retention(RUNTIME)
        public @interface OneToMany {
          Class targetEntity() default void.class;
          CascadeType[] cascade() default {};
          FetchType fetch() default LAZY;
          String mappedBy() default "";
        }

      </xsd:documentation>
    </xsd:annotation>
    <xsd:sequence>
      <xsd:element name="order-by" type="orm:order-by" 
                   minOccurs="0"/>
      <xsd:element name="map-key" type="orm:map-key" 
                   minOccurs="0"/>
      <xsd:choice>       
        <xsd:element name="join-table" type="orm:join-table" 
                     minOccurs="0"/>
        <xsd:element name="join-column" type="orm:join-column" 
                     minOccurs="0" maxOccurs="unbounded"/>
      </xsd:choice>       
      <xsd:element name="cascade" type="orm:cascade-type" 
                   minOccurs="0"/>
    </xsd:sequence>
    <xsd:attribute name="name" type="xsd:string" use="required"/>
    <xsd:attribute name="target-entity" type="xsd:string"/>
    <xsd:attribute name="fetch" type="orm:fetch-type"/>
    <xsd:attribute name="mapped-by" type="xsd:string"/>
  </xsd:complexType>

  <!-- **************************************************** -->

  <xsd:complexType name="join-table">
    <xsd:annotation>
      <xsd:documentation>

        @Target({METHOD, FIELD}) @Retention(RUNTIME)
        public @interface JoinTable {
          String name() default "";
          String catalog() default "";
          String schema() default "";
          JoinColumn[] joinColumns() default {};
          JoinColumn[] inverseJoinColumns() default {};
          UniqueConstraint[] uniqueConstraints() default {};
        }

      </xsd:documentation>
    </xsd:annotation>
    <xsd:sequence>
      <xsd:element name="join-column" type="orm:join-column" 
                   minOccurs="0" maxOccurs="unbounded"/>
      <xsd:element name="inverse-join-column" type="orm:join-column" 
                   minOccurs="0" maxOccurs="unbounded"/>
      <xsd:element name="unique-constraint" type="orm:unique-constraint" 
                   minOccurs="0" maxOccurs="unbounded"/>
    </xsd:sequence>
    <xsd:attribute name="name" type="xsd:string"/>
    <xsd:attribute name="catalog" type="xsd:string"/>
    <xsd:attribute name="schema" type="xsd:string"/>
  </xsd:complexType>

  <!-- **************************************************** -->

  <xsd:complexType name="many-to-many">
    <xsd:annotation>
      <xsd:documentation>

        @Target({METHOD, FIELD}) @Retention(RUNTIME)
        public @interface ManyToMany {
          Class targetEntity() default void.class;
          CascadeType[] cascade() default {};
          FetchType fetch() default LAZY;
          String mappedBy() default "";
        }

      </xsd:documentation>
    </xsd:annotation>
    <xsd:sequence>
      <xsd:element name="order-by" type="orm:order-by" 
                   minOccurs="0"/>
      <xsd:element name="map-key" type="orm:map-key" 
                   minOccurs="0"/>
      <xsd:element name="join-table" type="orm:join-table" 
                   minOccurs="0"/>
      <xsd:element name="cascade" type="orm:cascade-type" 
                   minOccurs="0"/>
    </xsd:sequence>
    <xsd:attribute name="name" type="xsd:string" use="required"/>
    <xsd:attribute name="target-entity" type="xsd:string"/>
    <xsd:attribute name="fetch" type="orm:fetch-type"/>
    <xsd:attribute name="mapped-by" type="xsd:string"/>
  </xsd:complexType>

  <!-- **************************************************** -->

  <xsd:complexType name="generated-value">
    <xsd:annotation>
      <xsd:documentation>

        @Target({METHOD, FIELD}) @Retention(RUNTIME)
        public @interface GeneratedValue {
          GenerationType strategy() default AUTO;
          String generator() default "";
        }

      </xsd:documentation>
    </xsd:annotation>
    <xsd:attribute name="strategy" type="orm:generation-type"/>
    <xsd:attribute name="generator" type="xsd:string"/>
  </xsd:complexType>

  <!-- **************************************************** -->

  <xsd:complexType name="map-key">
    <xsd:annotation>
      <xsd:documentation>

        @Target({METHOD, FIELD}) @Retention(RUNTIME)
        public @interface MapKey {
          String name() default "";
        }

      </xsd:documentation>
    </xsd:annotation>
    <xsd:attribute name="name" type="xsd:string"/>
  </xsd:complexType>

  <!-- **************************************************** -->

  <xsd:simpleType name="order-by">
    <xsd:annotation>
      <xsd:documentation>

        @Target({METHOD, FIELD}) @Retention(RUNTIME)
        public @interface OrderBy {
          String value() default "";
        }

      </xsd:documentation>
    </xsd:annotation>
    <xsd:restriction base="xsd:string"/>
  </xsd:simpleType>

  <!-- **************************************************** -->

  <xsd:complexType name="inheritance">
    <xsd:annotation>
      <xsd:documentation>

        @Target({TYPE}) @Retention(RUNTIME)
        public @interface Inheritance {
          InheritanceType strategy() default SINGLE_TABLE;
        }

      </xsd:documentation>
    </xsd:annotation>
    <xsd:attribute name="strategy" type="orm:inheritance-type"/>
  </xsd:complexType>

  <!-- **************************************************** -->

  <xsd:simpleType name="inheritance-type">
    <xsd:annotation>
      <xsd:documentation>

        public enum InheritanceType
          { SINGLE_TABLE, JOINED, TABLE_PER_CLASS};

      </xsd:documentation>
    </xsd:annotation>
    <xsd:restriction base="xsd:token">
      <xsd:enumeration value="SINGLE_TABLE"/>
      <xsd:enumeration value="JOINED"/>
      <xsd:enumeration value="TABLE_PER_CLASS"/>
    </xsd:restriction>
  </xsd:simpleType>

  <!-- **************************************************** -->

  <xsd:simpleType name="discriminator-value">
    <xsd:annotation>
      <xsd:documentation>

        @Target({TYPE}) @Retention(RUNTIME)
        public @interface DiscriminatorValue {
          String value();
        }

      </xsd:documentation>
    </xsd:annotation>
    <xsd:restriction base="xsd:string"/>
  </xsd:simpleType>

  <!-- **************************************************** -->

  <xsd:simpleType name="discriminator-type">
    <xsd:annotation>
      <xsd:documentation>

        public enum DiscriminatorType { STRING, CHAR, INTEGER };

      </xsd:documentation>
    </xsd:annotation>
    <xsd:restriction base="xsd:token">
      <xsd:enumeration value="STRING"/>
      <xsd:enumeration value="CHAR"/>
      <xsd:enumeration value="INTEGER"/>
    </xsd:restriction>
  </xsd:simpleType>

  <!-- **************************************************** -->

  <xsd:complexType name="primary-key-join-column">
    <xsd:annotation>
      <xsd:documentation>

        @Target({TYPE, METHOD, FIELD}) @Retention(RUNTIME)
        public @interface PrimaryKeyJoinColumn {
          String name() default "";
          String referencedColumnName() default "";
          String columnDefinition() default "";
        }

      </xsd:documentation>
    </xsd:annotation>
    <xsd:attribute name="name" type="xsd:string"/>
    <xsd:attribute name="referenced-column-name" type="xsd:string"/>
    <xsd:attribute name="column-definition" type="xsd:string"/>
  </xsd:complexType>

  <!-- **************************************************** -->

  <xsd:complexType name="discriminator-column">
    <xsd:annotation>
      <xsd:documentation>

        @Target({TYPE}) @Retention(RUNTIME)
        public @interface DiscriminatorColumn {
          String name() default "DTYPE";
          DiscriminatorType discriminatorType() default STRING;
          String columnDefinition() default "";
          int length() default 31;
        }

      </xsd:documentation>
    </xsd:annotation>
    <xsd:attribute name="name" type="xsd:string"/>
    <xsd:attribute name="discriminator-type" type="orm:discriminator-type"/>
    <xsd:attribute name="column-definition" type="xsd:string"/>
    <xsd:attribute name="length" type="xsd:int"/>
  </xsd:complexType>

  <!-- **************************************************** -->

  <xsd:complexType name="embeddable">
    <xsd:annotation>
      <xsd:documentation>

        Defines the settings and mappings for embeddable objects. Is 
        allowed to be sparsely populated and used in conjunction with 
        the annotations. Alternatively, the metadata-complete attribute 
        can be used to indicate that no annotations are to be processed 
        in the class. If this is the case then the defaulting rules will 
        be recursively applied.

        @Target({TYPE}) @Retention(RUNTIME)
        public @interface Embeddable {}

      </xsd:documentation>
    </xsd:annotation>
    <xsd:sequence>
      <xsd:element name="description" type="xsd:string" minOccurs="0"/>
      <xsd:element name="attributes" type="orm:embeddable-attributes" 
                   minOccurs="0"/>
    </xsd:sequence>
    <xsd:attribute name="class" type="xsd:string" use="required"/>
    <xsd:attribute name="access" type="orm:access-type"/>
    <xsd:attribute name="metadata-complete" type="xsd:boolean"/>
  </xsd:complexType>

  <!-- **************************************************** -->

  <xsd:complexType name="embeddable-attributes">
    <xsd:sequence>
      <xsd:element name="basic" type="orm:basic" 
                   minOccurs="0" maxOccurs="unbounded"/>
      <xsd:element name="transient" type="orm:transient" 
                   minOccurs="0" maxOccurs="unbounded"/>
    </xsd:sequence>
  </xsd:complexType>

  <!-- **************************************************** -->

  <xsd:complexType name="embedded">
    <xsd:annotation>
      <xsd:documentation>

        @Target({METHOD, FIELD}) @Retention(RUNTIME)
        public @interface Embedded {}

      </xsd:documentation>
    </xsd:annotation>
    <xsd:sequence>
      <xsd:element name="attribute-override" type="orm:attribute-override" 
                   minOccurs="0" maxOccurs="unbounded"/>
    </xsd:sequence>
    <xsd:attribute name="name" type="xsd:string" use="required"/>
  </xsd:complexType>

  <!-- **************************************************** -->

  <xsd:complexType name="mapped-superclass">
    <xsd:annotation>
      <xsd:documentation>

        Defines the settings and mappings for a mapped superclass. Is 
        allowed to be sparsely populated and used in conjunction with 
        the annotations. Alternatively, the metadata-complete attribute 
        can be used to indicate that no annotations are to be processed 
        If this is the case then the defaulting rules will be recursively 
        applied.

        @Target(TYPE) @Retention(RUNTIME)
        public @interface MappedSuperclass{}

      </xsd:documentation>
    </xsd:annotation>
    <xsd:sequence>
      <xsd:element name="description" type="xsd:string" minOccurs="0"/>
      <xsd:element name="id-class" type="orm:id-class" minOccurs="0"/>
      <xsd:element name="exclude-default-listeners" type="orm:emptyType" 
                   minOccurs="0"/>
      <xsd:element name="exclude-superclass-listeners" type="orm:emptyType" 
                   minOccurs="0"/>
      <xsd:element name="entity-listeners" type="orm:entity-listeners" 
                   minOccurs="0"/>
      <xsd:element name="pre-persist" type="orm:pre-persist" minOccurs="0"/>
      <xsd:element name="post-persist" type="orm:post-persist" 
                   minOccurs="0"/>
      <xsd:element name="pre-remove" type="orm:pre-remove" minOccurs="0"/>
      <xsd:element name="post-remove" type="orm:post-remove" minOccurs="0"/>
      <xsd:element name="pre-update" type="orm:pre-update" minOccurs="0"/>
      <xsd:element name="post-update" type="orm:post-update" minOccurs="0"/>
      <xsd:element name="post-load" type="orm:post-load" minOccurs="0"/>
      <xsd:element name="attributes" type="orm:attributes" minOccurs="0"/>
    </xsd:sequence>
    <xsd:attribute name="class" type="xsd:string" use="required"/>
    <xsd:attribute name="access" type="orm:access-type"/>
    <xsd:attribute name="metadata-complete" type="xsd:boolean"/>
  </xsd:complexType>

  <!-- **************************************************** -->

  <xsd:complexType name="sequence-generator">
    <xsd:annotation>
      <xsd:documentation>

        @Target({TYPE, METHOD, FIELD}) @Retention(RUNTIME)
        public @interface SequenceGenerator {
          String name();
          String sequenceName() default "";
          int initialValue() default 1;
          int allocationSize() default 50;
        }

      </xsd:documentation>
    </xsd:annotation>
    <xsd:attribute name="name" type="xsd:string" use="required"/>
    <xsd:attribute name="sequence-name" type="xsd:string"/>
    <xsd:attribute name="initial-value" type="xsd:int"/>
    <xsd:attribute name="allocation-size" type="xsd:int"/>
  </xsd:complexType>

  <!-- **************************************************** -->

  <xsd:complexType name="table-generator">
    <xsd:annotation>
      <xsd:documentation>

        @Target({TYPE, METHOD, FIELD}) @Retention(RUNTIME)
        public @interface TableGenerator {
          String name();
          String table() default "";
          String catalog() default "";
          String schema() default "";
          String pkColumnName() default "";
          String valueColumnName() default "";
          String pkColumnValue() default "";
          int initialValue() default 0;
          int allocationSize() default 50;
          UniqueConstraint[] uniqueConstraints() default {};
        }

      </xsd:documentation>
    </xsd:annotation>
    <xsd:sequence>
      <xsd:element name="unique-constraint" type="orm:unique-constraint" 
                   minOccurs="0" maxOccurs="unbounded"/>
    </xsd:sequence>
    <xsd:attribute name="name" type="xsd:string" use="required"/>
    <xsd:attribute name="table" type="xsd:string"/>
    <xsd:attribute name="catalog" type="xsd:string"/>
    <xsd:attribute name="schema" type="xsd:string"/>
    <xsd:attribute name="pk-column-name" type="xsd:string"/>
    <xsd:attribute name="value-column-name" type="xsd:string"/>
    <xsd:attribute name="pk-column-value" type="xsd:string"/>
    <xsd:attribute name="initial-value" type="xsd:int"/>
    <xsd:attribute name="allocation-size" type="xsd:int"/>
  </xsd:complexType>

</xsd:schema>

4.  Conclusion

That exhausts persistence metadata annotations. We present the class definitions for our sample model below:

Example 5.2.  Complete Metadata

package org.mag;

@Entity
@IdClass(Magazine.MagazineId.class)
public class Magazine {

    @Id private String isbn;
    @Id private String title;
    @Version private int version;

    private double price;   // defaults to @Basic
    private int copiesSold; // defaults to @Basic

    @OneToOne(fetch=FetchType.LAZY, 
        cascade={CascadeType.PERSIST,CascadeType.REMOVE})
    private Article coverArticle;

    @OneToMany(cascade={CascadeType.PERSIST,CascadeType.REMOVE})
    @OrderBy
    private Collection<Article> articles;

    @ManyToOne(fetch=FetchType.LAZY, cascade=CascadeType.PERSIST)
    private Company publisher; 

    @Transient private byte[] data;

    ...

    public static class MagazineId {
        ...
    }
}

@Entity
public class Article {

    @Id private long id;
    @Version private int version;
    
    private String title;   // defaults to @Basic
    private byte[] content; // defaults to @Basic

    @ManyToMany(cascade=CascadeType.PERSIST)
    @OrderBy("lastName, firstName")
    private Collection<Author> authors;

    ...
}


package org.mag.pub;

@Entity
public class Company {

    @Id private long id;
    @Version private int version;

    private String name;     // defaults to @Basic
    private double revenue;  // defaults to @Basic
    private Address address; // defaults to @Embedded
    
    @OneToMany(mappedBy="publisher", cascade=CascadeType.PERSIST)
    private Collection<Magazine> mags;

    @OneToMany(cascade={CascadeType.PERSIST,CascadeType.REMOVE})
    private Collection<Subscription> subscriptions;

    ...
}

@Entity
public class Author {

    @Id private long id;
    @Version private int version;

    private String firstName; // defaults to @Basic
    private double lastName;  // defaults to @Basic
    private Address address;  // defaults to @Embedded
    
    @ManyToMany(mappedBy="authors", cascade=CascadeType.PERSIST)
    private Collection<Article> arts;

    ...
}

@Embeddable
public class Address {

    private String street; // defaults to @Basic
    private String city;   // defaults to @Basic
    private String state;  // defaults to @Basic
    private String zip;    // defaults to @Basic

    ...
}


package org.mag.subscribe;

@MappedSuperclass
public abstract class Document {

    @Id private long id;
    @Version private int version;

    ...
}

@Entity
public class Contract
    extends Document {

    private String terms; // defaults to @Basic

    ...
}

@Entity
public class Subscription {

    @Id private long id;
    @Version private int version;

    private Date startDate; // defaults to @Basic
    private double payment; // defaults to @Basic

    @OneToMany(cascade={CascadeType.PERSIST,CascadeType.REMOVE})
    @MapKey(name="num")
    private Map<Long,LineItem> lineItems;

    ...

    @Entity
    public static class LineItem
        extends Contract {

        private String comments; // defaults to @Basic
        private double price;    // defaults to @Basic
        private long num;        // defaults to @Basic

        @ManyToOne
        private Magazine magazine;

        ...
    }
}

@Entity(name="Lifetime")
public class LifetimeSubscription
    extends Subscription {

    @Basic(fetch=FetchType.LAZY)
    private boolean getEliteClub() { ... }
    public void setEliteClub(boolean elite) { ... }

    ...
}

@Entity(name="Trial")
public class TrialSubscription
    extends Subscription {

    public Date getEndDate() { ... }
    public void setEndDate(Date end) { ... }

    ...
}

The same metadata declarations in XML:

<entity-mappings>
    <!-- declares a default access type for all entities -->
    <access-type>FIELD</access-type>
    <mapped-superclass class="org.mag.subscribe.Document">
        <attributes>
            <id name="id">
                <generated-value strategy="IDENTITY"/>
            </id>
            <version name="version"/>
        </attributes>
    </mapped-superclass>
    <entity class="org.mag.Magazine">
        <id-class="org.mag.Magazine$MagazineId"/>
        <attributes>
            <id name="isbn"/>
            <id name="title"/>
            <basic name="name"/>
            <basic name="price"/>
            <basic name="copiesSold"/>
            <version name="version"/>
            <many-to-one name="publisher" fetch="LAZY">
                <cascade>
                    <cascade-persist/>
                </cascade>
            </many-to-one>
            <one-to-many name="articles">
                <order-by/>
                <cascade>
                    <cascade-persist/>
                    <cascade-remove/>
                </cascade>
            </one-to-many>
            <one-to-one name="coverArticle" fetch="LAZY">
                <cascade>
                    <cascade-persist/>
                    <cascade-remove/>
                </cascade>
            </one-to-one>
            <transient name="data"/>
        </attributes>
    </entity>
    <entity class="org.mag.Article">
        <attributes>
            <id name="id"/>
            <basic name="title"/>
            <basic name="content"/>
            <version name="version"/>
            <many-to-many name="articles">
                <order-by>lastName, firstName</order-by>
            </many-to-many>
        </attributes>
    </entity>
    <entity class="org.mag.pub.Company">
        <attributes>
            <id name="id"/>
            <basic name="name"/>
            <basic name="revenue"/>
            <version name="version"/>
            <one-to-many name="mags" mapped-by="publisher">
                <cascade>
                    <cascade-persist/>
                </cascade>
            </one-to-many>
            <one-to-many name="subscriptions">
                <cascade>
                    <cascade-persist/>
                    <cascade-remove/>
                </cascade>
            </one-to-many>
        </attributes>
    </entity>
    <entity class="org.mag.pub.Author">
        <attributes>
            <id name="id"/>
            <basic name="firstName"/>
            <basic name="lastName"/>
            <version name="version"/>
            <many-to-many name="arts" mapped-by="authors">
                <cascade>
                    <cascade-persist/>
                </cascade>
            </many-to-many>
        </attributes>
    </entity>
    <entity class="org.mag.subcribe.Contract">
        <attributes>
            <basic name="terms"/>
        </attributes>
    </entity>
    <entity class="org.mag.subcribe.Subscription">
        <attributes>
            <id name="id"/>
            <basic name="payment"/>
            <basic name="startDate"/>
            <version name="version"/>
            <one-to-many name="items">
                <map-key name="num">
                <cascade>
                    <cascade-persist/>
                    <cascade-remove/>
                </cascade>
            </one-to-many>
        </attributes>
    </entity>
    <entity class="org.mag.subscribe.Subscription.LineItem">
        <attributes>
            <basic name="comments"/>
            <basic name="price"/>
            <basic name="num"/>
            <many-to-one name="magazine"/>
        </attributes>
    </entity>
    <entity class="org.mag.subscribe.LifetimeSubscription" name="Lifetime"
        access="PROPERTY">
        <attributes>
            <basic name="eliteClub" fetch="LAZY"/>
        </attributes>
    </entity>
    <entity class="org.mag.subscribe.TrialSubscription" name="Trial">
        <attributes>
            <basic name="endDate"/>
        </attributes>
    </entity>
    <embeddable class="org.mag.pub.Address">
        <attributes>
            <basic name="street"/>
            <basic name="city"/>
            <basic name="state"/>
            <basic name="zip"/>
        </attributes>
    </embeddable>
</entity-mappings>

Chapter 12, Mapping Metadata will show you how to map your persistent classes to the datastore using additional annotations and XML markup. First, however, we turn to the JPA runtime APIs.

Chapter 6.  Persistence

Note

OpenJPA also includes the OpenJPAPersistence helper class to provide additional utility methods.

Within a container, you will typically use injection to access an EntityManagerFactory. Applications operating of a container, however, can use the Persistence class to obtain EntityManagerFactory objects in a vendor-neutral fashion.

public static EntityManagerFactory createEntityManagerFactory(String name);
public static EntityManagerFactory createEntityManagerFactory(String name, Map props);

Each createEntityManagerFactory method searches the system for an EntityManagerFactory definition with the given name. Use null for an unnamed factory. The optional map contains vendor-specific property settings used to further configure the factory.

persistence.xml files define EntityManagerFactories. The createEntityManagerFactory methods search for persistence.xml files within the META-INF directory of any CLASSPATH element. For example, if your CLASSPATH contains the conf directory, you could place an EntityManagerFactory definition in conf/META-INF/persistence.xml.

1.  persistence.xml

The persistence.xml file format obeys the following Document Type Descriptor (DTD):

<!ELEMENT persistence (persistence-unit*)>
<!ELEMENT persistence-unit (description?,provider?,jta-data-source?,
  non-jta-data-source?,(class|jar-file|mapping-file)*,
  exclude-unlisted-classes?,properties?)>
<!ATTLIST persistence-unit name CDATA #REQUIRED>
<!ATTLIST persistence-unit transaction-type (JTA|RESOURCE_LOCAL) "JTA">
<!ELEMENT description (#PCDATA)>
<!ELEMENT provider (#PCDATA)>
<!ELEMENT jta-data-source (#PCDATA)>
<!ELEMENT non-jta-data-source (#PCDATA)>
<!ELEMENT mapping-file (#PCDATA)>
<!ELEMENT jar-file (#PCDATA)>
<!ELEMENT class (#PCDATA)>
<!ELEMENT exclude-unlisted-classes EMPTY>
<!ELEMENT properties (property*)>
<!ELEMENT property EMPTY>
<!ATTLIST property name CDATA #REQUIRED>
<!ATTLIST property value CDATA #REQUIRED>

The root element of a persistence.xml file is persistence, which then contains one or more persistence-unit definitions. Each persistence unit describes the configuration for the entity managers created by the persistence unit's entity manager factory. The persistence unit can specify these elements and attribtues.

  • name: This is the name you pass to the Persistence.createEntityManagerFactory methods described above. The name attribute is required.

  • transaction-type: Whether to use managed (JTA) or local (RESOURCE_LOCAL) transaction management.

  • provider: If you are using a third-party JPA vendor, this element names its implementation of the PersistenceProvider bootstrapping interface.

    Note

    Set the provider to org.apache.openjpa.persistence.PersistenceProviderImpl to use OpenJPA.

  • jta-data-source: The JNDI name of a JDBC DataSource that is automatically enlisted in JTA transactions. This may be an XA DataSource.

  • non-jta-data-source: The JNDI name of a JDBC DataSource that is not enlisted in JTA transactions.

  • mapping-file*: The resource names of XML mapping files for entities and embeddable classes. You can also specify mapping information in an orm.xml file in your META-INF directory. If present, the orm.xml mapping file will be read automatically.

  • jar-file*: The names of jar files containing entities and embeddable classes. The implementation will scan the jar for annotated classes.

  • class*: The class names of entities and embeddable classes.

  • properties: This element contains nested property elements used to specify vendor-specific settings. Each property has a name attribute and a value attribute.

    Note

    The Reference Guide's Chapter 2, Configuration describes OpenJPA's configuration properties.

Here is a typical persistence.xml file for a non-EE environment:

Example 6.1.  persistence.xml

<?xml version="1.0"?>
<persistence>
  <persistence-unit name="openjpa">
    <provider>org.apache.openjpa.persistence.PersistenceProviderImpl</provider>
    <class>tutorial.Animal</class>
    <class>tutorial.Dog</class>
    <class>tutorial.Rabbit</class>
    <class>tutorial.Snake</class>
    <properties>
      <property name="openjpa.ConnectionURL" value="jdbc:hsqldb:tutorial_database"/>
      <property name="openjpa.ConnectionDriverName" value="org.hsqldb.jdbcDriver"/>
      <property name="openjpa.ConnectionUserName" value="sa"/>
      <property name="openjpa.ConnectionPassword" value=""/>
      <property name="openjpa.Log" value="DefaultLevel=WARN, Tool=INFO"/>
    </properties>
  </persistence-unit>
</persistence>

2.  Non-EE Use

The example below demonstrates the Persistence class in action. You will typically execute code like this on application startup, then cache the resulting factory for future use. This bootstrapping code is only necessary in non-EE environments; in an EE environment EntityManagerFactories are typically injected.

Example 6.2.  Obtaining an EntityManagerFactory

// if your persistence.xml file does not contain all settings already, you
// can add vendor settings to a map
Properties props = new Properties();
...

// create the factory defined by the "openjpa" entity-manager entry
EntityManagerFactory emf = Persistence.createEntityManagerFactory("openjpa", props);

Chapter 7.  EntityManagerFactory

The EntityManagerFactory creates EntityManager instances for application use.

Note

OpenJPA extends the standard EntityManagerFactory interface with the OpenJPAEntityManagerFactory to provide additional functionality.

1.  Obtaining an EntityManagerFactory

Within a container, you will typically use injection to access an EntityManagerFactory. There are, however, alternative mechanisms for EntityManagerFactory construction.

Some vendors may supply public constructors for their EntityManagerFactory implementations, but we recommend using the Java Connector Architecture (JCA) in a managed environment, or the Persistence class' createEntityManagerFactory methods in an unmanaged environment, as described in Chapter 6, Persistence . These strategies allow vendors to pool factories, cutting down on resource utilization.

JPA allows you to create and configure an EntityManagerFactory, then store it in a Java Naming and Directory Interface (JNDI) tree for later retrieval and use.

2.  Obtaining EntityManagers

public EntityManager createEntityManager();
public EntityManager createEntityManager(Map map);

The two createEntityManager methods above create a new EntityManager each time they are invoked. The optional Map is used to to supply vendor-specific settings. If you have configured your implementation for JTA transactions and a JTA transaction is active, the returned EntityManager will be synchronized with that transaction.

Note

OpenJPA recognizes the following string keys in the map supplied to createEntityManager:

The last option uses reflection to configure any property of OpenJPA's EntityManager implementation with the value supplied in your map. The first options correspond exactly to the same-named OpenJPA configuration keys described in Chapter 2, Configuration of the Reference Guide.

3.  Persistence Context

A persistence context is a set of entities such that for any persistent identity there is a unique entity instance. Within a persistence context, entities are managed. The EntityManager controls their lifecycle, and they can access datastore resources.

When a persistence context ends, previously-managed entities become detached. A detached entity is no longer under the control of the EntityManager, and no longer has access to datastore resources. We discuss detachment is detail in Section 2, “ Entity Lifecycle Management ”. For now, it is sufficient to know that detachment as has two obvious consequences:

  1. The detached entity cannot load any additional persistent state.

  2. The EntityManager will not return the detached entity from find, nor will queries include the detached entity in their results. Instead, find method invocations and query executions that would normally incorporate the detached entity will create a new managed entity with the same identity.

Note

OpenJPA offers several features related to detaching entities. See Section 1, “ Detach and Attach ” in the Reference Guide. Section 1.3, “ Defining the Detached Object Graph ” in particular describes how to use the DetachState setting to boost the performance of merging detached entities.

Injected EntityManagers have use a transaction , while EntityManagers obtained through the EntityManagerFactory have an extended persistence context. We describe these persistence context types below.

3.1.  Transaction Persistence Context

Under the transaction persistence context model, an EntityManager begins a new persistence context with each transaction, and ends the context when the transaction commits or rolls back. Within the transaction, entities you retrieve through the EntityManager or via Queries are managed entities. They can access datastore resources to lazy-load additional persistent state as needed, and only one entity may exist for any persistent identity.

When the transaction completes, all entities lose their association with the EntityManager and become detached. Traversing a persistent field that wasn't already loaded now has undefined results. And using the EntityManager or a Query to retrieve additional objects may now create new instances with the same persistent identities as detached instances.

If you use an EntityManager with a transaction persistence context model outside of an active transaction, each method invocation creates a new persistence context, performs the method action, and ends the persistence context. For example, consider using the EntityManager.find method outside of a transaction. The EntityManager will create a temporary persistence context, perform the find operation, end the persistence context, and return the detached result object to you. A second call with the same id will return a second detached object.

When the next transaction begins, the EntityManager will begin a new persistence context, and will again start returning managed entities. As you'll see in Chapter 8, EntityManager , you can also merge the previously-detached entites back into the new persistence context.

Example 7.1.  Behavior of Transaction Persistence Context

The following code illustrates the behavior of entites under an EntityManager using a transaction persistence context.

EntityManager em; // injected
...

// outside a transaction:

// each operation occurs in a separate persistence context, and returns 
// a new detached instance
Magazine mag1 = em.find(Magazine.class, magId);
Magazine mag2 = em.find(Magazine.class, magId);
assertTrue(mag2 != mag1);
...

// transaction begins:

// within a transaction, a subsequent lookup doesn't return any of the
// detached objects.  however, two lookups within the same transaction
// return the same instance, because the persistence context spans the
// transaction
Magazine mag3 = em.find(Magazine.class, magId);
assertTrue(mag3 != mag1 && mag3 != mag2);
Magazine mag4 = em.find(Magazine.class (magId);
assertTrue(mag4 == mag3);
...

// transaction commits:

// once again, each operation returns a new instance
Magazine mag5 = em.find(Magazine.class, magId);
assertTrue(mag5 != mag3);

3.2.  Extended Persistence Context

An EntityManager using an extended persistence context maintains the same persistence context for its entire lifecycle. Whether inside a transaction or not, all entities returned from the EntityManager are managed, and the EntityManager never creates two entity instances to represent the same persistent identity. Entities only become detached when you finally close the EntityManager (or when they are serialized).

Example 7.2.  Behavior of Extended Persistence Context

The following code illustrates the behavior of entites under an EntityManager using an extended persistence context.

EntityManagerFactory emf = ...
EntityManager em = emf.createEntityManager();

// persistence context active for entire life of EM, so only one entity
// for a given persistent identity
Magazine mag1 = em.find(Magazine.class, magId);
Magazine mag2 = em.find(Magazine.class, magId);
assertTrue(mag2 == mag1);

em.getTransaction().begin();

// same persistence context active within the transaction
Magazine mag3 = em.find(Magazine.class, magId);
assertTrue(mag3 == mag1);
Magazine mag4 = em.find(Magazine.class (magId);
assertTrue(mag4 == mag1);

em.getTransaction.commit ();

// when the transaction commits, instance still managed
Magazine mag5 = em.find(Magazine.class, magId);
assertTrue(mag5 == mag1);

// instance finally becomes detached when EM closes
em.close();

4.  Closing the EntityManagerFactory

public boolean isOpen ();
public void close ();

EntityManagerFactory instances are heavyweight objects. Each factory might maintain a metadata cache, object state cache, EntityManager pool, connection pool, and more. If your application no longer needs an EntityManagerFactory, you should close it to free these resources. When an EntityManagerFactory closes, all EntityManagers from that factory, and by extension all entities managed by those EntityManager s, become invalid. Attempting to close an EntityManagerFactory while one or more of its EntityManagers has an active transaction may result in an IllegalStateException.

Closing an EntityManagerFactory should not be taken lightly. It is much better to keep a factory open for a long period of time than to repeatedly create and close new factories. Thus, most applications will never close the factory, or only close it when the application is exiting. Only applications that require multiple factories with different configurations have an obvious reason to create and close multiple EntityManagerFactory instances. Once a factory is closed, all methods except isOpen throw an IllegalStateException.

Chapter 8.  EntityManager

The diagram above presents an overview of the EntityManager interface. For a complete treatment of the EntityManager API, see the Javadoc documentation. Methods whose parameter signatures consist of an ellipsis (...) are overloaded to take multiple parameter types.

Note

OpenJPA extends the standard EntityManager interface with the org.apache.openjpa.persistence.OpenJPAEntityManager interface to provide additional functionality.

The EntityManager is the primary interface used by application developers to interact with the JPA runtime. The methods of the EntityManager can be divided into the following functional categories:

  • Transaction association.

  • Entity lifecycle management.

  • Entity identity management.

  • Cache management.

  • Query factory.

  • Closing.

1.  Transaction Association

public EntityTransaction getTransaction ();

Every EntityManager has a one-to-one relation with an EntityTransaction instance. In fact, many vendors use a single class to implement both the EntityManager and EntityTransaction interfaces. If your application requires multiple concurrent transactions, you will use multiple EntityManagers.

You can retrieve the EntityTransaction associated with an EntityManager through the getTransaction method. Note that most most JPA implementations can integrate with an application server's managed transactions. If you take advantage of this feature, you will control transactions by declarative demarcation or through the Java Transaction API (JTA) rather than through the EntityTransaction.

2.  Entity Lifecycle Management

EntityManagers perform several actions that affect the lifecycle state of entity instances.

public void persist(Object entity);

Transitions new instances to managed. On the next flush or commit, the newly persisted instances will be inserted into the datastore.

For a given entity A, the persist method behaves as follows:

  • If A is a new entity, it becomes managed.

  • If A is an existing managed entity, it is ignored. However, the persist operation cascades as defined below.

  • If A is a removed entity, it becomes managed.

  • If A is a detached entity, an IllegalArgumentException is thrown.

  • The persist operation recurses on all relation fields of A whose cascades include CascadeType.PERSIST.

This action can only be used in the context of an active transaction.

public void remove(Object entity);

Transitions managed instances to removed. The instances will be deleted from the datastore on the next flush or commit. Accessing a removed entity has undefined results.

For a given entity A, the remove method behaves as follows:

  • If A is a new entity, it is ignored. However, the remove operation cascades as defined below.

  • If A is an existing managed entity, it becomes removed.

  • If A is a removed entity, it is ignored.

  • If A is a detached entity, an IllegalArgumentException is thrown.

  • The remove operation recurses on all relation fields of A whose cascades include CascadeType.REMOVE.

This action can only be used in the context of an active transaction.

public void refresh(Object entity);

Use the refresh action to make sure the persistent state of an instance is synchronized with the values in the datastore. refresh is intended for long-running optimistic transactions in which there is a danger of seeing stale data.

For a given entity A, the refresh method behaves as follows:

  • If A is a new entity, it is ignored. However, the remove operation cascades as defined below.

  • If A is an existing managed entity, its state is refreshed from the datastore.

  • If A is a removed entity, it is ignored.

  • If A is a detached entity, an IllegalArgumentException is thrown.

  • The refresh operation recurses on all relation fields of A whose cascades include CascadeType.REFRESH.

public Object merge(Object entity);

A common use case for an application running in a servlet or application server is to "detach" objects from all server resources, modify them, and then "attach" them again. For example, a servlet might store persistent data in a user session between a modification based on a series of web forms. Between each form request, the web container might decide to serialize the session, requiring that the stored persistent state be disassociated from any other resources. Similarly, a client/server application might transfer persistent objects to a client via serialization, allow the client to modify their state, and then have the client return the modified data in order to be saved. This is sometimes referred to as the data transfer object or value object pattern, and it allows fine-grained manipulation of data objects without incurring the overhead of multiple remote method invocations.

JPA provides support for this pattern by automatically detaching entities when they are serialized or when a persistence context ends (see Section 3, “ Persistence Context ” for an exploration of persistence contexts). The JPA merge API re-attaches detached entities. This allows you to detach a persistent instance, modify the detached instance offline, and merge the instance back into an EntityManager (either the same one that detached the instance, or a new one). The changes will then be applied to the existing instance from the datastore.

A detached entity maintains its persistent identity, but cannot load additional state from the datastore. Accessing any persistent field or property that was not loaded at the time of detachment has undefined results. Also, be sure not to alter the version or identity fields of detached instances if you plan on merging them later.

The merge method returns a managed copy of the given detached entity. Changes made to the persistent state of the detached entity are applied to this managed instance. Because merging involves changing persistent state, you can only merge within a transaction.

If you attempt to merge an instance whose representation has changed in the datastore since detachment, the merge operation will throw an exception, or the transaction in which you perform the merge will fail on commit, just as if a normal optimistic conflict were detected.

Note

OpenJPA offers enhancements to JPA detachment functionality, including additional options to control which fields are detached. See Section 1, “ Detach and Attach ” in the Reference Guide for details.

For a given entity A, the merge method behaves as follows:

  • If A is a detached entity, its state is copied into existing managed instance A' of the same entity identity, or a new managed copy of A is created.

  • If A is a new entity, a new managed entity A' is created and the state of A is copied into A'.

  • If A is an existing managed entity, it is ignored. However, the merge operation still cascades as defined below.

  • If A is a removed entity, an IllegalArgumentException is thrown.

  • The merge operation recurses on all relation fields of A whose cascades include CascadeType.MERGE.

public void lock (Object entity, LockModeType mode);

This method locks the given entity using the named mode. The javax.persistence.LockModeType enum defines two modes:

  • READ: Other transactions may concurrently read the object, but cannot concurrently update it.

  • WRITE: Other transactions cannot concurrently read or write the object. When a transaction is committed that holds WRITE locks on any entites, those entites will have their version incremented even if the entities themselves did not change in the transaction.

Note

OpenJPA has additional APIs for controlling object locking. See Section 3, “ Object Locking ” in the Reference Guide for details.

The following diagram illustrates the lifecycle of an entity with respect to the APIs presented in this section.

3.  Lifecycle Examples

The examples below demonstrate how to use the lifecycle methods presented in the previous section. The examples are appropriate for out-of-container use. Within a container, EntityManagers are usually injected, and transactions are usually managed. You would therefore omit the createEntityManager and close calls, as well as all transaction demarcation code.

Example 8.1.  Persisting Objects

// create some objects
Magazine mag = new Magazine("1B78-YU9L", "JavaWorld");

Company pub = new Company("Weston House");
pub.setRevenue(1750000D);
mag.setPublisher(pub);
pub.addMagazine(mag);

Article art = new Article("JPA Rules!", "Transparent Object Persistence");
art.addAuthor(new Author("Fred", "Hoyle"));
mag.addArticle(art);

// persist
EntityManager em = emf.createEntityManager();
em.getTransaction().begin();
em.persist(mag);
em.persist(pub);
em.persist(art);
em.getTransaction().commit();

// or we could continue using the EntityManager...
em.close();

Example 8.2.  Updating Objects

Magazine.MagazineId mi = new Magazine.MagazineId();
mi.isbn = "1B78-YU9L";
mi.title = "JavaWorld";

// updates should always be made within transactions; note that
// there is no code explicitly linking the magazine or company
// with the transaction; JPA automatically tracks all changes
EntityManager em = emf.createEntityManager();
em.getTransaction().begin();
Magazine mag = em.find(Magazine.class, mi);
mag.setPrice(5.99);
Company pub = mag.getPublisher();
pub.setRevenue(1750000D);
em.getTransaction().commit();

// or we could continue using the EntityManager...
em.close();

Example 8.3.  Removing Objects

// assume we have an object id for the company whose subscriptions
// we want to delete
Object oid = ...;

// deletes should always be made within transactions
EntityManager em = emf.createEntityManager();
em.getTransaction().begin();
Company pub = (Company) em.find(Company.class, oid);
for (Subscription sub : pub.getSubscriptions())
    em.remove(sub);
pub.getSubscriptions().clear();
em.getTransaction().commit();

// or we could continue using the EntityManager...
em.close();

Example 8.4.  Detaching and Merging

This example demonstrates a common client/server scenario. The client requests objects and makes changes to them, while the server handles the object lookups and transactions.

// CLIENT:
// requests an object with a given oid
Record detached = (Record) getFromServer(oid);

...

// SERVER:
// send object to client; object detaches on EM close
Object oid = processClientRequest();
EntityManager em = emf.createEntityManager();
Record record = em.find(Record.class, oid);
em.close();
sendToClient(record);

...

// CLIENT:
// makes some modifications and sends back to server
detached.setSomeField("bar");
sendToServer(detached);

...

// SERVER:
// merges the instance and commit the changes
Record modified = (Record) processClientRequest();
EntityManager em = emf.createEntityManager();
em.getTransaction().begin();
Record merged = (Record) em.merge(modified);
merged.setLastModified(System.currentTimeMillis());
merged.setModifier(getClientIdentityCode());
em.getTransaction().commit();
em.close();

4.  Entity Identity Management

Each EntityManager is responsible for managing the persistent identities of the managed objects in the persistence context. The following methods allow you to interact with the management of persistent identities. The behavior of these methods is deeply affected by the persistence context type of the EntityManager; see Section 3, “ Persistence Context ” for an explanation of persistence contexts.

public <T> T find(Class<T> cls, Object oid);

This method returns the persistent instance of the given type with the given persistent identity. If the instance is already present in the current persistence context, the cached version will be returned. Otherwise, a new instance will be constructed and loaded with state from the datastore. If no entity with the given type and identity exists in the datastore, this method returns null.

public <T> T getReference(Class<T> cls, Object oid);

This method is similar to find, but does not necessarily go to the database when the entity is not found in cache. The implementation may construct a hollow entity and return it to you instead. Hollow entities do not have any state loaded. The state only gets loaded when you attempt to access a persistent field. At that time, the implementation may throw an EntityNotFoundException if it discovers that the entity does not exist in the datastore. The implementation may also throw an EntityNotFoundException from the getReference method itself. Unlike find, getReference does not return null.

public boolean contains(Object entity);

Returns true if the given entity is part of the current persistence context, and false otherwise. Removed entities are not considered part of the current persistence context.

5.  Cache Management

public void flush();

The flush method writes any changes that have been made in the current transaction to the datastore. If the EntityManager does not already have a connection to the datastore, it obtains one for the flush and retains it for the duration of the transaction. Any exceptions during flush cause the transaction to be marked for rollback. See Chapter 9, Transaction .

Flushing requires an active transaction. If there isn't a transaction in progress, the flush method throws a TransactionRequiredException.

public FlushModeType getFlushMode();
public void setFlushMode(FlushModeType flushMode);

The EntityManager's FlushMode property controls whether to flush transactional changes before executing queries. This allows the query results to take into account changes you have made during the current transaction. Available javax.persistence.FlushModeType constants are:

  • COMMIT: Only flush when committing, or when told to do so through the flush method. Query results may not take into account changes made in the current transaction.

  • AUTO: The implementation is permitted to flush before queries to ensure that the results reflect the most recent object state.

You can also set the flush mode on individual Query instances.

Note

OpenJPA only flushes before a query if the query might be affected by data changed in the current transaction. Additionally, OpenJPA allows fine-grained control over flushing behavior. See the Reference Guide's Section 8, “ Configuring the Use of JDBC Connections ”.

public void clear();

Clearing the EntityManager effectively ends the persistence context. All entities managed by the EntityManager become detached.

6.  Query Factory

public Query createQuery(String query);

Query objects are used to find entities matching certain criteria. The createQuery method creates a query using the given Java Persistence Query Language (JPQL) string. See Chapter 10, JPA Query for details.

public Query createNamedQuery(String name);

This method retrieves a query defined in metadata by name. The returned Query instance is initialized with the information declared in metadata. For more information on named queries, read Section 1.9, “ Named Queries ”.

public Query createNativeQuery(String sql);
public Query createNativeQuery(String sql, Class resultCls);
public Query createNativeQuery(String sql, String resultMapping);

Native queries are queries in the datastore's native language. For relational databases, this the Structured Query Language (SQL). Chapter 11, SQL Queries elaborates on JPA's native query support.

7.  Closing

public boolean isOpen();
public void close();

When an EntityManager is no longer needed, you should call its close method. Closing an EntityManager releases any resources it is using. The persistence context ends, and the entities managed by the EntityManager become detached. Any Query instances the EntityManager created become invalid. Calling any method other than isOpen on a closed EntityManager results in an IllegalStateException. You cannot close a EntityManager that is in the middle of a transaction.

If you are in a managed environment using injected entity managers, you should not close them.

Chapter 9.  Transaction

Transactions are critical to maintaining data integrity. They are used to group operations into units of work that act in an all-or-nothing fashion. Transactions have the following qualities:

  • Atomicity. Atomicity refers to the all-or-nothing property of transactions. Either every data update in the transaction completes successfully, or they all fail, leaving the datastore in its original state. A transaction cannot be only partially successful.

  • Consistency. Each transaction takes the datastore from one consistent state to another consistent state.

  • Isolation. Transactions are isolated from each other. When you are reading persistent data in one transaction, you cannot "see" the changes that are being made to that data in other transactions. Similarly, the updates you make in one transaction cannot conflict with updates made in concurrent transactions. The form of conflict resolution employed depends on whether you are using pessimistic or optimistic transactions. Both types are described later in this chapter.

  • Durability. The effects of successful transactions are durable; the updates made to persistent data last for the lifetime of the datastore.

Together, these qualities are called the ACID properties of transactions. To understand why these properties are so important to maintaining data integrity, consider the following example:

Suppose you create an application to manage bank accounts. The application includes a method to transfer funds from one user to another, and it looks something like this:

public void transferFunds(User from, User to, double amnt) {
    from.decrementAccount(amnt);
    to.incrementAccount(amnt);
}

Now suppose that user Alice wants to transfer 100 dollars to user Bob. No problem; you simply invoke your transferFunds method, supplying Alice in the from parameter, Bob in the to parameter, and 100.00 as the amnt . The first line of the method is executed, and 100 dollars is subtracted from Alice's account. But then, something goes wrong. An unexpected exception occurs, or the hardware fails, and your method never completes.

You are left with a situation in which the 100 dollars has simply disappeared. Thanks to the first line of your method, it is no longer in Alice's account, and yet it was never transferred to Bob's account either. The datastore is in an inconsistent state.

The importance of transactions should now be clear. If the two lines of the transferFunds method had been placed together in a transaction, it would be impossible for only the first line to succeed. Either the funds would be transferred properly or they would not be transferred at all, and an exception would be thrown. Money could never vanish into thin air, and the data store could never get into an inconsistent state.

1.  Transaction Types

There are two major types of transactions: pessimistic transactions and optimistic transactions. Each type has both advantages and disadvantages.

Pessimistic transactions generally lock the datastore records they act on, preventing other concurrent transactions from using the same data. This avoids conflicts between transactions, but consumes database resources. Additionally, locking records can result in deadlock, a situation in which two transactions are both waiting for the other to release its locks before completing. The results of a deadlock are datastore-dependent; usually one transaction is forcefully rolled back after some specified timeout interval, and an exception is thrown.

This document will often use the term datastore transaction in place of pessimistic transaction. This is to acknowledge that some datastores do not support pessimistic semantics, and that the exact meaning of a non-optimistic JPA transaction is dependent on the datastore. Most of the time, a datastore transaction is equivalent to a pessimistic transaction.

Optimistic transactions consume less resources than pessimistic/datastore transactions, but only at the expense of reliability. Because optimistic transactions do not lock datastore records, two transactions might change the same persistent information at the same time, and the conflict will not be detected until the second transaction attempts to flush or commit. At this time, the second transaction will realize that another transaction has concurrently modified the same records (usually through a timestamp or versioning system), and will throw an appropriate exception. Note that optimistic transactions still maintain data integrity; they are simply more likely to fail in heavily concurrent situations.

Despite their drawbacks, optimistic transactions are the best choice for most applications. They offer better performance, better scalability, and lower risk of hanging due to deadlock.

Note

OpenJPA uses optimistic semantics by default, but supports both optimistic and datastore transactions. OpenJPA also offers advanced locking and versioning APIs for fine-grained control over database resource allocation and object versioning. See Section 3, “ Object Locking ” of the Reference Guide for details on locking. Section 2.5, “ Version ” of this document covers standard object versioning, while Section 7, “ Additional JPA Mappings ” of the Reference Guide discusses additional versioning strategies available in OpenJPA.

2.  The EntityTransaction Interface

JPA integrates with your container's managed transactions, allowing you to use the container's declarative transaction demarcation and its Java Transaction API (JTA) implementation for transaction management. Outside of a container, though, you must demarcate transactions manually through JPA. The EntityTransaction interface controls unmanaged transactions in JPA.

public void begin();
public void commit();
public void rollback();

The begin, commit, and rollback methods demarcate transaction boundaries. The methods should be self-explanatory: begin starts a transaction, commit attempts to commit the transaction's changes to the datastore, and rollback aborts the transaction, in which case the datastore is "rolled back" to its previous state. JPA implementations will automatically roll back transactions if any exception is thrown during the commit process.

Unless you are using an extended persistence context, committing or rolling back also ends the persistence context. All managed entites will be detached from the EntityManager.

public boolean isActive();

Finally, the isActive method returns true if the transaction is in progress (begin has been called more recently than commit or rollback), and false otherwise.

Example 9.1.  Grouping Operations with Transactions

public void transferFunds(EntityManager em, User from, User to, double amnt) {
    // note: it would be better practice to move the transaction demarcation
    // code out of this method, but for the purposes of example...
    Transaction trans = em.getTransaction();
    trans.begin();
    try
    {
        from.decrementAccount(amnt);
        to.incrementAccount(amnt);
        trans.commit();
    }
    catch (RuntimeException re)
    {
        if (trans.isActive())
            trans.rollback();   // or could attempt to fix error and retry
        throw re;
    }
}

Chapter 10.  JPA Query

Table of Contents

1. JPQL API
1.1. Query Basics
1.2. Relation Traversal
1.3. Fetch Joins
1.4. JPQL Functions
1.5. Polymorphic Queries
1.6. Query Parameters
1.7. Ordering
1.8. Aggregates
1.9. Named Queries
1.10. Delete By Query
1.11. Update By Query
2. JPQL Language Reference
2.1. JPQL Statement Types
2.1.1. JPQL Select Statement
2.1.2. JPQL Update and Delete Statements
2.2. JPQL Abstract Schema Types and Query Domains
2.2.1. JPQL Entity Naming
2.2.2. JPQL Schema Example
2.3. JPQL FROM Clause and Navigational Declarations
2.3.1. JPQL FROM Identifiers
2.3.2. JPQL Identification Variables
2.3.3. JPQL Range Declarations
2.3.4. JPQL Path Expressions
2.3.5. JPQL Joins
2.3.5.1. JPQL Inner Joins (Relationship Joins)
2.3.5.2. JPQL Outer Joins
2.3.5.3. JPQL Fetch Joins
2.3.6. JPQL Collection Member Declarations
2.3.7. JPQL Polymorphism
2.4. JPQL WHERE Clause
2.5. JPQL Conditional Expressions
2.5.1. JPQL Literals
2.5.2. JPQL Identification Variables
2.5.3. JPQL Path Expressions
2.5.4. JPQL Input Parameters
2.5.4.1. JPQL Positional Parameters
2.5.4.2. JPQL Named Parameters
2.5.5. JPQL Conditional Expression Composition
2.5.6. JPQL Operators and Operator Precedence
2.5.7. JPQL Between Expressions
2.5.8. JPQL In Expressions
2.5.9. JPQL Like Expressions
2.5.10. JPQL Null Comparison Expressions
2.5.11. JPQL Empty Collection Comparison Expressions
2.5.12. JPQL Collection Member Expressions
2.5.13. JPQL Exists Expressions
2.5.14. JPQL All or Any Expressions
2.5.15. JPQL Subqueries
2.5.16. JPQL Functional Expressions
2.5.16.1. JPQL String Functions
2.5.16.2. JPQL Arithmetic Functions
2.5.16.3. JPQL Datetime Functions
2.6. JPQL GROUP BY, HAVING
2.7. JPQL SELECT Clause
2.7.1. JPQL Result Type of the SELECT Clause
2.7.2. JPQL Constructor Expressions
2.7.3. JPQL Null Values in the Query Result
2.7.4. JPQL Aggregate Functions
2.7.4.1. JPQL Aggregate Examples
2.8. JPQL ORDER BY Clause
2.9. JPQL Bulk Update and Delete
2.10. JPQL Null Values
2.11. JPQL Equality and Comparison Semantics
2.12. JPQL BNF

The javax.persistence.Query interface is the mechanism for issuing queries in JPA. The primary query language used is the Java Persistence Query Language, or JPQL. JPQL is syntactically very similar to SQL, but is object-oriented rather than table-oriented.

The API for executing JPQL queries will be discussed in Section 1, “ JPQL API ”, and a full language reference will be covered in Section 2, “ JPQL Language Reference ”.

1.  JPQL API

1.1.  Query Basics

SELECT x FROM Magazine x

The preceding is a simple JPQL query for all Magazine entities.

public Query createQuery(String jpql);

The EntityManager.createQuery method creates a Query instance from a given JPQL string.

public List getResultList();

Invoking Query.getResultList executes the query and returns a List containing the matching objects. The following example executes our Magazine query above:

EntityManager em = ...
Query q = em.createQuery("SELECT x FROM Magazine x");
List<Magazine> results = (List<Magazine>) q.getResultList();

A JPQL query has an internal namespace declared in the from clause of the query. Arbitrary identifiers are assigned to entities so that they can be referenced elsewhere in the query. In the query example above, the identifier x is assigned to the entity Magazine .

Note

The as keyword can optionally be used when declaring identifiers in the from clause. SELECT x FROM Magazine x and SELECT x FROM Magazine AS x are synonymous.

Following the select clause of the query is the object or objects that the query returns. In the case of the query above, the query's result list will contain instances of the Magazine class.

Note

When selecting entities, you can optional use the keyword object . The clauses select x and SELECT OBJECT(x) are synonymous.

The optional where clause places criteria on matching results. For example:

SELECT x FROM Magazine x WHERE x.title = 'JDJ'

Keywords in JPQL expressions are case-insensitive, but entity, identifier, and member names are not. For example, the expression above could also be expressed as:

SELECT x FROM Magazine x WHERE x.title = 'JDJ'

But it could not be expressed as:

SELECT x FROM Magazine x WHERE x.TITLE = 'JDJ'

As with the select clause, alias names in the where clause are resolved to the entity declared in the from clause. The query above could be described in English as "for all Magazine instances x, return a list of every x such that x's title field is equal to 'JDJ'".

JPQL uses SQL-like syntax for query criteria. The and and or logical operators chain multiple criteria together:

SELECT x FROM Magazine x WHERE x.title = 'JDJ' OR x.title = 'JavaPro'

The = operator tests for equality. <> tests for inequality. JPQL also supports the following arithmetic operators for numeric comparisons: >, >=, <, <=. For example:

SELECT x FROM Magazine x WHERE x.price > 3.00 AND x.price <= 5.00

This query returns all magazines whose price is greater than 3.00 and less than or equal to 5.00.

SELECT x FROM Magazine x WHERE x.price <> 3.00

This query returns all Magazines whose price is not equals to 3.00.

You can group expressions together using parentheses in order to specify how they are evaluated. This is similar to how parentheses are used in Java. For example:

SELECT x FROM Magazine x WHERE (x.price > 3.00 AND x.price <= 5.00) OR x.price = 7.00

This expression would match magazines whose price is 4.00, 5.00, or 7.00, but not 6.00. Alternately:

SELECT x FROM Magazine x WHERE x.price > 3.00 AND (x.price <= 5.00 OR x.price = 7.00)

This expression will magazines whose price is 5.00 or 7.00, but not 4.00 or 6.00.

JPQL also includes the following conditionals:

  • [NOT] BETWEEN: Shorthand for expressing that a value falls between two other values. The following two statements are synonymous:

    SELECT x FROM Magazine x WHERE x.price >= 3.00 AND x.price <= 5.00
    
    SELECT x FROM Magazine x WHERE x.price BETWEEN 3.00 AND 5.00
    
  • [NOT] LIKE: Performs a string comparison with wildcard support. The special character '_' in the parameter means to match any single character, and the special character '%' means to match any sequence of characters. The following statement matches title fields "JDJ" and "JavaPro", but not "IT Insider":

    SELECT x FROM Magazine x WHERE x.title LIKE 'J%'
    

    The following statement matches the title field "JDJ" but not "JavaPro":

    SELECT x FROM Magazine x WHERE x.title LIKE 'J__'
    
  • [NOT] IN: Specifies that the member must be equal to one element of the provided list. The following two statements are synonymous:

    SELECT x FROM Magazine x WHERE x.title IN ('JDJ', 'JavaPro', 'IT Insider')
    
    SELECT x FROM Magazine x WHERE x.title = 'JDJ' OR x.title = 'JavaPro' OR x.title = 'IT Insider'
    
  • IS [NOT] EMPTY: Specifies that the collection field holds no elements. For example:

    SELECT x FROM Magazine x WHERE x.articles is empty
    

    This statement will return all magazines whose articles member contains no elements.

  • IS [NOT] NULL: Specifies that the field is equal to null. For example:

    SELECT x FROM Magazine x WHERE x.publisher is null
    

    This statement will return all Magazine instances whose "publisher" field is set to null.

  • NOT: Negates the contained expression. For example, the following two statements are synonymous:

    SELECT x FROM Magazine x WHERE NOT(x.price = 10.0)
    
    SELECT x FROM Magazine x WHERE x.price <> 10.0
    

1.2.  Relation Traversal

Relations between objects can be traversed using Java-like syntax. For example, if the Magazine class has a field named "publisher" or type Company, that relation can be queried as follows:

SELECT x FROM Magazine x WHERE x.publisher.name = 'Random House'

This query returns all Magazine instances whose publisher field is set to a Company instance whose name is "Random House".

Single-valued relation traversal implies that the relation is not null. In SQL terms, this is known as an inner join. If you want to also include relations that are null, you can specify:

SELECT x FROM Magazine x WHERE x.publisher.name = 'Random House' or x.publisher is null

You can also traverse collection fields in queries, but you must declare each traversal in the from clause. Consider:

SELECT x FROM Magazine x, IN(x.articles) y WHERE y.authorName = 'John Doe'

This query says that for each Magazine x , traverse the articles relation and check each Articley, and pass the filter if y's authorName field is equal to "John Doe". In short, this query will return all magazines that have any articles written by John Doe.

Note

The IN() syntax can also be expressed with the keywords inner join. The statements SELECT x FROM Magazine x, IN(x.articles) y WHERE y.authorName = 'John Doe' and SELECT x FROM Magazine x inner join x.articles y WHERE y.authorName = 'John Doe' are synonymous.

1.3.  Fetch Joins

JPQL queries may specify one or more join fetch declarations, which allow the query to specify which fields in the returned instances will be pre-fetched.

SELECT x FROM Magazine x join fetch x.articles WHERE x.title = 'JDJ'

The query above returns Magazine instances and guarantees that the articles field will already be fetched in the returned instances.

Multiple fields may be specified in separate join fetch declarations:

SELECT x FROM Magazine x join fetch x.articles join fetch x.authors WHERE x.title = 'JDJ'

Note

Specifying the join fetch declaration is functionally equivalent to adding the fields to the Query's FetchConfiguration. See Section 6, “ Fetch Groups ”.

1.4.  JPQL Functions

As well as supporting direct field and relation comparisons, JPQL supports a pre-defined set of functions that you can apply.

  • CONCAT(string1, string2): Concatenates two string fields or literals. For example:

    SELECT x FROM Magazine x WHERE CONCAT(x.title, 's') = 'JDJs'
    
  • SUBSTRING(string, startIndex, length): Returns the part of the string argument starting at startIndex (1-based) and ending at length characters past startIndex.

    SELECT x FROM Magazine x WHERE SUBSTRING(x.title, 1, 1) = 'J'
    
  • TRIM([LEADING | TRAILING | BOTH] [character FROM] string: Trims the specified character from either the beginning ( LEADING ) end ( TRAILING) or both ( BOTH ) of the string argument. If no trim character is specified, the space character will be trimmed.

    SELECT x FROM Magazine x WHERE TRIM(BOTH 'J' FROM x.title) = 'D'
    
  • LOWER(string): Returns the lower-case of the specified string argument.

    SELECT x FROM Magazine x WHERE LOWER(x.title) = 'jdj'
    
  • UPPER(string): Returns the upper-case of the specified string argument.

    SELECT x FROM Magazine x WHERE UPPER(x.title) = 'JAVAPRO'
    
  • LENGTH(string): Returns the number of characters in the specified string argument.

    SELECT x FROM Magazine x WHERE LENGTH(x.title) = 3
    
  • LOCATE(searchString, candidateString [, startIndex]): Returns the first index of searchString in candidateString. Positions are 1-based. If the string is not found, returns 0.

    SELECT x FROM Magazine x WHERE LOCATE('D', x.title) = 2
    
  • ABS(number): Returns the absolute value of the argument.

    SELECT x FROM Magazine x WHERE ABS(x.price) >= 5.00
    
  • SQRT(number): Returns the square root of the argument.

    SELECT x FROM Magazine x WHERE SQRT(x.price) >= 1.00
    
  • MOD(number, divisor): Returns the modulo of number and divisor.

    SELECT x FROM Magazine x WHERE MOD(x.price, 10) = 0
    
  • CURRENT_DATE: Returns the current date.

  • CURRENT_TIME: Returns the current time.

  • CURRENT_TIMESTAMP: Returns the current timestamp.

1.5.  Polymorphic Queries

All JPQL queries are polymorphic, which means the from clause of a query includes not only instances of the specific entity class to which it refers, but all subclasses of that class as well. The instances returned by a query include instances of the subclasses that satisfy the query conditions. For example, the following query may return instances of Magazine , as well as Tabloid and Digest instances, where Tabloid and Digest are Magazine subclasses.

SELECT x FROM Magazine x WHERE x.price < 5

1.6.  Query Parameters

JPQL provides support for parameterized queries. Either named parameters or positional parameters may be specified in the query string. Parameters allow you to re-use query templates where only the input parameters vary. A single query can declare either named parameters or positional parameters, but is not allowed to declare both named and positional parameters.

public Query setParameter (int pos, Object value);

Specify positional parameters in your JPQL string using an integer prefixed by a question mark. You can then populate the Query object with positional parameter values via calls to the setParameter method above. The method returns the Query instance for optional method chaining.

EntityManager em = ...
Query q = em.createQuery("SELECT x FROM Magazine x WHERE x.title = ?1 and x.price > ?2");
q.setParameter(1, "JDJ").setParameter(2, 5.0);
List<Magazine> results = (List<Magazine>) q.getResultList();

This code will substitute JDJ for the ?1 parameter and 5.0 for the ?2 parameter, then execute the query with those values.

public Query setParameter(String name, Object value);

Named parameter are denoted by prefixing an arbitrary name with a colon in your JPQL string. You can then populate the Query object with parameter values using the method above. Like the positional parameter method, this method returns the Query instance for optional method chaining.

EntityManager em = ...
Query q = em.createQuery("SELECT x FROM Magazine x WHERE x.title = :titleParam and x.price > :priceParam");
q.setParameter("titleParam", "JDJ").setParameter("priceParam", 5.0);
List<Magazine> results = (List<Magazine>) q.getResultList();

This code substitutes JDJ for the :titleParam parameter and 5.0 for the :priceParam parameter, then executes the query with those values.

1.7.  Ordering

JPQL queries may optionally contain an order by clause which specifies one or more fields to order by when returning query results. You may follow the order by field clause with the asc or desc keywords, which indicate that ordering should be ascending or descending, respectively. If the direction is omitted, ordering is ascending by default.

SELECT x FROM Magazine x order by x.title asc, x.price desc

The query above returns Magazine instances sorted by their title in ascending order. In cases where the titles of two or more magazines are the same, those instances will be sorted by price in descending order.

1.8.  Aggregates

JPQL queries can select aggregate data as well as objects. JPQL includes the min, max, avg, and count aggregates. These functions can be used for reporting and summary queries.

The following query will return the average of all the prices of all the magazines:

EntityManager em = ...
Query q = em.createQuery("SELECT AVG(x.price) FROM Magazine x");
Number result = (Number) q.getSingleResult();

The following query will return the highest price of all the magazines titled "JDJ":

EntityManager em = ...
Query q = em.createQuery("SELECT MAX(x.price) FROM Magazine x WHERE x.title = 'JDJ'");
Number result = (Number) q.getSingleResult();

1.9.  Named Queries

Query templates can be statically declared using the NamedQuery and NamedQueries annotations. For example:

@Entity
@NamedQueries({
    @NamedQuery(name="magsOverPrice",
        query="SELECT x FROM Magazine x WHERE x.price > ?1"),
    @NamedQuery(name="magsByTitle",
        query="SELECT x FROM Magazine x WHERE x.title = :titleParam")
})
public class Magazine {
    ...
}

These declarations will define two named queries called magsOverPrice and magsByTitle.

public Query createNamedQuery(String name);

You retrieve named queries with the above EntityManager method. For example:

EntityManager em = ...
Query q = em.createNamedQuery("magsOverPrice");
q.setParameter(1, 5.0f);
List<Magazine> results = (List<Magazine>) q.getResultList();
EntityManager em = ...
Query q = em.createNamedQuery("magsByTitle");
q.setParameter("titleParam", "JDJ");
List<Magazine> results = (List<Magazine>) q.getResultList();

1.10.  Delete By Query

Queries are useful not only for finding objects, but for efficiently deleting them as well. For example, you might delete all records created before a certain date. Rather than bring these objects into memory and delete them individually, JPA allows you to perform a single bulk delete based on JPQL criteria.

Delete by query uses the same JPQL syntax as normal queries, with one exception: begin your query string with the delete keyword instead of the select keyword. To then execute the delete, you call the following Query method:

public int executeUpdate();

This method returns the number of objects deleted. The following example deletes all subscriptions whose expiration date has passed.

Example 10.1.  Delete by Query

Query q = em.createQuery("DELETE s FROM Subscription s WHERE s.subscriptionDate < :today");
q.setParameter("today", new Date());
int deleted = q.executeUpdate();

1.11.  Update By Query

Similar to bulk deletes, it is sometimes necessary to perform updates against a large number of queries in a single operation, without having to bring all the instances down to the client. Rather than bring these objects into memory and modifying them individually, JPA allows you to perform a single bulk update based on JPQL criteria.

Update by query uses the same JPQL syntax as normal queries, except that the query string begins with the update keyword instead of select. To execute the update, you call the following Query method:

public int executeUpdate();

This method returns the number of objects updated. The following example updates all subscriptions whose expiration date has passed to have the "paid" field set to true..

Example 10.2.  Update by Query

Query q = em.createQuery("UPDATE Subscription s SET s.paid = :paid WHERE s.subscriptionDate < :today");
q.setParameter("today", new Date());
q.setParameter("paid", true);
int updated = q.executeUpdate();

2.  JPQL Language Reference

2.1. JPQL Statement Types
2.1.1. JPQL Select Statement
2.1.2. JPQL Update and Delete Statements
2.2. JPQL Abstract Schema Types and Query Domains
2.2.1. JPQL Entity Naming
2.2.2. JPQL Schema Example
2.3. JPQL FROM Clause and Navigational Declarations
2.3.1. JPQL FROM Identifiers
2.3.2. JPQL Identification Variables
2.3.3. JPQL Range Declarations
2.3.4. JPQL Path Expressions
2.3.5. JPQL Joins
2.3.5.1. JPQL Inner Joins (Relationship Joins)
2.3.5.2. JPQL Outer Joins
2.3.5.3. JPQL Fetch Joins
2.3.6. JPQL Collection Member Declarations
2.3.7. JPQL Polymorphism
2.4. JPQL WHERE Clause
2.5. JPQL Conditional Expressions
2.5.1. JPQL Literals
2.5.2. JPQL Identification Variables
2.5.3. JPQL Path Expressions
2.5.4. JPQL Input Parameters
2.5.4.1. JPQL Positional Parameters
2.5.4.2. JPQL Named Parameters
2.5.5. JPQL Conditional Expression Composition
2.5.6. JPQL Operators and Operator Precedence
2.5.7. JPQL Between Expressions
2.5.8. JPQL In Expressions
2.5.9. JPQL Like Expressions
2.5.10. JPQL Null Comparison Expressions
2.5.11. JPQL Empty Collection Comparison Expressions
2.5.12. JPQL Collection Member Expressions
2.5.13. JPQL Exists Expressions
2.5.14. JPQL All or Any Expressions
2.5.15. JPQL Subqueries
2.5.16. JPQL Functional Expressions
2.5.16.1. JPQL String Functions
2.5.16.2. JPQL Arithmetic Functions
2.5.16.3. JPQL Datetime Functions
2.6. JPQL GROUP BY, HAVING
2.7. JPQL SELECT Clause
2.7.1. JPQL Result Type of the SELECT Clause
2.7.2. JPQL Constructor Expressions
2.7.3. JPQL Null Values in the Query Result
2.7.4. JPQL Aggregate Functions
2.7.4.1. JPQL Aggregate Examples
2.8. JPQL ORDER BY Clause
2.9. JPQL Bulk Update and Delete
2.10. JPQL Null Values
2.11. JPQL Equality and Comparison Semantics
2.12. JPQL BNF

The Java Persistence Query Language (JPQL) is used to define searches against persistent entities independent of the mechanism used to store those entities. As such, JPQL is "portable", and not constrained to any particular data store. The Java Persistence query language is an extension of the Enterprise JavaBeans query language, EJB QL, adding operations such as bulk deletes and updates, join operations, aggregates, projections, and subqueries. Furthermore, JPQL queries can be declared statically in metadata, or can be dynamically built in code. This chapter provides the full definition of the language.

Note

Much of this section is paraphrased or taken directly from Chapter 4 of the JSR 220 specification.

2.1.  JPQL Statement Types

A JPQL statement may be either a SELECT statement, an UPDATE statement, or a DELETE statement. This chapter refers to all such statements as "queries". Where it is important to distinguish among statement types, the specific statement type is referenced. In BNF syntax, a query language statement is defined as:

  • QL_statement ::= select_statement | update_statement | delete_statement

The complete BNF for JPQL is defined in Section 2.12, “ JPQL BNF ”. Any JPQL statement may be constructed dynamically or may be statically defined in a metadata annotation or XML descriptor element. All statement types may have parameters, as discussed in Section 2.5.4, “ JPQL Input Parameters ”.

2.1.1.  JPQL Select Statement

A select statement is a string which consists of the following clauses:

  • a SELECT clause, which determines the type of the objects or values to be selected.

  • a FROM clause, which provides declarations that designate the domain to which the expressions specified in the other clauses of the query apply.

  • an optional WHERE clause, which may be used to restrict the results that are returned by the query.

  • an optional GROUP BY clause, which allows query results to be aggregated in terms of groups.

  • an optional HAVING clause, which allows filtering over aggregated groups.

  • an optional ORDER BY clause, which may be used to order the results that are returned by the query.

In BNF syntax, a select statement is defined as:

  • select_statement ::= select_clause from_clause [where_clause] [groupby_clause] [having_clause] [orderby_clause]

A select statement must always have a SELECT and a FROM clause. The square brackets [] indicate that the other clauses are optional.

2.1.2.  JPQL Update and Delete Statements

Update and delete statements provide bulk operations over sets of entities. In BNF syntax, these operations are defined as:

  • update_statement ::= update_clause [where_clause]

  • delete_statement ::= delete_clause [where_clause]

The update and delete clauses determine the type of the entities to be updated or deleted. The WHERE clause may be used to restrict the scope of the update or delete operation. Update and delete statements are described further in Section 2.9, “ JPQL Bulk Update and Delete ”.

2.2.  JPQL Abstract Schema Types and Query Domains

The Java Persistence query language is a typed language, and every expression has a type. The type of an expression is derived from the structure of the expression, the abstract schema types of the identification variable declarations, the types to which the persistent fields and relationships evaluate, and the types of literals. The abstract schema type of an entity is derived from the entity class and the metadata information provided by Java language annotations or in the XML descriptor.

Informally, the abstract schema type of an entity can be characterized as follows:

  • For every persistent field or get accessor method (for a persistent property) of the entity class, there is a field ("state-field") whose abstract schema type corresponds to that of the field or the result type of the accessor method.

  • For every persistent relationship field or get accessor method (for a persistent relationship property) of the entity class, there is a field ("association-field") whose type is the abstract schema type of the related entity (or, if the relationship is a one-to-many or many-to-many, a collection of such). Abstract schema types are specific to the query language data model. The persistence provider is not required to implement or otherwise materialize an abstract schema type. The domain of a query consists of the abstract schema types of all entities that are defined in the same persistence unit. The domain of a query may be restricted by the navigability of the relationships of the entity on which it is based. The association-fields of an entity's abstract schema type determine navigability. Using the association-fields and their values, a query can select related entities and use their abstract schema types in the query.

2.2.1.  JPQL Entity Naming

Entities are designated in query strings by their entity names. The entity name is defined by the name element of the Entity annotation (or the entity-name XML descriptor element), and defaults to the unqualified name of the entity class. Entity names are scoped within the persistence unit and must be unique within the persistence unit.

2.2.2.  JPQL Schema Example

This example assumes that the application developer provides several entity classes, representing magazines, publishers, authors, and articles. The abstract schema types for these entities are Magazine, Publisher, Author, and Article.

Several Entities with Abstract Persistence Schemas Defined in the Same Persistence Unit. The entity Publisher has a one-to-many relationships with Magazine. There is also a one-to-many relationship between Magazine and Article . The entity Article is related to Author in a one-to-one relationship.

Queries to select magazines can be defined by navigating over the association-fields and state-fields defined by Magazine and Author. A query to find all magazines that have unpublished articles is as follows:

SELECT DISTINCT mag FROM Magazine AS mag JOIN mag.articles AS art WHERE art.published = FALSE

This query navigates over the association-field authors of the abstract schema type Magazine to find articles, and uses the state-field published of Article to select those magazines that have at least one article that is published. Although predefined reserved identifiers, such as DISTINCT, FROM, AS, JOIN, WHERE, and FALSE appear in upper case in this example, predefined reserved identifiers are case insensitive. The SELECT clause of this example designates the return type of this query to be of type Magazine. Because the same persistence unit defines the abstract persistence schemas of the related entities, the developer can also specify a query over articles that utilizes the abstract schema type for products, and hence the state-fields and association-fields of both the abstract schema types Magazine and Author. For example, if the abstract schema type Author has a state-field named firstName, a query over articles can be specified using this state-field. Such a query might be to find all magazines that have articles authored by someone with the first name "John".

SELECT DISTINCT mag FROM Magazine mag JOIN mag.articles art JOIN art.author auth WHERE auth.firstName = 'John'

Because Magazine is related to Author by means of the relationships between Magazine and Article and between Article and Author, navigation using the association-fields authors and product is used to express the query. This query is specified by using the abstract schema name Magazine, which designates the abstract schema type over which the query ranges. The basis for the navigation is provided by the association-fields authors and product of the abstract schema types Magazine and Article respectively.

2.3.  JPQL FROM Clause and Navigational Declarations

The FROM clause of a query defines the domain of the query by declaring identification variables. An identification variable is an identifier declared in the FROM clause of a query. The domain of the query may be constrained by path expressions. Identification variables designate instances of a particular entity abstract schema type. The FROM clause can contain multiple identification variable declarations separated by a comma (,).

  • from_clause ::= FROM identification_variable_declaration {, {identification_variable_declaration | collection_member_declaration}}*

  • identification_variable_declaration ::= range_variable_declaration { join | fetch_join }*

  • range_variable_declaration ::= abstract_schema_name [AS] identification_variable

  • join ::= join_spec join_association_path_expression [AS] identification_variable

  • fetch_join ::= join_spec FETCH join_association_path_expression

  • join_association_path_expression ::= join_collection_valued_path_expression | join_single_valued_association_path_expression

  • join_spec ::= [ LEFT [OUTER] | INNER ] JOIN

  • collection_member_declaration ::= IN (collection_valued_path_expression) [AS] identification_variable

2.3.1.  JPQL FROM Identifiers

An identifier is a character sequence of unlimited length. The character sequence must begin with a Java identifier start character, and all other characters must be Java identifier part characters. An identifier start character is any character for which the method Character.isJavaIdentifierStart returns true. This includes the underscore (_) character and the dollar sign ($) character. An identifier part character is any character for which the method Character.isJavaIdentifierPart returns true. The question mark (?) character is reserved for use by the Java Persistence query language. The following are reserved identifiers:

  • SELECT

  • FROM

  • WHERE

  • UPDATE

  • DELETE

  • JOIN

  • OUTER

  • INNER

  • LEFT

  • GROUP

  • BY

  • HAVING

  • FETCH

  • DISTINCT

  • OBJECT

  • NULL

  • TRUE

  • FALSE

  • NOT

  • AND

  • OR

  • BETWEEN

  • LIKE

  • IN

  • AS

  • UNKNOWN

  • EMPTY

  • MEMBER

  • OF

  • IS

  • AVG

  • MAX

  • MIN

  • SUM

  • COUNT

  • ORDER

  • BY

  • ASC

  • DESC

  • MOD

  • UPPER

  • LOWER

  • TRIM

  • POSITION

  • CHARACTER_LENGTH

  • CHAR_LENGTH

  • BIT_LENGTH

  • CURRENT_TIME

  • CURRENT_DATE

  • CURRENT_TIMESTAMP

  • NEW

  • EXISTS

  • ALL

  • ANY

  • SOME

Reserved identifiers are case insensitive. Reserved identifiers must not be used as identification variables. It is recommended that other SQL reserved words also not be as identification variables in queries because they may be used as reserved identifiers in future releases of the specification.

2.3.2.  JPQL Identification Variables

An identification variable is a valid identifier declared in the FROM clause of a query. All identification variables must be declared in the FROM clause. Identification variables cannot be declared in other clauses. An identification variable must not be a reserved identifier or have the same name as any entity in the same persistence unit: Identification variables are case insensitive. An identification variable evaluates to a value of the type of the expression used in declaring the variable. For example, consider the previous query:

SELECT DISTINCT mag FROM Magazine mag JOIN mag.articles art JOIN art.author auth WHERE auth.firstName = 'John'
In the FROM clause declaration mag.articlesart, the identification variable art evaluates to any Article value directly reachable from Magazine. The association-field articles is a collection of instances of the abstract schema type Article and the identification variable art refers to an element of this collection. The type of auth is the abstract schema type of Author. An identification variable ranges over the abstract schema type of an entity. An identification variable designates an instance of an entity abstract schema type or an element of a collection of entity abstract schema type instances. Identification variables are existentially quantified in a query. An identification variable always designates a reference to a single value. It is declared in one of three ways: in a range variable declaration, in a join clause, or in a collection member declaration. The identification variable declarations are evaluated from left to right in the FROM clause, and an identification variable declaration can use the result of a preceding identification variable declaration of the query string.

2.3.3.  JPQL Range Declarations

The syntax for declaring an identification variable as a range variable is similar to that of SQL; optionally, it uses the AS keyword.

  • range_variable_declaration ::= abstract_schema_name [AS] identification_variable

Range variable declarations allow the developer to designate a "root" for objects which may not be reachable by navigation. In order to select values by comparing more than one instance of an entity abstract schema type, more than one identification variable ranging over the abstract schema type is needed in the FROM clause.

The following query returns magazines whose price is greater than the price of magazines published by "Adventure" publishers. This example illustrates the use of two different identification variables in the FROM clause, both of the abstract schema type Magazine. The SELECT clause of this query determines that it is the magazines with prices greater than those of "Adventure" publisher's that are returned.

SELECT DISTINCT mag1 FROM Magazine mag1, Magazine mag2
WHERE mag1.price > mag2.price AND mag2.publisher.name = 'Adventure'

2.3.4.  JPQL Path Expressions

An identification variable followed by the navigation operator (.) and a state-field or association-field is a path expression. The type of the path expression is the type computed as the result of navigation; that is, the type of the state-field or association-field to which the expression navigates. Depending on navigability, a path expression that leads to a association-field may be further composed. Path expressions can be composed from other path expressions if the original path expression evaluates to a single-valued type (not a collection) corresponding to a association-field. Path expression navigability is composed using "inner join" semantics. That is, if the value of a non-terminal association-field in the path expression is null, the path is considered to have no value, and does not participate in the determination of the result. The syntax for single-valued path expressions and collection valued path expressions is as follows:

  • single_valued_path_expression ::= state_field_path_expression | single_valued_association_path_expression

  • state_field_path_expression ::= {identification_variable | single_valued_association_path_expression}.state_field

  • single_valued_association_path_expression ::= identification_variable.{single_valued_association_field.}*single_valued_association_field

  • collection_valued_path_expression ::= identification_variable.{single_valued_association_field.}*collection_valued_association_field

  • state_field ::= {embedded_class_state_field.}*simple_state_field

A single_valued_association_field is designated by the name of an association-field in a one-to-one or many-to-one relationship. The type of a single_valued_association_field and thus a single_valued_association_path_expression is the abstract schema type of the related entity. A collection_valued_association_field is designated by the name of an association-field in a one-to-many or a many-to-many relationship. The type of a collection_valued_association_field is a collection of values of the abstract schema type of the related entity. An embedded_class_state _field is designated by the name of an entity state field that corresponds to an embedded class. Navigation to a related entity results in a value of the related entity's abstract schema type.

The evaluation of a path expression terminating in a state-field results in the abstract schema type corresponding to the Java type designated by the state-field. It is syntactically illegal to compose a path expression from a path expression that evaluates to a collection. For example, if mag designates Magazine, the path expression mag.articles.author is illegal since navigation to authors results in a collection. This case should produce an error when the query string is verified. To handle such a navigation, an identification variable must be declared in the FROM clause to range over the elements of the articles collection. Another path expression must be used to navigate over each such element in the WHERE clause of the query, as in the following query which returns all authors that have any articles in any magazines:

SELECT DISTINCT art.author FROM Magazine AS mag, IN(mag.articles) art

2.3.5.  JPQL Joins

An inner join may be implicitly specified by the use of a cartesian product in the FROM clause and a join condition in the WHERE clause.

The syntax for explicit join operations is as follows:

  • join ::= join_spec join_association_path_expression [AS] identification_variable

  • fetch_join ::= join_spec FETCH join_association_path_expression

  • join_association_path_expression ::= join_collection_valued_path_expression | join_single_valued_association_path_expression

  • join_spec ::= [ LEFT [OUTER] | INNER ] JOIN

The following inner and outer join operation types are supported.

2.3.5.1.  JPQL Inner Joins (Relationship Joins)

The syntax for the inner join operation is

[ INNER ] JOIN join_association_path_expression [AS] identification_variable
For example, the query below joins over the relationship between publishers and magazines. This type of join typically equates to a join over a foreign key relationship in the database.

SELECT pub FROM Publisher pub JOIN pub.magazines mag WHERE pub.revenue > 1000000

The keyword INNER may optionally be used:

SELECT pub FROM Publisher pub INNER JOIN pub.magazines mag WHERE pub.revenue > 1000000

This is equivalent to the following query using the earlier IN construct. It selects those publishers with revenue of over 1 million for which at least one magazine exists:

SELECT OBJECT(pub) FROM Publisher pub, IN(pub.magazines) mag WHERE pub.revenue > 1000000
2.3.5.2.  JPQL Outer Joins

LEFT JOIN and LEFT OUTER JOIN are synonymous. They enable the retrieval of a set of entities where matching values in the join condition may be absent. The syntax for a left outer join is:

LEFT [OUTER] JOIN join_association_path_expression [AS] identification_variable

For example:

SELECT pub FROM Publisher pub LEFT JOIN pub.magazines mag WHERE pub.revenue > 1000000
The keyword OUTER may optionally be used:
SELECT pub FROM Publisher pub LEFT OUTER JOIN pub.magazines mags WHERE pub.revenue > 1000000
An important use case for LEFT JOIN is in enabling the prefetching of related data items as a side effect of a query. This is accomplished by specifying the LEFT JOIN as a FETCH JOIN.

2.3.5.3.  JPQL Fetch Joins

A FETCH JOIN enables the fetching of an association as a side effect of the execution of a query. A FETCH JOIN is specified over an entity and its related entities. The syntax for a fetch join is

  • fetch_join ::= [ LEFT [OUTER] | INNER ] JOIN FETCH join_association_path_expression

The association referenced by the right side of the FETCH JOIN clause must be an association that belongs to an entity that is returned as a result of the query. It is not permitted to specify an identification variable for the entities referenced by the right side of the FETCH JOIN clause, and hence references to the implicitly fetched entities cannot appear elsewhere in the query. The following query returns a set of magazines. As a side effect, the associated articles for those magazines are also retrieved, even though they are not part of the explicit query result. The persistent fields or properties of the articles that are eagerly fetched are fully initialized. The initialization of the relationship properties of the articles that are retrieved is determined by the metadata for the Article entity class.

SELECT mag FROM Magazine mag LEFT JOIN FETCH mag.articles WHERE mag.id = 1

A fetch join has the same join semantics as the corresponding inner or outer join, except that the related objects specified on the right-hand side of the join operation are not returned in the query result or otherwise referenced in the query. Hence, for example, if magazine id 1 has five articles, the above query returns five references to the magazine 1 entity.

2.3.6.  JPQL Collection Member Declarations

An identification variable declared by a collection_member_declaration ranges over values of a collection obtained by navigation using a path expression. Such a path expression represents a navigation involving the association-fields of an entity abstract schema type. Because a path expression can be based on another path expression, the navigation can use the association-fields of related entities. An identification variable of a collection member declaration is declared using a special operator, the reserved identifier IN . The argument to the IN operator is a collection-valued path expression. The path expression evaluates to a collection type specified as a result of navigation to a collection-valued association-field of an entity abstract schema type. The syntax for declaring a collection member identification variable is as follows:

  • collection_member_declaration ::= IN (collection_valued_path_expression) [AS] identification_variable

For example, the query

SELECT DISTINCT mag FROM Magazine mag
    JOIN mag.articles art
    JOIN art.author auth
    WHERE auth.lastName = 'Grisham'
may equivalently be expressed as follows, using the IN operator:
SELECT DISTINCT mag FROM Magazine mag,
    IN(mag.articles) art
    WHERE art.author.lastName = 'Grisham'
In this example, articles is the name of an association-field whose value is a collection of instances of the abstract schema type Article. The identification variable art designates a member of this collection, a single Article abstract schema type instance. In this example, mag is an identification variable of the abstract schema type Magazine.

2.3.7.  JPQL Polymorphism

Java Persistence queries are automatically polymorphic. The FROM clause of a query designates not only instances of the specific entity classes to which explicitly refers but of subclasses as well. The instances returned by a query include instances of the subclasses that satisfy the query criteria.

2.4.  JPQL WHERE Clause

The WHERE clause of a query consists of a conditional expression used to select objects or values that satisfy the expression. The WHERE clause restricts the result of a select statement or the scope of an update or delete operation. A WHERE clause is defined as follows:

  • where_clause ::= WHERE conditional_expression

The GROUP BY construct enables the aggregation of values according to the properties of an entity class. The HAVING construct enables conditions to be specified that further restrict the query result as restrictions upon the groups. The syntax of the HAVING clause is as follows:

  • having_clause ::= HAVING conditional_expression

The GROUP BY and HAVING constructs are further discussed in Section 2.6, “ JPQL GROUP BY, HAVING ”.

2.5.  JPQL Conditional Expressions

The following sections describe the language constructs that can be used in a conditional expression of the WHERE clause or HAVING clause. State-fields that are mapped in serialized form or as lobs may not be portably used in conditional expressions.

Note

The implementation is not expected to perform such query operations involving such fields in memory rather than in the database.

2.5.1.  JPQL Literals

A string literal is enclosed in single quotes--for example: 'literal'. A string literal that includes a single quote is represented by two single quotes--for example: 'literal''s'. String literals in queries, like Java String literals, use unicode character encoding. The use of Java escape notation is not supported in query string literals Exact numeric literals support the use of Java integer literal syntax as well as SQL exact numeric literal syntax. Approximate literals support the use Java floating point literal syntax as well as SQL approximate numeric literal syntax. Enum literals support the use of Java enum literal syntax. The enum class name must be specified. Appropriate suffixes may be used to indicate the specific type of a numeric literal in accordance with the Java Language Specification. The boolean literals are TRUE and FALSE. Although predefined reserved literals appear in upper case, they are case insensitive.

2.5.2.  JPQL Identification Variables

All identification variables used in the WHERE or HAVING clause of a SELECT or DELETE statement must be declared in the FROM clause, as described in Section 2.3.2, “ JPQL Identification Variables ”. The identification variables used in the WHERE clause of an UPDATE statement must be declared in the UPDATE clause. Identification variables are existentially quantified in the WHERE and HAVING clause. This means that an identification variable represents a member of a collection or an instance of an entity's abstract schema type. An identification variable never designates a collection in its entirety.

2.5.3.  JPQL Path Expressions

It is illegal to use a collection_valued_path_expression within a WHERE or HAVING clause as part of a conditional expression except in an empty_collection_comparison_expression, in a collection_member_expression, or as an argument to the SIZE operator.

2.5.4.  JPQL Input Parameters

Either positional or named parameters may be used. Positional and named parameters may not be mixed in a single query. Input parameters can only be used in the WHERE clause or HAVING clause of a query.

Note that if an input parameter value is null, comparison operations or arithmetic operations involving the input parameter will return an unknown value. See Section 2.10, “ JPQL Null Values ”.

2.5.4.1.  JPQL Positional Parameters

The following rules apply to positional parameters.

  • Input parameters are designated by the question mark (?) prefix followed by an integer. For example: ?1.

  • Input parameters are numbered starting from 1. Note that the same parameter can be used more than once in the query string and that the ordering of the use of parameters within the query string need not conform to the order of the positional parameters.

2.5.4.2.  JPQL Named Parameters

A named parameter is an identifier that is prefixed by the ":" symbol. It follows the rules for identifiers defined in Section 2.3.1, “ JPQL FROM Identifiers ”. Named parameters are case sensitive.

Example:

SELECT pub FROM Publisher pub WHERE pub.revenue > :rev

2.5.5.  JPQL Conditional Expression Composition

Conditional expressions are composed of other conditional expressions, comparison operations, logical operations, path expressions that evaluate to boolean values, boolean literals, and boolean input parameters. Arithmetic expressions can be used in comparison expressions. Arithmetic expressions are composed of other arithmetic expressions, arithmetic operations, path expressions that evaluate to numeric values, numeric literals, and numeric input parameters. Arithmetic operations use numeric promotion. Standard bracketing () for ordering expression evaluation is supported. Conditional expressions are defined as follows:

  • conditional_expression ::= conditional_term | conditional_expression OR conditional_term

  • conditional_term ::= conditional_factor | conditional_term AND conditional_factor

  • conditional_factor ::= [ NOT ] conditional_primary

  • conditional_primary ::= simple_cond_expression | (conditional_expression)

  • simple_cond_expression ::= comparison_expression | between_expression | like_expression | in_expression | null_comparison_expression | empty_collection_comparison_expression | collection_member_expression | exists_expression

Aggregate functions can only be used in conditional expressions in a HAVING clause. See Section 2.6, “ JPQL GROUP BY, HAVING ”.

2.5.6.  JPQL Operators and Operator Precedence

The operators are listed below in order of decreasing precedence.

  • Navigation operator (.)

  • Arithmetic operators: +, - unary *, / multiplication and division +, - addition and subtraction

  • Comparison operators: =, >, >=, <, <=, <> (not equal), [ NOT ] BETWEEN, [ NOT ] LIKE, [ NOT ] IN, IS [ NOT ] NULL, IS [ NOT ] EMPTY, [ NOT ] MEMBER [ OF ]

  • Logical operators: NOTANDOR

The following sections describe other operators used in specific expressions.

2.5.7.  JPQL Between Expressions

The syntax for the use of the comparison operator [ NOT ] BETWEEN in a conditional expression is as follows:

arithmetic_expression [NOT] BETWEEN arithmetic_expression AND arithmetic_expression | string_expression [NOT] BETWEEN string_expression AND string_expression | datetime_expression [NOT] BETWEEN datetime_expression AND datetime_expression

The BETWEEN expression

x BETWEEN y AND z
is semantically equivalent to:
y <= x AND x <= z
The rules for unknown and NULL values in comparison operations apply. See Section 2.10, “ JPQL Null Values ” . Examples are:
p.age BETWEEN 15 and 19
is equivalent to
p.age >= 15 AND p.age <= 19

p.age NOT BETWEEN 15 and 19
is equivalent to
p.age < 15 OR p.age > 19

2.5.8.  JPQL In Expressions

The syntax for the use of the comparison operator [ NOT ] IN in a conditional expression is as follows:

  • in_expression ::= state_field_path_expression [NOT] IN ( in_item {, in_item}* | subquery)

  • in_item ::= literal | input_parameter

The state_field_path_expression must have a string, numeric, or enum value. The literal and/or input_parameter values must be like the same abstract schema type of the state_field_path_expression in type. (See Section 2.11, “ JPQL Equality and Comparison Semantics ” ).

The results of the subquery must be like the same abstract schema type of the state_field_path_expression in type. Subqueries are discussed in Section 2.5.15, “ JPQL Subqueries ”. Examples are:

o.country IN ('UK', 'US', 'France')
is true for UK and false for Peru, and is equivalent to the expression:
(o.country = 'UK') OR (o.country = 'US') OR (o.country = ' France')
In the following expression:
o.country NOT IN ('UK', 'US', 'France')
is false for UK and true for Peru, and is equivalent to the expression:
NOT ((o.country = 'UK') OR (o.country = 'US') OR (o.country = 'France'))
There must be at least one element in the comma separated list that defines the set of values for the IN expression. If the value of a state_field_path_expression in an IN or NOT IN expression is NULL or unknown, the value of the expression is unknown.

2.5.9.  JPQL Like Expressions

The syntax for the use of the comparison operator [ NOT ] LIKE in a conditional expression is as follows:

string_expression [NOT] LIKE pattern_value [ESCAPE escape_character]

The string_expression must have a string value. The pattern_value is a string literal or a string-valued input parameter in which an underscore (_) stands for any single character, a percent (%) character stands for any sequence of characters (including the empty sequence), and all other characters stand for themselves. The optional escape_character is a single-character string literal or a character-valued input parameter (i.e., char or Character) and is used to escape the special meaning of the underscore and percent characters in pattern_value. Examples are:

  • address.phone LIKE '12%3'
    
    is true for '123' '12993' and false for '1234'

  • asentence.word LIKE 'l_se'
    is true for 'lose' and false for 'loose'

  • aword.underscored LIKE '\_%' ESCAPE '\'
    is true for '_foo' and false for 'bar'

  • address.phone NOT LIKE '12%3'
    is false for '123' and '12993' and true for '1234' If the value of the string_expression or pattern_value is NULL or unknown, the value of the LIKE expression is unknown. If the escape_character is specified and is NULL, the value of the LIKE expression is unknown.

2.5.10.  JPQL Null Comparison Expressions

The syntax for the use of the comparison operator IS NULL in a conditional expression is as follows:

{single_valued_path_expression | input_parameter } IS [NOT] NULL

A null comparison expression tests whether or not the single-valued path expression or input parameter is a NULL value.

2.5.11.  JPQL Empty Collection Comparison Expressions

The syntax for the use of the comparison operator IS EMPTY in an empty_collection_comparison_expression is as follows:

collection_valued_path_expression IS [NOT] EMPTY

This expression tests whether or not the collection designated by the collection-valued path expression is empty (i.e, has no elements).

For example, the following query will return all magazines that don't have any articles at all:

SELECT mag FROM Magazine mag WHERE mag.articles IS EMPTY
If the value of the collection-valued path expression in an empty collection comparison expression is unknown, the value of the empty comparison expression is unknown.

2.5.12.  JPQL Collection Member Expressions

The use of the comparison collection_member_expression is as follows: syntax for the operator MEMBER OF in an

  • collection_member_expression ::= entity_expression [NOT] MEMBER [OF] collection_valued_path_expression

  • entity_expression ::= single_valued_association_path_expression | simple_entity_expression

  • simple_entity_expression ::= identification_variable | input_parameter

This expression tests whether the designated value is a member of the collection specified by the collection-valued path expression. If the collection valued path expression designates an empty collection, the value of the MEMBER OF expression is FALSE and the value of the NOT MEMBER OF expression is TRUE. Otherwise, if the value of the collection-valued path expression or single-valued association-field path expression in the collection member expression is NULL or unknown, the value of the collection member expression is unknown.

2.5.13.  JPQL Exists Expressions

An EXISTS expression is a predicate that is true only if the result of the subquery consists of one or more values and that is false otherwise. The syntax of an exists expression is

  • exists_expression ::= [NOT] EXISTS (subquery)

The use of the reserved word OF is optional in this expression.

Example:

SELECT DISTINCT auth FROM Author auth
    WHERE EXISTS
        (SELECT spouseAuthor FROM Author spouseAuthor WHERE spouseAuthor = auth.spouse)
The result of this query consists of all authors whose spouse is also an author.

2.5.14.  JPQL All or Any Expressions

An ALL conditional expression is a predicate that is true if the comparison operation is true for all values in the result of the subquery or the result of the subquery is empty. An ALL conditional expression is false if the result of the comparison is false for at least one row, and is unknown if neither true nor false. An ANY conditional expression is a predicate that is true if the comparison operation is true for some value in the result of the subquery. An ANY conditional expression is false if the result of the subquery is empty or if the comparison operation is false for every value in the result of the subquery, and is unknown if neither true nor false. The keyword SOME is synonymous with ANY. The comparison operators used with ALL or ANY conditional expressions are =, <, <=, >, >=, <>. The result of the subquery must be like that of the other argument to the comparison operator in type. See Section 2.11, “ JPQL Equality and Comparison Semantics ”. The syntax of an ALL or ANY expression is specified as follows:

  • all_or_any_expression ::= { ALL | ANY | SOME} (subquery)

The following example select the authors who make the highest salary for their magazine:

SELECT auth FROM Author auth
    WHERE auth.salary >= ALL(SELECT a.salary FROM Author a WHERE a.magazine = auth.magazine)

2.5.15.  JPQL Subqueries

Subqueries may be used in the WHERE or HAVING clause. The syntax for subqueries is as follows:

  • subquery ::= simple_select_clause subquery_from_clause [where_clause] [groupby_clause] [having_clause]

Subqueries are restricted to the WHERE and HAVING clauses in this release. Support for subqueries in the FROM clause will be considered in a later release of the specification.

  • simple_select_clause ::= SELECT [DISTINCT] simple_select_expression

  • subquery_from_clause ::= FROM subselect_identification_variable_declaration {, subselect_identification_variable_declaration}*

  • subselect_identification_variable_declaration ::= identification_variable_declaration | association_path_expression [AS] identification_variable | collection_member_declaration

  • simple_select_expression ::= single_valued_path_expression | aggregate_expression | identification_variable

Examples:

SELECT DISTINCT auth FROM Author auth
    WHERE EXISTS (SELECT spouseAuth FROM Author spouseAuth WHERE spouseAuth = auth.spouse)
SELECT mag FROM Magazine mag
    WHERE (SELECT COUNT(art) FROM mag.articles art) > 10
Note that some contexts in which a subquery can be used require that the subquery be a scalar subquery (i.e., produce a single result). This is illustrated in the following example involving a numeric comparison operation.
SELECT goodPublisher FROM Publisher goodPublisher
    WHERE goodPublisher.revenue < (SELECT AVG(p.revenue) FROM Publisher p)

2.5.16.  JPQL Functional Expressions

The JPQL includes the following built-in functions, which may be used in the WHERE or HAVING clause of a query. If the value of any argument to a functional expression is null or unknown, the value of the functional expression is unknown.

2.5.16.1.  JPQL String Functions

  • functions_returning_strings ::= CONCAT(string_primar y, string_primary) | SUBSTRING(string_primar y, simple_arithmetic_expression, simple_arithmetic_expression) | TRIM([[trim_specification] [trim_character] FROM] string_primary) | LOWER(string_primar y) | UPPER(string_primar y)

  • trim_specification ::= LEADING | TRAILING | BOTH

  • functions_returning_numerics ::= LENGTH(string_primar y) | LOCATE(string_primar y, string_primar y[, simple_arithmetic_expression])

The CONCAT function returns a string that is a concatenation of its arguments. The second and third arguments of the SUBSTRING function denote the starting position and length of the substring to be returned. These arguments are integers. The first position of a string is denoted by 1. The SUBSTRING function returns a string. The TRIM function trims the specified character from a string. If the character to be trimmed is not specified, it is assumed to be space (or blank). The optional trim_character is a single-character string literal or a character-valued input parameter (i.e., char or Character). If a trim specification is not provided, BOTH is assumed. The TRIM function returns the trimmed string. The LOWER and UPPER functions convert a string to lower and upper case, respectively. They return a string. The LOCATE function returns the position of a given string within a string, starting the search at a specified position. It returns the first position at which the string was found as an integer. The first argument is the string to be located; the second argument is the string to be searched; the optional third argument is an integer that represents the string position at which the search is started (by default, the beginning of the string to be searched). The first position in a string is denoted by 1. If the string is not found, 0 is returned. The LENGTH function returns the length of the string in characters as an integer.

2.5.16.2.  JPQL Arithmetic Functions

  • functions_returning_numerics ::= ABS(simple_arithmetic_expression) | SQRT(simple_arithmetic_expression) | MOD(simple_arithmetic_expression, simple_arithmetic_expression) | SIZE(collection_valued_path_expression)

The ABS function takes a numeric argument and returns a number (integer, float, or double) of the same type as the argument to the function. The SQRT function takes a numeric argument and returns a double.

Note that not all databases support the use of a trim character other than the space character; use of this argument may result in queries that are not portable. Note that not all databases support the use of the third argument to LOCATE; use of this argument may result in queries that are not portable.

The MOD function takes two integer arguments and returns an integer. The SIZE function returns an integer value, the number of elements of the collection. If the collection is empty, the SIZE function evaluates to zero. Numeric arguments to these functions may correspond to the numeric Java object types as well as the primitive numeric types.

2.5.16.3.  JPQL Datetime Functions

functions_returning_datetime:= CURRENT_DATE | CURRENT_TIME | CURRENT_TIMESTAMP

The datetime functions return the value of current date, time, and timestamp on the database server.

2.6.  JPQL GROUP BY, HAVING

The GROUP BY construct enables the aggregation of values according to a set of properties. The HAVING construct enables conditions to be specified that further restrict the query result. Such conditions are restrictions upon the groups. The syntax of the GROUP BY and HAVING clauses is as follows:

  • groupby_clause ::= GROUP BY groupby_item {, groupby_item}*

  • groupby_item ::= single_valued_path_expression | identification_variable

  • having_clause ::= HAVING conditional_expression

If a query contains both a WHERE clause and a GROUP BY clause, the effect is that of first applying the where clause, and then forming the groups and filtering them according to the HAVING clause. The HAVING clause causes those groups to be retained that satisfy the condition of the HAVING clause. The requirements for the SELECT clause when GROUP BY is used follow those of SQL: namely, any item that appears in the SELECT clause (other than as an argument to an aggregate function) must also appear in the GROUP BY clause. In forming the groups, null values are treated as the same for grouping purposes. Grouping by an entity is permitted. In this case, the entity must contain no serialized state fields or lob-valued state fields. The HAVING clause must specify search conditions over the grouping items or aggregate functions that apply to grouping items.

If there is no GROUP BY clause and the HAVING clause is used, the result is treated as a single group, and the select list can only consist of aggregate functions. When a query declares a HAVING clause, it must always also declare a GROUP BY clause.

2.7.  JPQL SELECT Clause

The SELECT clause denotes the query result. More than one value may be returned from the SELECT clause of a query. The SELECT clause may contain one or more of the following elements: a single range variable or identification variable that ranges over an entity abstract schema type, a single-valued path expression, an aggregate select expression, a constructor expression. The SELECT clause has the following syntax:

  • select_clause ::= SELECT [DISTINCT] select_expression {, select_expression}*

  • select_expression ::= single_valued_path_expression | aggregate_expression | identification_variable | OBJECT(identification_variable) | constructor_expression

  • constructor_expression ::= NEW constructor_name ( constructor_item {, constructor_item}*)

  • constructor_item ::= single_valued_path_expression | aggregate_expression

  • aggregate_expression ::= { AVG | MAX | MIN | SUM } ([DISTINCT] state_field_path_expression) | COUNT ([DISTINCT] identification_variable | state_field_path_expression | single_valued_association_path_expression)

For example:

SELECT pub.id, pub.revenue
    FROM Publisher pub JOIN pub.magazines mag WHERE mag.price > 5.00

Note that the SELECT clause must be specified to return only single-valued expressions. The query below is therefore not valid:

SELECT mag.authors FROM Magazine AS mag
The DISTINCT keyword is used to specify that duplicate values must be eliminated from the query result. If DISTINCT is not specified, duplicate values are not eliminated. Standalone identification variables in the SELECT clause may optionally be qualified by the OBJECT operator. The SELECT clause must not use the OBJECT operator to qualify path expressions.

2.7.1.  JPQL Result Type of the SELECT Clause

The type of the query result specified by the SELECT clause of a query is an entity abstract schema type, a state-field type, the result of an aggregate function, the result of a construction operation, or some sequence of these. The result type of the SELECT clause is defined by the the result types of the select_expressions contained in it. When multiple select_expressions are used in the SELECT clause, the result of the query is of type Object[], and the elements in this result correspond in order to the order of their specification in the SELECT clause and in type to the result types of each of the select_expressions. The type of the result of a select_expression is as follows:

  • A single_valued_path_expression that is a state_field_path_expression results in an object of the same type as the corresponding state field of the entity. If the state field of the entity is a primitive type, the corresponding object type is returned.

  • single_valued_path_expression that is a single_valued_association_path_expression results in an entity object of the type of the relationship field or the subtype of the relationship field of the entity object as determined by the object/relational mapping.

  • The result type of an identification_variable is the type of the entity to which that identification variable corresponds or a subtype as determined by the object/relational mapping.

  • The result type of aggregate_expression is defined in section Section 2.7.4, “ JPQL Aggregate Functions ”.

  • The result type of a constructor_expression is the type of the class for which the constructor is defined. The types of the arguments to the constructor are defined by the above rules.

2.7.2.  JPQL Constructor Expressions

in the SELECT Clause A constructor may be used in the SELECT list to return one or more Java instances. The specified class is not required to be an entity or to be mapped to the database. The constructor name must be fully qualified.

If an entity class name is specified in the SELECT NEW clause, the resulting entity instances are in the new state.

SELECT NEW com.company.PublisherInfo(pub.id, pub.revenue, mag.price)
    FROM Publisher pub JOIN pub.magazines mag WHERE mag.price > 5.00

2.7.3.  JPQL Null Values in the Query Result

If the result of a query corresponds to a association-field or state-field whose value is null, that null value is returned in the result of the query method. The IS NOT NULL construct can be used to eliminate such null values from the result set of the query. Note, however, that state-field types defined in terms of Java numeric primitive types cannot produce NULL values in the query result. A query that returns such a state-field type as a result type must not return a null value.

2.7.4.  JPQL Aggregate Functions

in the SELECT Clause The result of a query may be the result of an aggregate function applied to a path expression. The following aggregate functions can be used in the SELECT clause of a query: AVG, COUNT, MAX, MIN, SUM. For all aggregate functions except COUNT, the path expression that is the argument to the aggregate function must terminate in a state-field. The path expression argument to COUNT may terminate in either a state-field or a association-field, or the argument to COUNT may be an identification variable. Arguments to the functions SUM and AVG must be numeric. Arguments to the functions MAX and MIN must correspond to orderable state-field types (i.e., numeric types, string types, character types, or date types). The Java type that is contained in the result of a query using an aggregate function is as follows:

  • COUNT returns Long.

  • MAX, MIN return the type of the state-field to which they are applied.

  • AVG returns Double.

  • SUM returns Long when applied to state-fields of integral types (other than BigInteger); Double when applied to state-fields of floating point types; BigInteger when applied to state-fields of type BigInteger; and BigDecimal when applied to state-fields of type BigDecimal. If SUM , AVG, MAX, or MIN is used, and there are no values to which the aggregate function can be applied, the result of the aggregate function is NULL. If COUNT is used, and there are no values to which COUNT can be applied, the result of the aggregate function is 0.

The argument to an aggregate function may be preceded by the keyword DISTINCT to specify that duplicate values are to be eliminated before the aggregate function is applied. Null values are eliminated before the aggregate function is applied, regardless of whether the keyword DISTINCT is specified.

2.7.4.1.  JPQL Aggregate Examples

Examples The following query returns the average price of all magazines:

SELECT AVG(mag.price) FROM Magazine mag
The following query returns the sum total cost of all the prices from all the magazines published by 'Larry':
SELECT SUM(mag.price) FROM Publisher pub JOIN pub.magazines mag pub.firstName = 'Larry'
The following query returns the total number of magazines:
SELECT COUNT(mag) FROM Magazine mag

2.8.  JPQL ORDER BY Clause

The ORDER BY clause allows the objects or values that are returned by the query to be ordered. The syntax of the ORDER BY clause is

  • orderby_clause ::= ORDER BY orderby_item {, orderby_item}*

  • orderby_item ::= state_field_path_expression [ASC | DESC]

It is legal to specify DISTINCT with MAX or MIN, but it does not affect the result.

When the ORDER BY clause is used in a query, each element of the SELECT clause of the query must be one of the following: an identification variable x, optionally denoted as OBJECT(x) , a single_valued_association_path_expression, or a state_field_path_expression. For example:

SELECT pub FROM Publisher pub JOIN pub.magazines mag ORDER BY o.revenue, o.name
If more than one orderby_item is specified, the left-to-right sequence of the orderby_item elements determines the precedence, whereby the leftmost orderby_item has highest precedence. The keyword ASC specifies that ascending ordering be used; the keyword DESC specifies that descending ordering be used. Ascending ordering is the default. SQL rules for the ordering of null values apply: that is, all null values must appear before all non-null values in the ordering or all null values must appear after all non-null values in the ordering, but it is not specified which. The ordering of the query result is preserved in the result of the query method if the ORDER BY clause is used.

2.9.  JPQL Bulk Update and Delete

Operations Bulk update and delete operations apply to entities of a single entity class (together with its subclasses, if any). Only one entity abstract schema type may be specified in the FROM or UPDATE clause. The syntax of these operations is as follows:

  • update_statement ::= update_clause [where_clause]

  • update_clause ::= UPDATE abstract_schema_name [[AS] identification_variable] SET update_item {, update_item}*

  • update_item ::= [identification_variable.]{state_field | single_valued_association_field} = new_value

  • new_value ::= simple_arithmetic_expression | string_primary | datetime_primary | boolean_primary | enum_primary simple_entity_expression | NULL

  • delete_statement ::= delete_clause [where_clause]

  • delete_clause ::= DELETE FROM abstract_schema_name [[AS] identification_variable]

The syntax of the WHERE clause is described in Section 2.4, “ JPQL WHERE Clause ”. A delete operation only applies to entities of the specified class and its subclasses. It does not cascade to related entities. The new_value specified for an update operation must be compatible in type with the state-field to which it is assigned. Bulk update maps directly to a database update operation, bypassing optimistic locking checks. Portable applications must manually update the value of the version column, if desired, and/or manually validate the value of the version column. The persistence context is not synchronized with the result of the bulk update or delete. Caution should be used when executing bulk update or delete operations because they may result in inconsistencies between the database and the entities in the active persistence context. In general, bulk update and delete operations should only be performed within a separate transaction or at the beginning of a transaction (before entities have been accessed whose state might be affected by such operations).

Examples:

DELETE FROM Publisher pub WHERE pub.revenue > 1000000.0
DELETE FROM Publisher pub WHERE pub.revenue = 0 AND pub.magazines IS EMPTY
UPDATE Publisher pub SET pub.status = 'outstanding'
    WHERE pub.revenue < 1000000 AND 20 > (SELECT COUNT(mag) FROM pub.magazines mag)

2.10.  JPQL Null Values

When the target of a reference does not exist in the database, its value is regarded as NULL. SQL 92 NULL semantics defines the evaluation of conditional expressions containing NULL values. The following is a brief description of these semantics:

  • Comparison or arithmetic operations with a NULL value always yield an unknown value.

  • Two NULL values are not considered to be equal, the comparison yields an unknown value.

  • Comparison or arithmetic operations with an unknown value always yield an unknown value.

  • The IS NULL and IS NOT NULL operators convert a NULL state-field or single-valued association-field value into the respective TRUE or FALSE value.

Note: The JPQL defines the empty string, "", as a string with 0 length, which is not equal to a NULL value. However, NULL values and empty strings may not always be distinguished when queries are mapped to some databases. Application developers should therefore not rely on the semantics of query comparisons involving the empty string and NULL value.

2.11.  JPQL Equality and Comparison Semantics

Only the values of like types are permitted to be compared. A type is like another type if they correspond to the same Java language type, or if one is a primitive Java language type and the other is the wrappered Java class type equivalent (e.g., int and Integer are like types in this sense). There is one exception to this rule: it is valid to compare numeric values for which the rules of numeric promotion apply. Conditional expressions attempting to compare non-like type values are disallowed except for this numeric case. Note that the arithmetic operators and comparison operators are permitted to be applied to state-fields and input parameters of the wrappered Java class equivalents to the primitive numeric Java types. Two entities of the same abstract schema type are equal if and only if they have the same primary key value. Only equality/inequality comparisons over enums are required to be supported.

2.12.  JPQL BNF

The following is the BNF for the Java Persistence query language, from section 4.14 of the JSR 220 specification.

  • QL_statement ::= select_statement | update_statement | delete_statement

  • select_statement ::= select_clause from_clause [where_clause] [groupby_clause] [having_clause] [orderby_clause]

  • update_statement ::= update_clause [where_clause]

  • delete_statement ::= delete_clause [where_clause]

  • from_clause ::= FROM identification_variable_declaration {, {identification_variable_declaration | collection_member_declaration}}*

  • identification_variable_declaration ::= range_variable_declaration { join | fetch_join }*

  • range_variable_declaration ::= abstract_schema_name [ AS ] identification_variable

  • join ::= join_spec join_association_path_expression [ AS ] identification_variable

  • fetch_join ::= join_spec FETCH join_association_path_expression

  • association_path_expression ::= collection_valued_path_expression | single_valued_association_path_expression

  • join_spec ::= [ LEFT [ OUTER ]| INNER ] JOIN

  • join_association_path_expression ::= join_collection_valued_path_expression | join_single_valued_association_path_expression

  • join_collection_valued_path_expression ::= identification_variable.collection_valued_association_field

  • join_single_valued_association_path_expression ::= identification_variable.single_valued_association_field

  • collection_member_declaration ::= IN (collection_valued_path_expression) [ AS ] identification_variable

  • single_valued_path_expression ::= state_field_path_expression | single_valued_association_path_expression

  • state_field_path_expression ::= {identification_variable | single_valued_association_path_expression}.state_field

  • single_valued_association_path_expression ::= identification_variable.{single_valued_association_field.}* single_valued_association_field

  • collection_valued_path_expression ::= identification_variable.{single_valued_association_field.}*collection_valued_association_field

  • state_field ::= {embedded_class_state_field.}*simple_state_field

  • update_clause ::= UPDATE abstract_schema_name [[ AS ] identification_variable] SET update_item {, update_item}*

  • update_item ::= [identification_variable.]{state_field | single_valued_association_field}= new_value

  • new_value ::= simple_arithmetic_expression | string_primary | datetime_primary | boolean_primary | enum_primary simple_entity_expression | NULL

  • delete_clause ::= DELETEFROM abstract_schema_name [[ AS ] identification_variable]

  • select_clause ::= SELECT [ DISTINCT ] select_expression {, select_expression}*

  • select_expression ::= single_valued_path_expression | aggregate_expression | identification_variable | OBJECT (identification_variable)| constructor_expression

  • constructor_expression ::= NEW constructor_name( constructor_item {, constructor_item}*)

  • constructor_item ::= single_valued_path_expression | aggregate_expression

  • aggregate_expression ::= { AVG | MAX | MIN | SUM }([ DISTINCT ] state_field_path_expression) | COUNT ([ DISTINCT ] identification_variable | state_field_path_expression | single_valued_association_path_expression)

  • where_clause ::= WHERE conditional_expression

  • groupby_clause ::= GROUPBY groupby_item {, groupby_item}*

  • groupby_item ::= single_valued_path_expression | identification_variable

  • having_clause ::= HAVING conditional_expression

  • orderby_clause ::= ORDERBY orderby_item {, orderby_item}*

  • orderby_item ::= state_field_path_expression [ ASC | DESC ]

  • subquery ::= simple_select_clause subquery_from_clause [where_clause] [groupby_clause] [having_clause]

  • subquery_from_clause ::= FROM subselect_identification_variable_declaration {, subselect_identification_variable_declaration}*

  • subselect_identification_variable_declaration ::= identification_variable_declaration | association_path_expression [ AS ] identification_variable | collection_member_declaration

  • simple_select_clause ::= SELECT [ DISTINCT ] simple_select_expression

  • simple_select_expression ::= single_valued_path_expression | aggregate_expression | identification_variable

  • conditional_expression ::= conditional_term | conditional_expression OR conditional_term

  • conditional_term ::= conditional_factor | conditional_term AND conditional_factor

  • conditional_factor ::= [ NOT ] conditional_primary

  • conditional_primary ::= simple_cond_expression |(conditional_expression)

  • simple_cond_expression ::= comparison_expression | between_expression | like_expression | in_expression | null_comparison_expression | empty_collection_comparison_expression | collection_member_expression | exists_expression

  • between_expression ::= arithmetic_expression [ NOT ] BETWEEN arithmetic_expression AND arithmetic_expression | string_expression [ NOT ] BETWEEN string_expression AND string_expression | datetime_expression [ NOT ] BETWEEN datetime_expression AND datetime_expression

  • in_expression ::= state_field_path_expression [ NOT ] IN ( in_item {, in_item}* | subquery)

  • in_item ::= literal | input_parameter

  • like_expression ::= string_expression [ NOT ] LIKE pattern_value [ ESCAPE escape_character]

  • null_comparison_expression ::= {single_valued_path_expression | input_parameter} IS [ NOT ] NULL

  • empty_collection_comparison_expression ::= collection_valued_path_expression IS [ NOT ] EMPTY

  • collection_member_expression ::= entity_expression [ NOT ] MEMBER [ OF ] collection_valued_path_expression

  • exists_expression ::= [ NOT ] EXISTS (subquery)

  • all_or_any_expression ::= { ALL | ANY | SOME }(subquery)

  • comparison_expression ::= string_expressioncomparison_operator{string_expression|all_or_any_expression}| boolean_expression {=|<>} {boolean_expression | all_or_any_expression} | enum_expression {=|<>} {enum_expression | all_or_any_expression} | datetime_expression comparison_operator {datetime_expression | all_or_any_expression} | entity_expression {= |<> } {entity_expression | all_or_any_expression} | arithmetic_expression comparison_operator {arithmetic_expression | all_or_any_expression}

  • comparison_operator ::== |> |>= |< |<= |<>

  • arithmetic_expression ::= simple_arithmetic_expression |(subquery)

  • simple_arithmetic_expression ::= arithmetic_term | simple_arithmetic_expression {+ |- } arithmetic_term

  • arithmetic_term ::= arithmetic_factor | arithmetic_term {* |/ } arithmetic_factor

  • arithmetic_factor ::= [{+ |-}] arithmetic_primary

  • arithmetic_primary ::= state_field_path_expression | numeric_literal | (simple_arithmetic_expression) | input_parameter | functions_returning_numerics | aggregate_expression

  • string_expression ::= string_primary |(subquery)

  • string_primary ::= state_field_path_expression | string_literal | input_parameter | functions_returning_strings | aggregate_expression

  • datetime_expression ::= datetime_primary |(subquery)

  • datetime_primary ::= state_field_path_expression | input_parameter | functions_returning_datetime | aggregate_expression

  • boolean_expression ::= boolean_primary |(subquery)

  • boolean_primary ::= state_field_path_expression | boolean_literal | input_parameter |

  • enum_expression ::= enum_primary |(subquery)

  • enum_primary ::= state_field_path_expression | enum_literal | input_parameter |

  • entity_expression ::= single_valued_association_path_expression | simple_entity_expression

  • simple_entity_expression ::= identification_variable | input_parameter

  • functions_returning_numerics ::= LENGTH (string_primary)| LOCATE (string_primary,string_primary [, simple_arithmetic_expression]) | ABS (simple_arithmetic_expression) | SQRT (simple_arithmetic_expression) | MOD (simple_arithmetic_expression, simple_arithmetic_expression) | SIZE (collection_valued_path_expression)

  • functions_returning_datetime ::= CURRENT_DATE | CURRENT_TIME | CURRENT_TIMESTAMP

  • functions_returning_strings ::= CONCAT (string_primary, string_primary) | SUBSTRING (string_primary, simple_arithmetic_expression,simple_arithmetic_expression)| TRIM ([[trim_specification] [trim_character] FROM ] string_primary) | LOWER (string_primary) | UPPER (string_primary)

  • trim_specification ::= LEADING | TRAILING | BOTH

Chapter 11.  SQL Queries

JPQL is a powerful query language, but there are times when it is not enough. Maybe you're migrating a JDBC application to JPA on a strict deadline, and you don't have time to translate your existing SQL selects to JPQL. Or maybe a certain query requires database-specific SQL your JPA implementation doesn't support. Or maybe your DBA has spent hours crafting the perfect select statement for a query in your application's critical path. Whatever the reason, SQL queries can remain an essential part of an application.

You are probably familiar with executing SQL queries by obtaining a java.sql.Connection, using the JDBC APIs to create a Statement, and executing that Statement to obtain a ResultSet. And of course, you are free to continue using this low-level approach to SQL execution in your JPA applications. However, JPA also supports executing SQL queries through the javax.persistence.Query interface introduced in Chapter 10, JPA Query . Using a JPA SQL query, you can retrieve either persistent objects or projections of column values. The following sections detail each use.

1.  Creating SQL Queries

The EntityManager has two factory methods suitable for creating SQL queries:

public Query createNativeQuery(String sqlString, Class resultClass);
public Query createNativeQuery(String sqlString, String resultSetMapping);

The first method is used to create a new Query instance that will return instances of the specified class.

The second method uses a SqlResultSetMapping to determine the type of object or objects to return. The example below shows these methods in action.

Example 11.1.  Creating a SQL Query

EntityManager em = ...;
Query query = em.createNativeQuery("SELECT * FROM MAG", Magazine.class);
processMagazines(query.getResultList());

Note

In addition to SELECT statements, OpenJPA supports stored procedure invocations as SQL queries. OpenJPA will assume any SQL that does not begin with the SELECT keyword (ignoring case) is a stored procedure call, and invoke it as such at the JDBC level.

2.  Retrieving Persistent Objects with SQL

When you give a SQL Query a candidate class, it will return persistent instances of that class. At a minimum, your SQL must select the class' primary key columns, discriminator column (if mapped), and version column (also if mapped). The JPA runtime uses the values of the primary key columns to construct each result object's identity, and possibly to match it with a persistent object already in the EntityManager's cache. When an object is not already cached, the implementation creates a new object to represent the current result row. It might use the discriminator column value to make sure it constructs an object of the correct subclass. Finally, the query records available version column data for use in optimistic concurrency checking, should you later change the result object and flush it back to the database.

Aside from the primary key, discriminator, and version columns, any columns you select are used to populate the persistent fields of each result object. JPA implementations will compete on how effectively they map your selected data to your persistent instance fields.

Let's make the discussion above concrete with an example. It uses the following simple mapping between a class and the database:

Example 11.2.  Retrieving Persistent Objects

Query query = em.createNativeQuery("SELECT ISBN, TITLE, PRICE, "
    + "VERS FROM MAG WHERE PRICE > 5 AND PRICE < 10", Magazine.class);
List<Magazine> results = (List<Magazine>) query.getResultList();
for (Magazine mag : results)
    processMagazine(mag);

The query above works as advertised, but isn't very flexible. Let's update it to take in parameters for the minimum and maximum price, so we can reuse it to find magazines in any price range:

Example 11.3.  SQL Query Parameters

Query query = em.createNativeQuery("SELECT ISBN, TITLE, PRICE, "
    + "VERS FROM MAG WHERE PRICE > ?1 AND PRICE < ?2", Magazine.class);

query.setParameter(1, 5d);
query.setParameter(2, 10d);

List<Magazine> results = (List<Magazine>) query.getResultList();
for (Magazine mag : results)
    processMagazine (mag);

Like JDBC prepared statements, SQL queries represent parameters with question marks, but are followed by an integer to represent its index.

Chapter 12.  Mapping Metadata

Object-relational mapping is the process of mapping entities to relational database tables. In JPA, you perform object/relational mapping through mapping metadata. Mapping metadata uses annotations to describe how to link your object model to your relational model.

Note

OpenJPA offers tools to automate mapping and schema creation. See Chapter 7, Mapping in the Reference Guide.

Throughout this chapter, we will draw on the object model introduced in Chapter 5, Metadata . We present that model again below. As we discuss various aspects of mapping metadata, we will zoom in on specific areas of the model and show how we map the object layer to the relational layer.

All mapping metadata is optional. Where no explicit mapping metadata is given, JPA uses the defaults defined by the specification. As we present each mapping throughout this chapter, we also describe the defaults that apply when the mapping is absent.

1.  Table

The Table annotation specifies the table for an entity class. If you omit the Table annotation, base entity classes default to a table with their unqualified class name. The default table of an entity subclass depends on the inheritance strategy, as you will see in Section 6, “ Inheritance ”.

Tables have the following properties:

  • String name: The name of the table. Defaults to the unqualified entity class name.

  • String schema: The table's schema. If you do not name a schema, JPA uses the default schema for the database connection.

  • String catalog: The table's catalog. If you do not name a catalog, JPA uses the default catalog for the database connection.

  • UniqueConstraint[] uniqueConstraints: An array of unique constraints to place on the table. We cover unique constraints below. Defaults to an empty array.

The equivalent XML element is table. It has the following attributes, which correspond to the annotation properties above:

  • name

  • schema

  • catalog

The table element also accepts nested unique-constraint elements representing unique constraints. We will detail unique constraints shortly.

Sometimes, some of the fields in a class are mapped to secondary tables. In that case, use the class' Table annotation to name what you consider the class' primary table. Later, we will see how to map certain fields to other tables.

The example below maps classes to tables according to the following diagram. The CONTRACT, SUB, and LINE_ITEM tables are in the CNTRCT schema; all other tables are in the default schema.

Note that the diagram does not include our model's Document and Address classes. Mapped superclasses and embeddable classes are never mapped to tables.

Example 12.1.  Mapping Classes

package org.mag;

@Entity
@IdClass(Magazine.MagazineId.class)
@Table(name="MAG")
public class Magazine {
    ...

    public static class MagazineId {
        ...
    }
}

@Entity
@Table(name="ART")
public class Article {
    ...
}


package org.mag.pub;

@Entity
@Table(name="COMP")
public class Company {
    ...
}

@Entity
@Table(name="AUTH")
public class Author {
    ...
}

@Embeddable
public class Address {
    ...
}


package org.mag.subscribe;

@MappedSuperclass
public abstract class Document {
    ...
}

@Entity
@Table(schema="CNTRCT")
public class Contract
    extends Document {
    ...
}

@Entity
@Table(name="SUB", schema="CNTRCT")
public class Subscription {
    ...

    @Entity
    @Table(name="LINE_ITEM", schema="CNTRCT")
    public static class LineItem
        extends Contract {
        ...
    }
}

@Entity(name="Lifetime")
public class LifetimeSubscription
    extends Subscription {
    ...
}

@Entity(name="Trial")
public class TrialSubscription
    extends Subscription {
    ...
}

The same mapping information expressed in XML:

<entity-mappings xmlns="http://java.sun.com/xml/ns/persistence/orm" 
    xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" 
    xsi:schemaLocation="http://java.sun.com/xml/ns/persistence/orm orm_1_0.xsd"
    version="1.0">
    <mapped-superclass class="org.mag.subscribe.Document">
        ...
    </mapped-superclass>
    <entity class="org.mag.Magazine">
        <table name="MAG"/>
        <id-class="org.mag.Magazine.MagazineId"/>
        ...
    </entity>
    <entity class="org.mag.Article">
        <table name="ART"/>
        ...
    </entity>
    <entity class="org.mag.pub.Company">
        <table name="COMP"/>
        ...
    </entity>
    <entity class="org.mag.pub.Author">
        <table name="AUTH"/>
        ...
    </entity>
    <entity class="org.mag.subcribe.Contract">
        <table schema="CNTRCT"/>
        ...
    </entity>
    <entity class="org.mag.subcribe.Subscription">
        <table name="SUB" schema="CNTRCT"/>
        ...
    </entity>
    <entity class="org.mag.subscribe.Subscription.LineItem">
        <table name="LINE_ITEM" schema="CNTRCT"/>
        ...
    </entity>
    <entity class="org.mag.subscribe.LifetimeSubscription" name="Lifetime">
        ...
    </entity>
    <entity class="org.mag.subscribe.TrialSubscription" name="Trial">
        ...
    </entity>
    <embeddable class="org.mag.pub.Address">
        ...
    </embeddable>
</entity-mappings>

2.  Unique Constraints

Unique constraints ensure that the data in a column or combination of columns is unique for each row. A table's primary key, for example, functions as an implicit unique constraint. In JPA, you represent other unique constraints with an array of UniqueConstraint annotations within the table annotation. The unique constraints you define are used during table creation to generate the proper database constraints, and may also be used at runtime to order INSERT, UPDATE , and DELETE statements. For example, suppose there is a unique constraint on the columns of field F. In the same transaction, you remove an object A and persist a new object B, both with the same F value. The JPA runtime must ensure that the SQL deleting A is sent to the database before the SQL inserting B to avoid a unique constraint violation.

UniqueConstraint has a single property:

  • String[] columnNames: The names of the columns the constraint spans.

In XML, unique constraints are represented by nesting unique-constraint elements within the table element. Each unique-constraint element in turn nests column-name text elements to enumerate the contraint's columns.

Example 12.2.  Defining a Unique Constraint

The following defines a unique constraint on the TITLE column of the ART table:

@Entity
@Table(name="ART", uniqueConstraints=@Unique(columnNames="TITLE"))
public class Article {
    ...
}

The same metadata expressed in XML form:

<entity class="org.mag.Article">
    <table name="ART">
        <unique-constraint>
            <column-name>TITLE</column-name>
        </unique-constraint>
    </table>
    ...
</entity>

3.  Column

In the previous section, we saw that a UniqueConstraint uses an array of column names. Field mappings, however, use full-fledged Column annotations. Column annotations have the following properties:

  • String name: The column name. Defaults to the field name.

  • String columnDefinition: The database-specific column type name. This property is only used by vendors that support creating tables from your mapping metadata. During table creation, the vendor will use the value of the columnDefinition as the declared column type. If no columnDefinition is given, the vendor will choose an appropriate default based on the field type combined with the column's length, precision, and scale.

  • int length: The column length. This property is typically only used during table creation, though some vendors might use it to validate data before flushing. CHAR and VARCHAR columns typically default to a length of 255; other column types use the database default.

  • int precision: The precision of a numeric column. This property is often used in conjunction with scale to form the proper column type name during table creation.

  • int scale: The number of decimal digits a numeric column can hold. This property is often used in conjunction with precision to form the proper column type name during table creation.

  • boolean nullable: Whether the column can store null values. Vendors may use this property both for table creation and at runtime; however, it is never required. Defaults to true.

  • boolean insertable: By setting this property to false, you can omit the column from SQL INSERT statements. Defaults to true.

  • boolean updatable: By setting this property to false, you can omit the column from SQL UPDATE statements. Defaults to true.

  • String table: Sometimes you will need to map fields to tables other than the primary table. This property allows you specify that the column resides in a secondary table. We will see how to map fields to secondary tables later in the chapter.

The equivalent XML element is column. This element has attributes that are exactly equivalent to the Column annotation's properties described above:

  • name

  • column-definition

  • length

  • precision

  • scale

  • insertable

  • updatable

  • table

4.  Identity Mapping

With our new knowledge of columns, we can map the identity fields of our entities. The diagram below now includes primary key columns for our model's tables. The primary key column for Author uses nonstandard type INTEGER64, and the Magazine.isbn field is mapped to a VARCHAR(9) column instead of a VARCHAR(255) column, which is the default for string fields. We do not need to point out either one of these oddities to the JPA implementation for runtime use. If, however, we want to use the JPA implementation to create our tables for us, it needs to know about any desired non-default column types. Therefore, the example following the diagram includes this data in its encoding of our mappings.

Note that many of our identity fields do not need to specify column information, because they use the default column name and type.

Example 12.3.  Identity Mapping

package org.mag;

@Entity
@IdClass(Magazine.MagazineId.class)
@Table(name="MAG")
public class Magazine {

    @Column(length=9)
    @Id private String isbn;
    @Id private String title;

    ...

    public static class MagazineId {
        ...
    }
}

@Entity
@Table(name="ART", uniqueConstraints=@Unique(columnNames="TITLE"))
public class Article {

    @Id private long id;

    ...
}


package org.mag.pub;

@Entity
@Table(name="COMP")
public class Company {

    @Column(name="CID")
    @Id private long id;

    ...
}

@Entity
@Table(name="AUTH")
public class Author {

    @Column(name="AID", columnDefinition="INTEGER64")
    @Id private long id;

    ...
}

@Embeddable
public class Address {
    ...
}


package org.mag.subscribe;

@MappedSuperclass
public abstract class Document {

    @Id
    @GeneratedValue(strategy=GenerationType.IDENTITY)
    private long id;

    ...
}

@Entity
@Table(schema="CNTRCT")
public class Contract
    extends Document {
    ...
}

@Entity
@Table(name="SUB", schema="CNTRCT")
public class Subscription {

    @Id private long id;

    ...

    @Entity
    @Table(name="LINE_ITEM", schema="CNTRCT")
    public static class LineItem
        extends Contract {
        ...
    }
}

@Entity(name="Lifetime")
public class LifetimeSubscription
    extends Subscription {
    ...
}

@Entity(name="Trial")
public class TrialSubscription
    extends Subscription {
    ...
}

The same metadata for Magazine and Company expressed in XML form:

<entity class="org.mag.Magazine">
    <id-class class="org.mag.Magazine.Magazine.MagazineId"/>
    <table name="MAG"/>
    <attributes>
        <id name="isbn">
            <column length="9"/>
        </id>
        <id name="title"/>
      ...  
    </attributes>
</entity>
<entity class="org.mag.pub.Company">
    <table name="COMP"/>
    <attributes>
        <id name="id">
            <column name="CID"/>
        </id>
        ...
    </attributes>
</entity>

5.  Generators

One aspect of identity mapping not covered in the previous section is JPA's ability to automatically assign a value to your numeric identity fields using generators. We discussed the available generator types in Section 2.2, “ Id ”. Now we show you how to define named generators.

5.1.  Sequence Generator

Most databases allow you to create native sequences. These are database structures that generate increasing numeric values. The SequenceGenerator annotation represents a named database sequence. You can place the annotation on any package, entity class, persistent field declaration (if your entity uses field access), or getter method for a persistent property (if your entity uses property access). SequenceGenerator has the following properties:

  • String name: The generator name. This property is required.

  • String sequenceName: The name of the database sequence. If you do not specify the database sequence, your vendor will choose an appropriate default.

  • int initialValue: The initial sequence value.

  • int allocationSize: Some databases can pre-allocate groups of sequence values. This allows the database to service sequence requests from cache, rather than physically incrementing the sequence with every request. This allocation size defaults to 50.

Note

OpenJPA allows you to use describe one of OpenJPA's built-in generator implementations in the sequenceName property. You can also set the sequenceName to system to use the system sequence defined by the openjpa.Sequence configuration property. See the Reference Guide's Section 6, “ Generators ” for details.

The XML element for a sequence generator is sequence-generator . Its attributes mirror the above annotation's properties:

  • name

  • sequence-name

  • initial-value

  • allocation-size

To use a sequence generator, set your GeneratedValue annotation's strategy property to GenerationType.SEQUENCE, and its generator property to the sequence generator's declared name. Or equivalently, set your generated-value XML element's strategy attribute to SEQUENCE and its generator attribute to the generator name.

5.2.  TableGenerator

A TableGenerator refers to a database table used to store increasing sequence values for one or more entities. As with SequenceGenerator, you can place the TableGenerator annotation on any package, entity class, persistent field declaration (if your entity uses field access), or getter method for a persistent property (if your entity uses property access). TableGenerator has the following properties:

  • String name: The generator name. This property is required.

  • String table: The name of the generator table. If left unspecified, your vendor will choose a default table.

  • String schema: The named table's schema.

  • String catalog: The named table's catalog.

  • String pkColumnName: The name of the primary key column in the generator table. If unspecified, your implementation will choose a default.

  • String valueColumnName: The name of the column that holds the sequence value. If unspecified, your implementation will choose a default.

  • String pkColumnValue: The primary key column value of the row in the generator table holding this sequence value. You can use the same generator table for multiple logical sequences by supplying different pkColumnValue s. If you do not specify a value, the implementation will supply a default.

  • int initialValue: The value of the generator's first issued number.

  • int allocationSize: The number of values to allocate in memory for each trip to the database. Allocating values in memory allows the JPA runtime to avoid accessing the database for every sequence request. This number also specifies the amount that the sequence value is incremented each time the generator table is updated. Defaults to 50.

The XML equivalent is the table-generator element. This element's attributes correspond exactly to the above annotation's properties:

  • name

  • table

  • schema

  • catalog

  • pk-column-name

  • value-column-name

  • pk-column-value

  • initial-value

  • allocation-size

To use a table generator, set your GeneratedValue annotation's strategy property to GenerationType.TABLE, and its generator property to the table generator's declared name. Or equivalently, set your generated-value XML element's strategy attribute to TABLE and its generator attribute to the generator name.

5.3.  Example

Let's take advantage of generators in our entity model. Here are our updated mappings.

Example 12.4.  Generator Mapping

package org.mag;

@Entity
@IdClass(Magazine.MagazineId.class)
@Table(name="MAG")
public class Magazine {

    @Column(length=9)
    @Id private String isbn;
    @Id private String title;

    ...

    public static class MagazineId {
        ...
    }
}

@Entity
@Table(name="ART", uniqueConstraints=@Unique(columnNames="TITLE"))
@SequenceGenerator(name="ArticleSeq", sequenceName="ART_SEQ")
public class Article {

    @Id
    @GeneratedValue(strategy=GenerationType.SEQUENCE, generator="ArticleSeq") 
    private long id;

    ...
}


package org.mag.pub;

@Entity
@Table(name="COMP")
public class Company {

    @Column(name="CID")
    @Id private long id;

    ...
}

@Entity
@Table(name="AUTH")
public class Author {

    @Id
    @GeneratedValue(strategy=GenerationType.TABLE, generator="AuthorGen")
    @TableGenerator(name="AuthorGen", table="AUTH_GEN", pkColumnName="PK",
        valueColumnName="AID")
    @Column(name="AID", columnDefinition="INTEGER64")
    private long id;

    ...
}

@Embeddable
public class Address {
    ...
}


package org.mag.subscribe;

@MappedSuperclass
public abstract class Document {

    @Id
    @GeneratedValue(generate=GenerationType.IDENTITY)
    private long id;
 
    ...
}

@Entity
@Table(schema="CNTRCT")
public class Contract
    extends Document {

    ...
}

@Entity
@Table(name="SUB", schema="CNTRCT")
public class Subscription {

    @Id
    @GeneratedValue(strategy=GenerationType.IDENTITY)
    private long id;

    ...

    @Entity
    @Table(name="LINE_ITEM", schema="CNTRCT")
    public static class LineItem
        extends Contract {
        ...
    }
}

@Entity(name="Lifetime")
public class LifetimeSubscription
    extends Subscription {
    ...
}

@Entity(name="Trial")
public class TrialSubscription
    extends Subscription {
    ...
}

The same metadata for Article and Author expressed in XML form:

<entity class="org.mag.Article">
    <table name="ART">
        <unique-constraint>
            <column-name>TITLE</column-name>
        </unique-constraint>
    </table>
    <sequence-generator name="ArticleSeq" sequence-name="ART_SEQ"/>
    <attributes>
        <id name="id">
            <generated-value strategy="SEQUENCE" generator="ArticleSeq"/>
        </id>
        ...
    </attributes>
</entity>
<entity class="org.mag.pub.Author">
    <table name="AUTH"/>
    <attributes>
        <id name="id">
            <column name="AID" column-definition="INTEGER64"/>
            <generated-value strategy="TABLE" generator="AuthorGen"/>
            <table-generator name="AuthorGen" table="AUTH_GEN" 
                pk-column-name="PK" value-column-name="AID"/>
        </id>
        ...
    </attributes>
</entity>

6.  Inheritance

In the 1990's programmers coined the term impedance mismatch to describe the difficulties in bridging the object and relational worlds. Perhaps no feature of object modeling highlights the impedance mismatch better than inheritance. There is no natural, efficient way to represent an inheritance relationship in a relational database.

Luckily, JPA gives you a choice of inheritance strategies, making the best of a bad situation. The base entity class defines the inheritance strategy for the hierarchy with the Inheritance annotation. Inheritance has the following properties:

  • InheritanceType strategy: Enum value declaring the inheritance strategy for the hierarchy. Defaults to InheritanceType.SINGLE_TABLE. We detail each of the available strategies below.

The corresponding XML element is inheritance, which has a single attribute:

  • strategy: One of SINGLE_TABLE, JOINED, or TABLE_PER_CLASS.

The following sections describe JPA's standard inheritance strategies.

Note

OpenJPA allows you to vary your inheritance strategy for each class, rather than forcing a single strategy per inheritance hierarchy. See Section 7, “ Additional JPA Mappings ” in the Reference Guide for details.

6.1.  Single Table

The InheritanceType.SINGLE_TABLE strategy maps all classes in the hierarchy to the base class' table.

In our model, Subscription is mapped to the CNTRCT.SUB table. LifetimeSubscription, which extends Subscription, adds its field data to this table as well.

Example 12.5.  Single Table Mapping

@Entity
@Table(name="SUB", schema="CNTRCT")
@Inheritance(strategy=InheritanceType.SINGLE_TABLE)
public class Subscription {
    ...
}

@Entity(name="Lifetime")
public class LifetimeSubscription
    extends Subscription {
    ...
}

The same metadata expressed in XML form:

<entity class="org.mag.subcribe.Subscription">
    <table name="SUB" schema="CNTRCT"/>
    <inheritance strategy="SINGLE_TABLE"/>
    ...
</entity>
<entity class="org.mag.subscribe.LifetimeSubscription">
    ...
</entity>

Single table inheritance is the default strategy. Thus, we could omit the @Inheritance annotation in the example above and get the same result.

Note

Mapping subclass state to the superclass table is often called flat inheritance mapping.

6.1.1.  Advantages

Single table inheritance mapping is the fastest of all inheritance models, since it never requires a join to retrieve a persistent instance from the database. Similarly, persisting or updating a persistent instance requires only a single INSERT or UPDATE statement. Finally, relations to any class within a single table inheritance hierarchy are just as efficient as relations to a base class.

6.1.2.  Disadvantages

The larger the inheritance model gets, the "wider" the mapped table gets, in that for every field in the entire inheritance hierarchy, a column must exist in the mapped table. This may have undesirable consequence on the database size, since a wide or deep inheritance hierarchy will result in tables with many mostly-empty columns.

6.2.  Joined

The InheritanceType.JOINED strategy uses a different table for each class in the hierarchy. Each table only includes state declared in its class. Thus to load a subclass instance, the JPA implementation must read from the subclass table as well as the table of each ancestor class, up to the base entity class.

Note

Using joined subclass tables is also called vertical inheritance mapping.

PrimaryKeyJoinColumn annotations tell the JPA implementation how to join each subclass table record to the corresponding record in its direct superclass table. In our model, the LINE_ITEM.ID column joins to the CONTRACT.ID column. The PrimaryKeyJoinColumn annotation has the following properties:

  • String name: The name of the subclass table column. When there is a single identity field, defaults to that field's column name.

  • String referencedColumnName: The name of the superclass table column this subclass table column joins to. When there is a single identity field, defaults to that field's column name.

  • String columnDefinition: This property has the same meaning as the columnDefinition property on the Column annotation, described in Section 3, “ Column ”.

The XML equivalent is the primary-key-join-column element. Its attributes mirror the annotation properties described above:

  • name

  • referenced-column-name

  • column-definition

The example below shows how we use InheritanceTable.JOINED and a primary key join column to map our sample model according to the diagram above. Note that a primary key join column is not strictly needed, because there is only one identity column, and the subclass table column has the same name as the superclass table column. In this situation, the defaults suffice. However, we include the primary key join column for illustrative purposes.

Example 12.6.  Joined Subclass Tables

@Entity
@Table(schema="CNTRCT")
@Inheritance(strategy=InheritanceType.JOINED)
public class Contract
    extends Document {
    ...
}

public class Subscription {
    ...

    @Entity
    @Table(name="LINE_ITEM", schema="CNTRCT")
    @PrimaryKeyJoinColumn(name="ID", referencedColumnName="ID")
    public static class LineItem
        extends Contract {
        ...
    }
}

The same metadata expressed in XML form:

<entity class="org.mag.subcribe.Contract">
    <table schema="CNTRCT"/>
    <inheritance strategy="JOINED"/>
    ...
</entity>
<entity class="org.mag.subscribe.Subscription.LineItem">
    <table name="LINE_ITEM" schema="CNTRCT"/>
    <primary-key-join-column name="ID" referenced-column-name="PK"/>
    ...
</entity>

When there are multiple identity columns, you must define multiple PrimaryKeyJoinColumns using the aptly-named PrimaryKeyJoinColumns annotation. This annotation's value is an array of PrimaryKeyJoinColumn s. We could rewrite LineItem's mapping as:

@Entity
@Table(name="LINE_ITEM", schema="CNTRCT")
@PrimaryKeyJoinColumns({
    @PrimaryKeyJoinColumn(name="ID", referencedColumnName="ID")
})
public static class LineItem
    extends Contract {
    ...
}

In XML, simply list as many primary-key-join-column elements as necessary.

6.2.1.  Advantages

The joined strategy has the following advantages:

  1. Using joined subclass tables results in the most normalized database schema, meaning the schema with the least spurious or redundant data.

  2. As more subclasses are added to the data model over time, the only schema modification that needs to be made is the addition of corresponding subclass tables in the database (rather than having to change the structure of existing tables).

  3. Relations to a base class using this strategy can be loaded through standard joins and can use standard foreign keys, as opposed to the machinations required to load polymorphic relations to table-per-class base types, described below.

6.2.2.  Disadvantages

Aside from certain uses of the table-per-class strategy described below, the joined strategy is often the slowest of the inheritance models. Retrieving any subclass requires one or more database joins, and storing subclasses requires multiple INSERT or UPDATE statements. This is only the case when persistence operations are performed on subclasses; if most operations are performed on the least-derived persistent superclass, then this mapping is very fast.

Note

When executing a select against a hierarchy that uses joined subclass table inheritance, you must consider how to load subclass state. Section 7, “ Eager Fetching ” in the Reference Guide describes OpenJPA's options for efficient data loading.

6.3.  Table Per Class

Like the JOINED strategy, the InheritanceType.TABLE_PER_CLASS strategy uses a different table for each class in the hierarchy. Unlike the JOINED strategy, however, each table includes all state for an instance of the corresponding class. Thus to load a subclass instance, the JPA implementation must only read from the subclass table; it does not need to join to superclass tables.

Suppose that our sample model's Magazine class has a subclass Tabloid. The classes are mapped using the table-per-class strategy, as in the diagram above. In a table-per-class mapping, Magazine's table MAG contains all state declared in the base Magazine class. Tabloid maps to a separate table, TABLOID. This table contains not only the state declared in the Tabloid subclass, but all the base class state from Magazine as well. Thus the TABLOID table would contain columns for isbn, title, and other Magazine fields. These columns would default to the names used in Magazine's mapping metadata. Section 8.3, “ Embedded Mapping ” will show you how to use AttributeOverrides and AssociationOverride s to override superclass field mappings.

Example 12.7.  Table Per Class Mapping

@Entity
@Table(name="MAG")
@Inheritance(strategy=InheritanceType.TABLE_PER_CLASS)
public class Magazine {
    ...
}

@Entity
@Table(name="TABLOID")
public class Tabloid
    extends Magazine {
    ...
}

And the same classes in XML:

<entity class="org.mag.Magazine">
    <table name="MAG"/>
    <inheritance strategy="TABLE_PER_CLASS"/>
    ...
</entity>
<entity class="org.mag.Tabloid">
    <table name="TABLOID"/>
    ...
</entity>

6.3.1.  Advantages

The table-per-class strategy is very efficient when operating on instances of a known class. Under these conditions, the strategy never requires joining to superclass or subclass tables. Reads, joins, inserts, updates, and deletes are all efficient in the absence of polymorphic behavior. Also, as in the joined strategy, adding additional classes to the hierarchy does not require modifying existing class tables.

6.3.2.  Disadvantages

Polymorphic relations to non-leaf classes in a table-per-class hierarchy have many limitations. When the concrete subclass is not known, the related object could be in any of the subclass tables, making joins through the relation impossible. This ambiguity also affects identity lookups and queries; these operations require multiple SQL SELECTs (one for each possible subclass), or a complex UNION.

Note

Section 8.1, “ Table Per Class ” in the Reference Guide describes the limitations OpenJPA places on table-per-class mapping.

6.4.  Putting it All Together

Now that we have covered JPA's inheritance strategies, we can update our mapping document with inheritance information. Here is the complete model:

And here is the corresponding mapping metadata:

Example 12.8.  Inheritance Mapping

package org.mag;

@Entity
@IdClass(Magazine.MagazineId.class)
@Table(name="MAG")
public class Magazine {

    @Column(length=9)
    @Id private String isbn;
    @Id private String title;

    ...

    public static class MagazineId {
        ...
    }
}

@Entity
@Table(name="ART", uniqueConstraints=@Unique(columnNames="TITLE"))
@SequenceGenerator(name="ArticleSeq", sequenceName="ART_SEQ")
public class Article {

    @Id
    @GeneratedValue(strategy=GenerationType.SEQUENCE, generator="ArticleSeq") 
    private long id;

    ...
}


package org.mag.pub;

@Entity
@Table(name="COMP")
public class Company {

    @Column(name="CID")
    @Id private long id;

    ...
}

@Entity
@Table(name="AUTH")
public class Author {

    @Id
    @GeneratedValue(strategy=GenerationType.TABLE, generator="AuthorGen")
    @TableGenerator(name="AuthorGen", table="AUTH_GEN", pkColumnName="PK",
        valueColumnName="AID")
    @Column(name="AID", columnDefinition="INTEGER64")
    private long id;

    ...
}

@Embeddable
public class Address {
    ...
}


package org.mag.subscribe;

@MappedSuperclass
public abstract class Document {

    @Id
    @GeneratedValue(strategy=GenerationType.IDENTITY)
    private long id;

    ...
}

@Entity
@Table(schema="CNTRCT")
@Inheritance(strategy=InheritanceType.JOINED)
public class Contract
    extends Document {
    ...
}

@Entity
@Table(name="SUB", schema="CNTRCT")
@Inheritance(strategy=InheritanceType.SINGLE_TABLE)
public class Subscription {

    @Id
    @GeneratedValue(strategy=GenerationType.IDENTITY)
    private long id;

    ...

    @Entity
    @Table(name="LINE_ITEM", schema="CNTRCT")
    @PrimaryKeyJoinColumn(name="ID", referencedColumnName="ID")
    public static class LineItem
        extends Contract {
        ...
    }
}

@Entity(name="Lifetime")
public class LifetimeSubscription
    extends Subscription {
    ...
}

@Entity(name="Trial")
public class TrialSubscription
    extends Subscription {
    ...
}

The same metadata expressed in XML form:

<entity-mappings xmlns="http://java.sun.com/xml/ns/persistence/orm" 
    xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" 
    xsi:schemaLocation="http://java.sun.com/xml/ns/persistence/orm orm_1_0.xsd"
    version="1.0">
    <mapped-superclass class="org.mag.subscribe.Document">
        <attributes>
            <id name="id">
                <generated-value strategy="IDENTITY"/>
            </id>
            ...
        </attributes>
    </mapped-superclass>
    <entity class="org.mag.Magazine">
        <table name="MAG"/>
        <id-class="org.mag.Magazine.MagazineId"/>
        <attributes>
            <id name="isbn">
                <column length="9"/>
            </id>
            <id name="title"/>
            ...
        </attributes>
    </entity>
    <entity class="org.mag.Article">
        <table name="ART">
            <unique-constraint>
               <column-name>TITLE</column-name>
            </unique-constraint>
        </table>
        <sequence-generator name="ArticleSeq" sequence-name="ART_SEQ"/>
        <attributes>
            <id name="id">
                <generated-value strategy="SEQUENCE" generator="ArticleSeq"/>
            </id>
           ...
        </attributes>
    </entity>
    <entity class="org.mag.pub.Company">
        <table name="COMP"/>
        <attributes>
            <id name="id">
                <column name="CID"/>
            </id>
            ...
        </attributes>
    </entity>
    <entity class="org.mag.pub.Author">
        <table name="AUTH"/>
        <attributes>
            <id name="id">
                <column name="AID" column-definition="INTEGER64"/>
                <generated-value strategy="TABLE" generator="AuthorGen"/>
                <table-generator name="AuthorGen" table="AUTH_GEN" 
                    pk-column-name="PK" value-column-name="AID"/>
            </id>
            ...
        </attributes>
    </entity>
    <entity class="org.mag.subcribe.Contract">
        <table schema="CNTRCT"/>
        <inheritance strategy="JOINED"/>
        <attributes>
            ...
        </attributes>
    </entity>
    <entity class="org.mag.subcribe.Subscription">
        <table name="SUB" schema="CNTRCT"/>
        <inheritance strategy="SINGLE_TABLE"/>
        <attributes>
            <id name="id">
                <generated-value strategy="IDENTITY"/>
            </id>
            ...
        </attributes>
    </entity>
    <entity class="org.mag.subscribe.Subscription.LineItem">
        <table name="LINE_ITEM" schema="CNTRCT"/>
        <primary-key-join-column name="ID" referenced-column-name="PK"/>
        ...
    </entity>
    <entity class="org.mag.subscribe.LifetimeSubscription" name="Lifetime">
        ...
    </entity>
    <entity class="org.mag.subscribe.TrialSubscription" name="Trial">
        ...
    </entity>
</entity-mappings>

7.  Discriminator

The single table inheritance strategy results in a single table containing records for two or more different classes in an inheritance hierarchy. Similarly, using the joined strategy results in the superclass table holding records for superclass instances as well as for the superclass state of subclass instances. When selecting data, JPA needs a way to differentiate a row representing an object of one class from a row representing an object of another. That is the job of the discriminator column.

The discriminator column is always in the table of the base entity. It holds a different value for records of each class, allowing the JPA runtime to determine what class of object each row represents.

The DiscriminatorColumn annotation represents a discriminator column. It has these properties:

  • String name: The column name. Defaults to DTYPE .

  • length: For string discriminator values, the length of the column. Defaults to 31.

  • String columnDefinition: This property has the same meaning as the columnDefinition property on the Column annotation, described in Section 3, “ Column ”.

  • DiscriminatorType discriminatorType: Enum value declaring the discriminator strategy of the hierarchy.

The corresponding XML element is discriminator-column. Its attribues mirror the annotation properties above:

  • name

  • length

  • column-definition

  • discriminator-type: One of STRING, CHAR, or INTEGER.

The DiscriminatorValue annotation specifies the discriminator value for each class. Though this annotation's value is always a string, the implementation will parse it according to the DiscriminatorColumn's discriminatorType property above. The type defaults to DiscriminatorType.STRING, but may be DiscriminatorType.CHAR or DiscriminatorType.INTEGER. If you do not specify a DiscriminatorValue, the provider will choose an appropriate default.

The corresponding XML element is discriminator-value. The text within this element is parsed as the discriminator value.

Note

OpenJPA assumes your model employs a discriminator column if any of the following are true:

  1. The base entity explicitly declares an inheritance type of SINGLE_TABLE.

  2. The base entity sets a discriminator value.

  3. The base entity declares a discriminator column.

Only SINGLE_TABLE inheritance hierarchies require a discriminator column and values. JOINED hierarchies can use a discriminator to make some operations more efficient, but do not require one. TABLE_PER_CLASS hierarchies have no use for a discriminator.

OpenJPA defines additional discriminator strategies; see Section 7, “ Additional JPA Mappings ” in the Reference Guide for details. OpenJPA also supports final entity classes. OpenJPA does not use a discriminator on final classes.

We can now translate our newfound knowledge of JPA discriminators into concrete JPA mappings. We first extend our diagram with discriminator columns:

Next, we present the updated mapping document. Notice that in this version, we have removed explicit inheritance annotations when the defaults sufficed. Also, notice that entities using the default DTYPE discriminator column mapping do not need an explicit DiscriminatorColumn annotation.

Example 12.9.  Discriminator Mapping

package org.mag;

@Entity
@IdClass(Magazine.MagazineId.class)
@Table(name="MAG")
@DiscriminatorValue("Mag")
public class Magazine {

    @Column(length=9)
    @Id private String isbn;
    @Id private String title;

    ...

    public static class MagazineId {
        ...
    }
}

@Entity
@Table(name="ART", uniqueConstraints=@Unique(columnNames="TITLE"))
@SequenceGenerator(name="ArticleSeq", sequenceName="ART_SEQ")
public class Article {

    @Id
    @GeneratedValue(strategy=GenerationType.SEQUENCE, generator="ArticleSeq") 
    private long id;

    ...
}


package org.mag.pub;

@Entity
@Table(name="COMP")
public class Company {

    @Column(name="CID")
    @Id private long id;

    ...
}

@Entity
@Table(name="AUTH")
public class Author {

    @Id
    @GeneratedValue(strategy=GenerationType.TABLE, generator="AuthorGen")
    @TableGenerator(name="AuthorGen", table="AUTH_GEN", pkColumnName="PK",
        valueColumnName="AID")
    @Column(name="AID", columnDefinition="INTEGER64")
    private long id;

    ...
}

@Embeddable
public class Address {
    ...
}


package org.mag.subscribe;

@MappedSuperclass
public abstract class Document {

    @Id
    @GeneratedValue(strategy=GenerationType.IDENTITY)
    private long id;

    ...
}

@Entity
@Table(schema="CNTRCT")
@Inheritance(strategy=InheritanceType.JOINED)
@DiscriminatorColumn(name="CTYPE")
public class Contract
    extends Document {
    ...
}

@Entity
@Table(name="SUB", schema="CNTRCT")
@DiscriminatorColumn(name="KIND", discriminatorType=DiscriminatorType.INTEGER)
@DiscriminatorValue("1")
public class Subscription {

    @Id
    @GeneratedValue(strategy=GenerationType.IDENTITY)
    private long id;

    ...

    @Entity
    @Table(name="LINE_ITEM", schema="CNTRCT")
    public static class LineItem
        extends Contract {
        ...
    }
}

@Entity(name="Lifetime")
@DiscriminatorValue("2")
public class LifetimeSubscription
    extends Subscription {
    ...
}

@Entity(name="Trial")
@DiscriminatorValue("3")
public class TrialSubscription
    extends Subscription {
    ...
}

The same metadata expressed in XML:

<entity-mappings xmlns="http://java.sun.com/xml/ns/persistence/orm" 
    xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" 
    xsi:schemaLocation="http://java.sun.com/xml/ns/persistence/orm orm_1_0.xsd"
    version="1.0">
    <mapped-superclass class="org.mag.subscribe.Document">
        <attributes>
            <id name="id">
                <generated-value strategy="IDENTITY"/>
            </id>
            ...
        </attributes>
    </mapped-superclass>
    <entity class="org.mag.Magazine">
        <table name="MAG"/>
        <id-class="org.mag.Magazine.MagazineId"/>
        <discriminator-value>Mag</discriminator-value>
        <attributes>
            <id name="isbn">
                <column length="9"/>
            </id>
            <id name="title"/>
            ...
        </attributes>
    </entity>
    <entity class="org.mag.Article">
        <table name="ART">
            <unique-constraint>
                <column-name>TITLE</column-name>
            </unique-constraint>
        </table>
        <sequence-generator name="ArticleSeq" sequence-name="ART_SEQ"/>
        <attributes>
            <id name="id">
                <generated-value strategy="SEQUENCE" generator="ArticleSeq"/>
            </id>
            ...
        </attributes>
    </entity>
    <entity class="org.mag.pub.Company">
        <table name="COMP"/>
        <attributes>
            <id name="id">
                <column name="CID"/>
            </id>
            ...
        </attributes>
    </entity>
    <entity class="org.mag.pub.Author">
        <table name="AUTH"/>
        <attributes>
            <id name="id">
                <column name="AID" column-definition="INTEGER64"/>
                <generated-value strategy="TABLE" generator="AuthorGen"/>
                <table-generator name="AuthorGen" table="AUTH_GEN" 
                    pk-column-name="PK" value-column-name="AID"/>
            </id>
            ...
        </attributes>
    </entity>
    <entity class="org.mag.subcribe.Contract">
        <table schema="CNTRCT"/>
        <inheritance strategy="JOINED"/>
        <discriminator-column name="CTYPE"/>
        <attributes>
            ...
        </attributes>
    </entity>
    <entity class="org.mag.subcribe.Subscription">
        <table name="SUB" schema="CNTRCT"/>
        <inheritance strategy="SINGLE_TABLE"/>
        <discriminator-value>1</discriminator-value>
        <discriminator-column name="KIND" discriminator-type="INTEGER"/>
        <attributes>
            <id name="id">
                <generated-value strategy="IDENTITY"/>
            </id>
            ...
        </attributes>
    </entity>
    <entity class="org.mag.subscribe.Subscription.LineItem">
        <table name="LINE_ITEM" schema="CNTRCT"/>
        <primary-key-join-column name="ID" referenced-column-name="PK"/>
        ...
    </entity>
    <entity class="org.mag.subscribe.LifetimeSubscription" name="Lifetime">
        <discriminator-value>2</discriminator-value>
        ...
    </entity>
    <entity class="org.mag.subscribe.TrialSubscription" name="Trial">
        <discriminator-value>3</discriminator-value>
        ...
    </entity>
</entity-mappings>

8.  Field Mapping

The following sections enumerate the myriad of field mappings JPA supports. JPA augments the persistence metadata covered in Chapter 5, Metadata with many new object-relational annotations. As we explore the library of standard mappings, we introduce each of these enhancements in context.

Note

OpenJPA supports many additional field types, and allows you to create custom mappings for unsupported field types or database schemas. See the Reference Guide's Chapter 7, Mapping for complete coverage of OpenJPA's mapping capabilities.

8.1.  Basic Mapping

A basic field mapping stores the field value directly into a database column. The following field metadata types use basic mapping. These types were defined in Section 2, “ Field and Property Metadata ”.

In fact, you have already seen examples of basic field mappings in this chapter - the mapping of all identity fields in Example 12.3, “ Identity Mapping ”. As you saw in that section, to write a basic field mapping you use the Column annotation to describe the column the field value is stored in. We discussed the Column annotation in Section 3, “ Column ”. Recall that the name of the column defaults to the field name, and the type of the column defaults to an appropriate type for the field type. These defaults allow you to sometimes omit the annotation altogether.

8.1.1.  LOBs

Adding the Lob marker annotation to a basic field signals that the data is to be stored as a LOB (Large OBject). If the field holds string or character data, it will map to a CLOB (Character Large OBject) database column. If the field holds any other data type, it will be stored as binary data in a BLOB (Binary Large OBject) column. The implementation will serialize the Java value if needed.

The equivalent XML element is lob, which has no children or attributes.

8.1.2.  Enumerated

You can apply the Enumerated annotation to your Enum fields to control how they map to the database. The Enumerated annotation's value one of the following constants from the EnumType enum:

  • EnumType.ORDINAL: The default. The persistence implementation places the ordinal value of the enum in a numeric column. This is an efficient mapping, but may break if you rearrange the Java enum declaration.

  • EnumType.STRING: Store the name of the enum value rather than the ordinal. This mapping uses a VARCHAR column rather than a numeric one.

The Enumerated annotation is optional. Any un-annotated enumeration field defaults to ORDINAL mapping.

The corresponding XML element is enumerated. Its embedded text must be one of STRING or ORIDINAL.

8.1.3.  Temporal Types

The Temporal annotation determines how the implementation handles your basic java.util.Date and java.util.Calendar fields at the JDBC level. The Temporal annotation's value is a constant from the TemporalType enum. Available values are:

  • TemporalType.TIMESTAMP: The default. Use JDBC's timestamp APIs to manipulate the column data.

  • TemporalType.DATE: Use JDBC's SQL date APIs to manipulate the column data.

  • TemporalType.TIME: Use JDBC's time APIs to manipulate the column data.

If the Temporal annotation is omitted, the implementation will treat the data as a timestamp.

The corresponding XML element is temporal, whose text value must be one of: TIME, DATE, or TIMESTAMP.

8.1.4.  The Updated Mappings

Below we present an updated diagram of our model and its associated database schema, followed by the corresponding mapping metadata. Note that the mapping metadata relies on defaults where possible. Also note that as a mapped superclass, Document can define mappings that will automatically transfer to its subclass' tables. In Section 8.3, “ Embedded Mapping ”, you will see how a subclass can override its mapped superclass' mappings.

Example 12.10.  Basic Field Mapping

package org.mag;

@Entity
@IdClass(Magazine.MagazineId.class)
@Table(name="MAG")
@DiscriminatorValue("Mag")
public class Magazine {

    @Column(length=9)
    @Id private String isbn;
    @Id private String title;

    @Column(name="VERS")
    @Version private int version;
    
    private String name;
    private double price;

    @Column(name="COPIES")
    private int copiesSold;

    ...

    public static class MagazineId {
        ...
    }
}

@Entity
@Table(name="ART", uniqueConstraints=@Unique(columnNames="TITLE"))
@SequenceGenerator(name="ArticleSeq", sequenceName="ART_SEQ")
public class Article {

    @Id
    @GeneratedValue(strategy=GenerationType.SEQUENCE, generator="ArticleSeq") 
    private long id;

    @Column(name="VERS")
    @Version private int version;

    private String title;
    private byte[] content;

    ...
}


package org.mag.pub;

@Entity
@Table(name="COMP")
public class Company {

    @Column(name="CID")
    @Id private long id;

    @Column(name="VERS")
    @Version private int version;

    private String name;

    @Column(name="REV")
    private double revenue;

    ...
}

@Entity
@Table(name="AUTH")
public class Author {

    @Id
    @GeneratedValue(strategy=GenerationType.TABLE, generator="AuthorGen")
    @TableGenerator(name="AuthorGen", table="AUTH_GEN", pkColumnName="PK",
        valueColumnName="AID")
    @Column(name="AID", columnDefinition="INTEGER64")
    private long id;

    @Column(name="VERS")
    @Version private int version;

    @Column(name="FNAME")
    private String firstName;

    @Column(name="LNAME")
    private String lastName;

    ...
}

@Embeddable
public class Address {
    ...
}

package org.mag.subscribe;

@MappedSuperclass
public abstract class Document {

    @Id
    @GeneratedValue(strategy=GenerationType.IDENTITY)
    private long id;

    @Column(name="VERS")
    @Version private int version;

    ...
}

@Entity
@Table(schema="CNTRCT")
@Inheritance(strategy=InheritanceType.JOINED)
@DiscriminatorColumn(name="CTYPE")
public class Contract
    extends Document {

    @Lob
    private String terms;

    ...
}

@Entity
@Table(name="SUB", schema="CNTRCT")
@DiscriminatorColumn(name="KIND", discriminatorType=DiscriminatorType.INTEGER)
@DiscriminatorValue("1")
public class Subscription {

    @Id
    @GeneratedValue(strategy=GenerationType.IDENTITY)
    private long id;

    @Column(name="VERS")
    @Version private int version;

    @Column(name="START")
    private Date startDate;

    @Column(name="PAY")
    private double payment;

    ...

    @Entity
    @Table(name="LINE_ITEM", schema="CNTRCT")
    public static class LineItem
        extends Contract {

        @Column(name="COMM")
        private String comments;

        private double price;
        private long num;
        ...
    }
}

@Entity(name="Lifetime")
@DiscriminatorValue("2")
public class LifetimeSubscription
    extends Subscription {

    @Basic(fetch=FetchType.LAZY)
    @Column(name="ELITE")
    private boolean getEliteClub () { ... }
    public void setEliteClub (boolean elite) { ... }

    ...
}

@Entity(name="Trial")
@DiscriminatorValue("3")
public class TrialSubscription
    extends Subscription {

    @Column(name="END")
    public Date getEndDate () { ... }
    public void setEndDate (Date end) { ... }

    ...
}

The same metadata expressed in XML:

<entity-mappings xmlns="http://java.sun.com/xml/ns/persistence/orm" 
    xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" 
    xsi:schemaLocation="http://java.sun.com/xml/ns/persistence/orm orm_1_0.xsd"
    version="1.0">
    <mapped-superclass class="org.mag.subscribe.Document">
        <attributes>
            <id name="id">
                <generated-value strategy="IDENTITY"/>
            </id>
            <version name="version">
                <column name="VERS"/>
            </version>
            ...
        </attributes>
    </mapped-superclass>
    <entity class="org.mag.Magazine">
        <table name="MAG"/>
        <id-class="org.mag.Magazine.MagazineId"/>
        <discriminator-value>Mag</discriminator-value>
        <attributes>
            <id name="isbn">
                <column length="9"/>
            </id>
            <id name="title"/>
            <basic name="name"/>
            <basic name="price"/>
            <basic name="copiesSold">
                <column name="COPIES"/>
            </basic>
            <version name="version">
                <column name="VERS"/>
            </version>
            ...
        </attributes>
    </entity>
    <entity class="org.mag.Article">
        <table name="ART">
            <unique-constraint>
               <column-name>TITLE</column-name>
            </unique-constraint>
        </table>
        <sequence-generator name="ArticleSeq", sequenceName="ART_SEQ"/>
        <attributes>
            <id name="id">
                <generated-value strategy="SEQUENCE" generator="ArticleSeq"/>
            </id>
            <basic name="title"/>
            <basic name="content"/>
            <version name="version">
                <column name="VERS"/>
            </version>
            ...
        </attributes>
    </entity>
    <entity class="org.mag.pub.Company">
        <table name="COMP"/>
        <attributes>
            <id name="id">
                <column name="CID"/>
            </id>
            <basic name="name"/>
            <basic name="revenue">
                <column name="REV"/>
            </basic>
        </attributes>
    </entity>
    <entity class="org.mag.pub.Author">
        <table name="AUTH"/>
        <attributes>
            <id name="id">
                <column name="AID" column-definition="INTEGER64"/>
                <generated-value strategy="TABLE" generator="AuthorGen"/>
                <table-generator name="AuthorGen" table="AUTH_GEN" 
                    pk-column-name="PK" value-column-name="AID"/>
            </id>
            <basic name="firstName">
                <column name="FNAME"/>
            </basic>
            <basic name="lastName">
                <column name="LNAME"/>
            </basic>
            <version name="version">
                <column name="VERS"/>
            </version>
            ...
        </attributes>
    </entity>
    <entity class="org.mag.subcribe.Contract">
        <table schema="CNTRCT"/>
        <inheritance strategy="JOINED"/>
        <discriminator-column name="CTYPE"/>
        <attributes>
            <basic name="terms">
                <lob/>
            </basic>
            ...
        </attributes>
    </entity>
    <entity class="org.mag.subcribe.Subscription">
        <table name="SUB" schema="CNTRCT"/>
        <inheritance strategy="SINGLE_TABLE"/>
        <discriminator-value>1</discriminator-value>
        <discriminator-column name="KIND" discriminator-type="INTEGER"/>
        <attributes>
            <id name="id">
                <generated-value strategy="IDENTITY"/>
            </id>
            <basic name="payment">
                <column name="PAY"/>
            </basic>
            <basic name="startDate">
                <column name="START"/>
            </basic>
            <version name="version">
                <column name="VERS"/>
            </version>
            ...
        </attributes>
    </entity>
    <entity class="org.mag.subscribe.Subscription.LineItem">
        <table name="LINE_ITEM" schema="CNTRCT"/>
        <primary-key-join-column name="ID" referenced-column-name="PK"/>
        <attributes>
            <basic name="comments">
                <column name="COMM"/>
            </basic>
            <basic name="price"/>
            <basic name="num"/>
            ...
        </attributes>
    </entity>
    <entity class="org.mag.subscribe.LifetimeSubscription" name="Lifetime">
        <discriminator-value>2</discriminator-value>
        <attributes>
            <basic name="eliteClub" fetch="LAZY">
                <column name="ELITE"/>
            </basic>
            ...
        </attributes>
    </entity>
    <entity class="org.mag.subscribe.TrialSubscription" name="Trial">
        <discriminator-value>3</discriminator-value>
        <attributes>
            <basic name="endDate">
                <column name="END"/>
            </basic>
            ...
        </attributes>
    </entity>
</entity-mappings>

8.2.  Secondary Tables

Sometimes a a logical record is spread over multiple database tables. JPA calls a class' declared table the primary table, and calls other tables that make up a logical record secondary tables. You can map any persistent field to a secondary table. Just write the standard field mapping, then perform these two additional steps:

  1. Set the table attribute of each of the field's columns or join columns to the name of the secondary table.

  2. Define the secondary table on the entity class declaration.

You define secondary tables with the SecondaryTable annotation. This annotation has all the properties of the Table annotation covered in Section 1, “ Table ” , plus a pkJoinColumns property.

The pkJoinColumns property is an array of PrimaryKeyJoinColumns dictating how to join secondary table records to their owning primary table records. Each PrimaryKeyJoinColumn joins a secondary table column to a primary key column in the primary table. See Section 6.2, “ Joined ” above for coverage of PrimaryKeyJoinColumn's properties.

The corresponding XML element is secondary-table. This element has all the attributes of the table element, but also accepts nested primary-key-join-column elements.

In the following example, we move the Article.content field we mapped in Section 8.1, “ Basic Mapping ” into a joined secondary table, like so:

Example 12.11.  Secondary Table Field Mapping

package org.mag;

@Entity
@Table(name="ART")
@SecondaryTable(name="ART_DATA", 
    pkJoinColumns=@PrimaryKeyJoinColumn(name="ART_ID", referencedColumnName="ID"))
public class Article {

    @Id private long id;

    @Column(table="ART_DATA")
    private byte[] content;

    ...
}

And in XML:

<entity class="org.mag.Article">
    <table name="ART">
    <secondary-table name="ART_DATA">
        <primary-key-join-column name="ART_ID" referenced-column-name="ID"/>
    </secondary-table>
    <attributes>
        <id name="id"/>
        <basic name="content">
            <column table="ART_DATA"/>
        </basic>
        ...
    </attributes>
</entity>

8.3.  Embedded Mapping

Chapter 5, Metadata describes JPA's concept of embeddable objects. The field values of embedded objects are stored as part of the owning record, rather than as a separate database record. Thus, instead of mapping a relation to an embeddable object as a foreign key, you map all the fields of the embeddable instance to columns in the owning field's table.

JPA defaults the embedded column names and descriptions to those of the embeddable class' field mappings. The AttributeOverride annotation overrides a basic embedded mapping. This annotation has the following properties:

  • String name: The name of the embedded class' field being mapped to this class' table.

  • Column column: The column defining the mapping of the embedded class' field to this class' table.

The corresponding XML element is attribute-override. It has a single name attribute to name the field being overridden, and a single column child element.

To declare multiple overrides, use the AttributeOverrides annotation, whose value is an array of AttributeOverride s. In XML, simply list multiple attribute-override elements in succession.

To override a many to one or one to one relationship, use the AssociationOverride annotation in place of AttributeOverride. AssociationOverride has the following properties:

  • String name: The name of the embedded class' field being mapped to this class' table.

  • JoinColumn[] joinColumns: The foreign key columns joining to the related record.

The corresponding XML element is association-override. It has a single name attribute to name the field being overridden, and one or more join-column child elements.

To declare multiple relation overrides, use the AssociationOverrides annotation, whose value is an array of AssociationOverride s. In XML, simply list multiple association-override elements in succession.

Example 12.12.  Embedded Field Mapping

In this example, Company overrides the default mapping of Address.street and Address.city. All other embedded mappings are taken from the Address embeddable class.

package org.mag.pub;

@Entity
@Table(name="COMP")
public class Company {

    @Embedded
    @AttributeOverrides({
        @AttributeOverride(name="street", column=@Column(name="STRT")),
        @AttributeOverride(name="city", column=@Column(name="ACITY"))
    })
    private Address address;
    
    ...
}

@Entity
@Table(name="AUTH")
public class Author {

    // use all defaults from Address class mappings
    private Address address;
    
    ...
}

@Embeddable
public class Address {

    private String street;
    private String city;
    @Column(columnDefinition="CHAR(2)")
    private String state;
    private String zip;
}

The same metadata expressed in XML:

<entity class="org.mag.pub.Company">
    <table name="COMP"/>
    <attributes>
        ...
        <embedded name="address">
            <attribute-override name="street">
               <column name="STRT"/>
            </attribute-override>
            <attribute-override name="city">
               <column name="ACITY"/>
            </attribute-override>
        </embedded>
    </attributes>
</entity>
<entity class="org.mag.pub.Author">
    <table name="AUTH"/>
    <attributes>
        <embedded name="address">
            <!-- use all defaults from Address -->
        </embedded>
    </attributes>
</entity>
<embeddable class="org.mag.pub.Address">
    <attributes>
        <basic name="street"/>
        <basic name="city"/>
        <basic name="state">
            <column column-definition="CHAR(2)"/>
        </basic>
        <basic name="zip"/>
    </attributes>
</embeddable>

You can also use attribute overrides on an entity class to override mappings defined by its mapped superclass or table-per-class superclass. The example below re-maps the Document.version field to the Contract table's CVERSION column.

Example 12.13.  Mapping Mapped Superclass Field

@MappedSuperclass
public abstract class Document {

    @Column(name="VERS")
    @Version private int version;

    ...
}

@Entity
@Table(schema="CNTRCT")
@Inheritance(strategy=InheritanceType.JOINED)
@DiscriminatorColumn(name="CTYPE")
@AttributeOverride(name="version", column=@Column(name="CVERSION"))
public class Contract
    extends Document {
    ...
}

The same metadata expressed in XML form:

<mapped-superclass class="org.mag.subcribe.Document">
    <attributes>
        <version name="version">
            <column name="VERS">
        </version>
        ...
    </attributes>
</mapped-superclass>
<entity class="org.mag.subcribe.Contract">
    <table schema="CNTRCT"/>
    <inheritance strategy="JOINED"/>
    <discriminator-column name="CTYPE"/>
    <attribute-override name="version">
        <column name="CVERSION"/>
    </attribute-override>
    <attributes>
        ...
    </attributes>
</entity>

8.4.  Direct Relations

A direct relation is a non-embedded persistent field that holds a reference to another entity. many to one and one to one metadata field types are mapped as direct relations. Our model has three direct relations: Magazine's publisher field is a direct relation to a Company, Magazine's coverArticle field is a direct relation to Article, and the LineItem.magazine field is a direct relation to a Magazine. Direct relations are represented in the database by foreign key columns:

You typically map a direct relation with JoinColumn annotations describing how the local foreign key columns join to the primary key columns of the related record. The JoinColumn annotation exposes the following properties:

  • String name: The name of the foreign key column. Defaults to the relation field name, plus an underscore, plus the name of the referenced primary key column.

  • String referencedColumnName: The name of the primary key column being joined to. If there is only one identity field in the related entity class, the join column name defaults to the name of the identity field's column.

  • boolean unique: Whether this column is guaranteed to hold unique values for all rows. Defaults to false.

JoinColumn also has the same nullable , insertable, updatable, columnDefinition, and table properties as the Column annotation. See Section 3, “ Column ” for details on these properties.

The join-column element represents a join column in XML. Its attributes mirror the above annotation's properties:

  • name

  • referenced-column-name

  • unique

  • nullable

  • insertable

  • updatable

  • column-definition

  • table

When there are multiple columns involved in the join, as when a LineItem references a Magazine in our model, the JoinColumns annotation allows you to specify an array of JoinColumn values. In XML, simply list multiple join-column elements.

Note

OpenJPA supports many non-standard joins. See Section 6, “ Non-Standard Joins ” in the Reference Guide for details.

Example 12.14.  Direct Relation Field Mapping

package org.mag;

@Table(name="AUTH")
public class Magazine {

    @Column(length=9)
    @Id private String isbn;
    @Id private String title;

    @OneToOne
    @JoinColumn(name="COVER_ID" referencedColumnName="ID")
    private Article coverArticle;

    @ManyToOne
    @JoinColumn(name="PUB_ID" referencedColumnName="CID")
    private Company publisher;

    ...
}

@Table(name="ART")
public class Article {

    @Id private long id;

    ...
}


package org.mag.pub;

@Table(name="COMP")
public class Company {

    @Column(name="CID")
    @Id private long id;

    ...
}


package org.mag.subscribe;

public class Subscription {
    ...

    @Table(name="LINE_ITEM", schema="CNTRCT")
    public static class LineItem
        extends Contract {

        @ManyToOne
        @JoinColumns({
            @JoinColumn(name="MAG_ISBN" referencedColumnName="ISBN"),
            @JoinColumn(name="MAG_TITLE" referencedColumnName="TITLE")
        })
        private Magazine magazine;

        ...
    }
}

The same metadata expressed in XML form:

<entity class="org.mag.Magazine">
    <table name="MAG"/>
    <id-class="org.mag.Magazine.MagazineId"/>
    <attributes>
        <id name="isbn">
            <column length="9"/>
        </id>
        <id name="title"/>
        <one-to-one name="coverArticle">
            <join-column name="COVER_ID" referenced-column-name="ID"/>
        </one-to-one>
        <many-to-one name="publisher">
            <join-column name="PUB_IC" referenced-column-name="CID"/>
        </many-to-one>
        ...
    </attributes>
</entity>
<entity class="org.mag.Article">
    <table name="ART"/>
    <attributes>
        <id name="id"/>
        ...
    </attributes>
</entity>
<entity class="org.mag.pub.Company">
    <table name="COMP"/>
    <attributes>
        <id name="id">
            <column name="CID"/>
        </id>
        ...
    </attributes>
</entity>
<entity class="org.mag.subscribe.Subscription.LineItem">
    <table name="LINE_ITEM" schema="CNTRCT"/>
    <primary-key-join-column name="ID" referenced-column-name="PK"/>
    <attributes>
        <many-to-one name="magazine">
            <join-column name="MAG_ISBN" referenced-column-name="ISBN"/>
            <join-column name="MAG_TITLE" referenced-column-name="TITLE"/>
        </many-to-one>
        ...
    </attributes>
</entity>

When the entities in a one to one relation join on shared primary key values rather than separate foreign key columns, use the PrimaryKeyJoinColumn(s) annotation or primary-key-join-column elements in place of JoinColumn(s) / join-column elements.

8.5.  Join Table

A join table consists of two foreign keys. Each row of a join table associates two objects together. JPA uses join tables to represent collections of entity objects: one foreign key refers back to the collection's owner, and the other refers to a collection element.

one to many and many to many metadata field types can map to join tables. Several fields in our model use join table mappings, including Magazine.articles and Article.authors.

You define join tables with the JoinTable annotation. This annotation has the following properties:

  • String name: Table name. If not given, the name of the table defaults to the name of the owning entity's table, plus an underscore, plus the name of the related entity's table.

  • String catalog: Table catalog.

  • String schema: Table schema.

  • JoinColumn[] joinColumns: Array of JoinColumn sshowing how to associate join table records with the owning row in the primary table. This property mirrors the pkJoinColumns property of the SecondaryTable annotation in functionality. See Section 8.2, “ Secondary Tables ” to refresh your memory on secondary tables.

    If this is a bidirectional relation (see Section 2.9.1, “ Bidirectional Relations ” ), the name of a join column defaults to the inverse field name, plus an underscore, plus the referenced primary key column name. Otherwise, the join column name defaults to the field's owning entity name, plus an underscore, plus the referenced primary key column name.

  • JoinColumn[] inverseJoinColumns: Array of JoinColumns showing how to associate join table records with the records that form the elements of the collection. These join columns are used just like the join columns for direct relations, and they have the same naming defaults. Read Section 8.4, “ Direct Relations ” for a review of direct relation mapping.

join-table is the corresponding XML element. It has the same attributes as the table element, but includes the ability to nest join-column and inverse-join-column elements as children. We have seen join-column elements already; inverse-join-column elements have the same attributes.

Here are the join table mappings for the diagram above.

Example 12.15.  Join Table Mapping

package org.mag;

@Entity
@Table(name="MAG")
public class Magazine {

    @Column(length=9)
    @Id private String isbn;
    @Id private String title;

    @OneToMany(...)
    @OrderBy
    @JoinTable(name="MAG_ARTS",
        joinColumns={
            @JoinColumn(name="MAG_ISBN", referencedColumnName="ISBN"),
            @JoinColumn(name="MAG_TITLE", referencedColumnName="TITLE")
        },
        inverseJoinColumns=@JoinColumn(name="ART_ID", referencedColumnName="ID"))
    private Collection<Article> articles;

    ...
}

@Entity
@Table(name="ART")
public class Article {

    @Id private long id;

    @ManyToMany(cascade=CascadeType.PERSIST)
    @OrderBy("lastName, firstName")
    @JoinTable(name="ART_AUTHS",
        joinColumns=@JoinColumn(name="ART_ID", referencedColumnName="ID"),
        inverseJoinColumns=@JoinColumn(name="AUTH_ID", referencedColumnName="AID"))
    private Collection<Author> authors;

    ...
}


package org.mag.pub;

@Entity
@Table(name="AUTH")
public class Author {

    @Column(name="AID", columnDefinition="INTEGER64")
    @Id private long id;

    ...
}

The same metadata expressed in XML:

<entity class="org.mag.Magazine">
    <table name="MAG"/>
    <attributes>
        <id name="isbn">
            <column length="9"/>
        </id>
        <id name="title"/>
        <one-to-many name="articles">
            <order-by/>
            <join-table name="MAG_ARTS">
                <join-column name="MAG_ISBN" referenced-column-name="ISBN"/>
                <join-column name="MAG_TITLE" referenced-column-name="TITLE"/>
            </join-table>
        </one-to-many>
        ...
    </attributes>
</entity>
<entity class="org.mag.Article">
    <table name="ART"/>
        <attributes>
            <id name="id"/>
            <many-to-many name="articles">
                <order-by>lastName, firstName</order-by>
                <join-table name="ART_AUTHS">
                    <join-column name="ART_ID" referenced-column-name="ID"/>
                    <inverse-join-column name="AUTH_ID" referenced-column-name="AID"/>
                </join-table>
            </many-to-many>
            ...
       </attributes>
</entity>
<entity class="org.mag.pub.Author">
    <table name="AUTH"/>
    <attributes>
        <id name="id">
            <column name="AID" column-definition="INTEGER64"/>
        </id>
        ...
    </attributes>
</entity>

8.6.  Bidirectional Mapping

Section 2.9.1, “ Bidirectional Relations ” introduced bidirectional relations. To map a bidirectional relation, you map one field normally using the annotations we have covered throughout this chapter. Then you use the mappedBy property of the other field's metadata annotation or the corresponding mapped-by XML attribute to refer to the mapped field. Look for this pattern in these bidirectional relations as you peruse the complete mappings below:

  • Magazine.publisher and Company.ags.

  • Article.authors and Author.articles.

8.7.  Map Mapping

All map fields in JPA are modeled on either one to many or many to many associations. The map key is always derived from an associated entity's field. Thus map fields use the same mappings as any one to many or many to many fields, namely dedicated join tables or bidirectional relations. The only additions are the MapKey annotation and map-key element to declare the key field. We covered these additions in in Section 2.13, “ Map Key ”.

The example below maps Subscription's map of LineItems to the SUB_ITEMS join table. The key for each map entry is the LineItem's num field value.

Example 12.16.  Join Table Map Mapping

package org.mag.subscribe;

@Entity
@Table(name="SUB", schema="CNTRCT")
public class Subscription {

    @OneToMany(cascade={CascadeType.PERSIST,CascadeType.REMOVE})
    @MapKey(name="num")
    @JoinTable(name="SUB_ITEMS", schema="CNTRCT",
        joinColumns=@JoinColumn(name="SUB_ID"),
        inverseJoinColumns=@JoinColumn(name="ITEM_ID"))
    private Map<Long,LineItem> items;

    ...

    @Entity
    @Table(name="LINE_ITEM", schema="CNTRCT")
    public static class LineItem
        extends Contract {

        private long num;

        ...
    }
}

The same metadata expressed in XML:

<entity class="org.mag.subscribe.Subscription">
    <table name="SUB" schema="CNTRCT"/>
    <attributes>
        ...
        <one-to-many name="items">
            <map-key name="num">
            <join-table name="MAG_ARTS">
                <join-column name="MAG_ISBN" referenced-column-name="ISBN"/>
                <join-column name="MAG_TITLE" referenced-column-name="TITLE"/>
            </join-table>
            <cascade>
                <cascade-persist/>
                <cascade-remove/>
            </cascade>
        </one-to-many>
        ...
    </attributes>
</entity>
<entity class="org.mag.subscribe.Subscription.LineItem">
    <table name="LINE_ITEM" schema="CNTRCT"/>
    <attributes>
        ...
        <basic name="num"/>
        ...
    </attributes>
</entity>

9.  The Complete Mappings

We began this chapter with the goal of mapping the following object model:

That goal has now been met. In the course of explaining JPA's object-relational mapping metadata, we slowly built the requisite schema and mappings for the complete model. First, the database schema:

And finally, the complete entity mappings. We have trimmed the mappings to take advantage of JPA defaults where possible.

Example 12.17.  Full Entity Mappings

package org.mag;

@Entity
@IdClass(Magazine.MagazineId.class)
@Table(name="MAG")
@DiscriminatorValue("Mag")
public class Magazine {

    @Column(length=9)
    @Id private String isbn;
    @Id private String title;

    @Column(name="VERS")
    @Version private int version;
    
    private String name;
    private double price;

    @Column(name="COPIES")
    private int copiesSold;

    @OneToOne(fetch=FetchType.LAZY, 
        cascade={CascadeType.PERSIST,CascadeType.REMOVE})
    @JoinColumn(name="COVER_ID")
    private Article coverArticle;

    @OneToMany(cascade={CascadeType.PERSIST,CascadeType.REMOVE})
    @OrderBy
    @JoinTable(name="MAG_ARTS",
        joinColumns={
            @JoinColumn(name="MAG_ISBN", referencedColumnName="ISBN"),
            @JoinColumn(name="MAG_TITLE", referencedColumnName="TITLE")
        },
        inverseJoinColumns=@JoinColumn(name="ART_ID"))
    private Collection<Article> articles;

    @ManyToOne(fetch=FetchType.LAZY, cascade=CascadeType.PERSIST)
    @JoinColumn(name="PUB_ID")
    private Company publisher; 

    @Transient private byte[] data;


    ...

    public static class MagazineId {
        ...
    }
}

@Entity
@Table(name="ART", uniqueConstraints=@Unique(columnNames="TITLE"))
@SequenceGenerator(name="ArticleSeq", sequenceName="ART_SEQ")
public class Article {

    @Id
    @GeneratedValue(strategy=GenerationType.SEQUENCE, generator="ArticleSeq") 
    private long id;

    @Column(name="VERS")
    @Version private int version;

    private String title;
    private byte[] content;

    @ManyToMany(cascade=CascadeType.PERSIST)
    @OrderBy("lastName, firstName")
    @JoinTable(name="ART_AUTHS",
        joinColumns=@JoinColumn(name="ART_ID"),
        inverseJoinColumns=@JoinColumn(name="AUTH_ID"))
    private Collection<Author> authors;

    ...
}


package org.mag.pub;

@Entity
@Table(name="COMP")
public class Company {

    @Column(name="CID")
    @Id private long id;

    @Column(name="VERS")
    @Version private int version;

    private String name;

    @Column(name="REV")
    private double revenue;

    @Embedded
    @AttributeOverrides({
        @AttributeOverride(name="street", column=@Column(name="STRT")),
        @AttributeOverride(name="city", column=@Column(name="ACITY"))
    })
    private Address address;
    
    @OneToMany(mappedBy="publisher", cascade=CascadeType.PERSIST)
    private Collection<Magazine> mags;

    @OneToMany(cascade=CascadeType.PERSIST,CascadeType.REMOVE)
    @JoinTable(name="COMP_SUBS",
        joinColumns=@JoinColumn(name="COMP_ID"),
        inverseJoinColumns=@JoinColumn(name="SUB_ID"))
    private Collection<Subscription> subscriptions;

    ...
}

@Entity
@Table(name="AUTH")
public class Author {

    @Id
    @GeneratedValue(strategy=GenerationType.TABLE, generator="AuthorGen")
    @TableGenerator(name="AuthorGen", tableName="AUTH_GEN", pkColumnName="PK",
        valueColumnName="AID")
    @Column(name="AID", columnDefinition="INTEGER64")
    private long id;

    @Column(name="VERS")
    @Version private int version;

    @Column(name="FNAME")
    private String firstName;

    @Column(name="LNAME")
    private String lastName;

    private Address address;
    
    @ManyToMany(mappedBy="authors", cascade=CascadeType.PERSIST)
    private Collection<Article> arts;

    ...
}

@Embeddable
public class Address {

    private String street;
    private String city;
    @Column(columnDefinition="CHAR(2)")
    private String state;
    private String zip;
}


package org.mag.subscribe;

@MappedSuperclass
public abstract class Document {

    @Id
    @GeneratedValue(strategy=GenerationType.IDENTITY)
    private long id;

    @Column(name="VERS")
    @Version private int version;

    ...
}

@Entity
@Table(schema="CNTRCT")
@Inheritance(strategy=InheritanceType.JOINED)
@DiscriminatorColumn(name="CTYPE")
public class Contract
    extends Document {

    @Lob
    private String terms;

    ...
}

@Entity
@Table(name="SUB", schema="CNTRCT")
@DiscriminatorColumn(name="KIND", discriminatorType=DiscriminatorType.INTEGER)
@DiscriminatorValue("1")
public class Subscription {

    @Id
    @GeneratedValue(strategy=GenerationType.IDENTITY)
    private long id;

    @Column(name="VERS")
    @Version private int version;

    @Column(name="START")
    private Date startDate;

    @Column(name="PAY")
    private double payment;

    @OneToMany(cascade={CascadeType.PERSIST,CascadeType.REMOVE})
    @MapKey(name="num")
    @JoinTable(name="SUB_ITEMS", schema="CNTRCT",
        joinColumns=@JoinColumn(name="SUB_ID"),
        inverseJoinColumns=@JoinColumn(name="ITEM_ID"))
    private Map<Long,LineItem> items;

    ...

    @Entity
    @Table(name="LINE_ITEM", schema="CNTRCT")
    public static class LineItem
        extends Contract {

        @Column(name="COMM")
        private String comments;

        private double price;
        private long num;

        @ManyToOne
        @JoinColumns({
             @JoinColumn(name="MAG_ISBN", referencedColumnName="ISBN"),
             @JoinColumn(name="MAG_TITLE", referencedColumnName="TITLE")
        })
        private Magazine magazine;
        ...
    }
}

@Entity(name="Lifetime")
@DiscriminatorValue("2")
public class LifetimeSubscription
    extends Subscription {

    @Basic(fetch=FetchType.LAZY)
    @Column(name="ELITE")
    private boolean getEliteClub () { ... }
    public void setEliteClub (boolean elite) { ... }

    ...
}

@Entity(name="Trial")
@DiscriminatorValue("3")
public class TrialSubscription
    extends Subscription {

    @Column(name="END")
    public Date getEndDate () { ... }
    public void setEndDate (Date end) { ... }

    ...
}

The same metadata expressed in XML form:

<entity-mappings xmlns="http://java.sun.com/xml/ns/persistence/orm" 
    xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" 
    xsi:schemaLocation="http://java.sun.com/xml/ns/persistence/orm orm_1_0.xsd"
    version="1.0">
    <mapped-superclass class="org.mag.subscribe.Document">
        <attributes>
            <id name="id">
                <generated-value strategy="IDENTITY"/>
            </id>
            <version name="version">
                <column name="VERS"/>
            </version>
        </attributes>
    </mapped-superclass>
    <entity class="org.mag.Magazine">
        <table name="MAG"/>
        <id-class="org.mag.Magazine.MagazineId"/>
        <discriminator-value>Mag</discriminator-value>
        <attributes>
            <id name="isbn">
                <column length="9"/>
            </id>
            <id name="title"/>
            <basic name="name"/>
            <basic name="price"/>
            <basic name="copiesSold">
                <column name="COPIES"/>
            </basic>
            <version name="version">
                <column name="VERS"/>
            </version>
            <many-to-one name="publisher" fetch="LAZY">
                <join-column name="PUB_ID"/>
                <cascade>
                    <cascade-persist/>
                </cascade>
            </many-to-one>
            <one-to-many name="articles">
                <order-by/>
                <join-table name="MAG_ARTS">
                    <join-column name="MAG_ISBN" referenced-column-name="ISBN"/>
                    <join-column name="MAG_TITLE" referenced-column-name="TITLE"/>
                    <inverse-join-column name="ART_ID"/>
                </join-table>
                <cascade>
                    <cascade-persist/>
                    <cascade-remove/>
                </cascade>
            </one-to-many>
            <one-to-one name="coverArticle" fetch="LAZY">
                <join-column name="COVER_ID"/>
                <cascade>
                    <cascade-persist/>
                    <cascade-remove/>
                </cascade>
            </one-to-one>
            <transient name="data"/>
        </attributes>
    </entity>
    <entity class="org.mag.Article">
        <table name="ART">
            <unique-constraint>
               <column-name>TITLE</column-name>
            </unique-constraint>
        </table>
        <sequence-generator name="ArticleSeq", sequenceName="ART_SEQ"/>
        <attributes>
            <id name="id">
                <generated-value strategy="SEQUENCE" generator="ArticleSeq"/>
            </id>
            <basic name="title"/>
            <basic name="content"/>
            <version name="version">
                <column name="VERS"/>
            </version>
            <many-to-many name="articles">
                <order-by>lastName, firstName</order-by>
                <join-table name="ART_AUTHS">
                    <join-column name="ART_ID" referenced-column-name="ID"/>
                    <inverse-join-column name="AUTH_ID" referenced-column-name="AID"/>
                </join-table>
            </many-to-many>
        </attributes>
    </entity>
    <entity class="org.mag.pub.Company">
        <table name="COMP"/>
        <attributes>
            <id name="id">
                <column name="CID"/>
            </id>
            <basic name="name"/>
            <basic name="revenue">
                <column name="REV"/>
            </basic>
            <version name="version">
                <column name="VERS"/>
            </version>
            <one-to-many name="mags" mapped-by="publisher">
                <cascade>
                    <cascade-persist/>
                </cascade>
            </one-to-many>
            <one-to-many name="subscriptions">
                <join-table name="COMP_SUBS">
                    <join-column name="COMP_ID"/>
                    <inverse-join-column name="SUB_ID"/>
                </join-table>
                <cascade>
                    <cascade-persist/>
                    <cascade-remove/>
                </cascade>
            </one-to-many>
            <embedded name="address">
                <attribute-override name="street">
                   <column name="STRT"/>
                </attribute-override>
                <attribute-override name="city">
                   <column name="ACITY"/>
                </attribute-override>
            </embedded>
        </attributes>
    </entity>
    <entity class="org.mag.pub.Author">
        <table name="AUTH"/>
        <attributes>
            <id name="id">
                <column name="AID" column-definition="INTEGER64"/>
                <generated-value strategy="TABLE" generator="AuthorGen"/>
                <table-generator name="AuthorGen" table="AUTH_GEN" 
                    pk-column-name="PK" value-column-name="AID"/>
            </id>
            <basic name="firstName">
                <column name="FNAME"/>
            </basic>
            <basic name="lastName">
                <column name="LNAME"/>
            </basic>
            <version name="version">
                <column name="VERS"/>
            </version>
            <many-to-many name="arts" mapped-by="authors">
                <cascade>
                    <cascade-persist/>
                </cascade>
            </many-to-many>
            <embedded name="address"/>
        </attributes>
    </entity>
    <entity class="org.mag.subcribe.Contract">
        <table schema="CNTRCT"/>
        <inheritance strategy="JOINED"/>
        <discriminator-column name="CTYPE"/>
        <attributes>
            <basic name="terms">
                <lob/>
            </basic>
        </attributes>
    </entity>
    <entity class="org.mag.subcribe.Subscription">
        <table name="SUB" schema="CNTRCT"/>
        <inheritance strategy="SINGLE_TABLE"/>
        <discriminator-value>1</discriminator-value>
        <discriminator-column name="KIND" discriminator-type="INTEGER"/>
        <attributes>
            <id name="id">
                <generated-value strategy="IDENTITY"/>
            </id>
            <basic name="payment">
                <column name="PAY"/>
            </basic>
            <basic name="startDate">
                <column name="START"/>
            </basic>
            <version name="version">
                <column name="VERS"/>
            </version>
            <one-to-many name="items">
                <map-key name="num">
                <join-table name="SUB_ITEMS" schema="CNTRCT">
                    <join-column name="SUB_ID"/>
                    <inverse-join-column name="ITEM_ID"/>
                </join-table>
                <cascade>
                    <cascade-persist/>
                    <cascade-remove/>
                </cascade>
            </one-to-many>
        </attributes>
    </entity>
    <entity class="org.mag.subscribe.Subscription.LineItem">
        <table name="LINE_ITEM" schema="CNTRCT"/>
        <attributes>
            <basic name="comments">
                <column name="COMM"/>
            </basic>
            <basic name="price"/>
            <basic name="num"/>
            <many-to-one name="magazine">
                <join-column name="MAG_ISBN" referenced-column-name="ISBN"/>
                <join-column name="MAG_TITLE" referenced-column-name="TITLE"/>
            </many-to-one>
        </attributes>
    </entity>
    <entity class="org.mag.subscribe.LifetimeSubscription" name="Lifetime">
        <discriminator-value>2</discriminator-value>
        <attributes>
            <basic name="eliteClub" fetch="LAZY">
                <column name="ELITE"/>
            </basic>
        </attributes>
    </entity>
    <entity class="org.mag.subscribe.TrialSubscription" name="Trial">
        <discriminator-value>3</discriminator-value>
        <attributes>
            <basic name="endDate">
                <column name="END"/>
            </basic>
        </attributes>
    </entity>
    <embeddable class="org.mag.pub.Address">
        <attributes>
            <basic name="street"/>
            <basic name="city"/>
            <basic name="state">
                <column column-definition="CHAR(2)"/>
            </basic>
            <basic name="zip"/>
        </attributes>
    </embeddable>
</entity-mappings>

Chapter 13.  Conclusion

This concludes our overview of the JPA specification. The OpenJPA Reference Guide contains detailed documentation on all aspects of the OpenJPA implementation and core development tools.

Reference Guide

Table of Contents

1. Introduction
1. Intended Audience
2. Configuration
1. Introduction
2. Runtime Configuration
3. Command Line Configuration
3.1. Code Formatting
4. Plugin Configuration
5. OpenJPA Properties
5.1. openjpa.AutoClear
5.2. openjpa.AutoDetach
5.3. openjpa.BrokerFactory
5.4. openjpa.BrokerImpl
5.5. openjpa.ClassResolver
5.6. openjpa.Compatibility
5.7. openjpa.ConnectionDriverName
5.8. openjpa.Connection2DriverName
5.9. openjpa.ConnectionFactory
5.10. openjpa.ConnectionFactory2
5.11. openjpa.ConnectionFactoryName
5.12. openjpa.ConnectionFactory2Name
5.13. openjpa.ConnectionFactoryMode
5.14. openjpa.ConnectionFactoryProperties
5.15. openjpa.ConnectionFactory2Properties
5.16. openjpa.ConnectionPassword
5.17. openjpa.Connection2Password
5.18. openjpa.ConnectionProperties
5.19. openjpa.Connection2Properties
5.20. openjpa.ConnectionURL
5.21. openjpa.Connection2URL
5.22. openjpa.ConnectionUserName
5.23. openjpa.Connection2UserName
5.24. openjpa.ConnectionRetainMode
5.25. openjpa.DataCache
5.26. openjpa.DataCacheManager
5.27. openjpa.DataCacheTimeout
5.28. openjpa.DetachState
5.29. openjpa.DynamicDataStructs
5.30. openjpa.FetchBatchSize
5.31. openjpa.FetchGroups
5.32. openjpa.FlushBeforeQueries
5.33. openjpa.IgnoreChanges
5.34. openjpa.Id
5.35. openjpa.InverseManager
5.36. openjpa.LockManager
5.37. openjpa.LockTimeout
5.38. openjpa.Log
5.39. openjpa.ManagedRuntime
5.40. openjpa.Mapping
5.41. openjpa.MaxFetchDepth
5.42. openjpa.MetaDataFactory
5.43. openjpa.Multithreaded
5.44. openjpa.Optimistic
5.45. openjpa.OrphanedKeyAction
5.46. openjpa.NontransactionalRead
5.47. openjpa.NontransactionalWrite
5.48. openjpa.ProxyManager
5.49. openjpa.QueryCache
5.50. openjpa.QueryCompilationCache
5.51. openjpa.ReadLockLevel
5.52. openjpa.RemoteCommitProvider
5.53. openjpa.RestoreState
5.54. openjpa.RetainState
5.55. openjpa.RetryClassRegistration
5.56. openjpa.SavepointManager
5.57. openjpa.Sequence
5.58. openjpa.TransactionMode
5.59. openjpa.WriteLockLevel
6. OpenJPA JDBC Properties
6.1. openjpa.jdbc.ConnectionDecorators
6.2. openjpa.jdbc.DBDictionary
6.3. openjpa.jdbc.DriverDataSource
6.4. openjpa.jdbc.EagerFetchMode
6.5. openjpa.jdbc.FetchDirection
6.6. openjpa.jdbc.JDBCListeners
6.7. openjpa.jdbc.LRSSize
6.8. openjpa.jdbc.MappingDefaults
6.9. openjpa.jdbc.MappingFactory
6.10. openjpa.jdbc.ResultSetType
6.11. openjpa.jdbc.Schema
6.12. openjpa.jdbc.SchemaFactory
6.13. openjpa.jdbc.Schemas
6.14. openjpa.jdbc.SQLFactory
6.15. openjpa.jdbc.SubclassFetchMode
6.16. openjpa.jdbc.SynchronizeMappings
6.17. openjpa.jdbc.TransactionIsolation
6.18. openjpa.jdbc.UpdateManager
3. Logging
1. Logging Channels
2. OpenJPA Logging
3. Disabling Logging
4. Log4J
5. Apache Commons Logging
5.1. JDK 1.4 java.util.logging
6. Custom Log
4. JDBC
1. Using the OpenJPA DataSource
2. Using a Third-Party DataSource
2.1. Managed and XA DataSources
3. Runtime Access to DataSource
4. Database Support
4.1. DBDictionary Properties
4.2. MySQLDictionary Properties
4.3. OracleDictionary Properties
5. Setting the Transaction Isolation
6. Setting the SQL Join Syntax
7. Accessing Multiple Databases
8. Configuring the Use of JDBC Connections
9. Large Result Sets
10. Default Schema
11. Schema Reflection
11.1. Schemas List
11.2. Schema Factory
12. Schema Tool
13. XML Schema Format
5. Persistent Classes
1. Persistent Class List
2. Enhancement
2.1. Enhancing at Build Time
2.2. Enhancing JPA Entities on Deployment
2.3. Enhancing at Runtime
2.4. Serializing Enhanced Types
3. Object Identity
3.1. Datastore Identity
3.2. Entities as Identity Fields
3.3. Application Identity Tool
3.4. Autoassign / Identity Strategy Caveats
4. Managed Inverses
5. Persistent Fields
5.1. Restoring State
5.2. Typing and Ordering
5.3. Calendar Fields and TimeZones
5.4. Proxies
5.4.1. Smart Proxies
5.4.2. Large Result Set Proxies
5.4.3. Custom Proxies
5.5. Externalization
5.5.1. External Values
6. Fetch Groups
6.1. Custom Fetch Groups
6.2. Custom Fetch Group Configuration
6.3. Per-field Fetch Configuration
6.4. Implementation Notes
7. Eager Fetching
7.1. Configuring Eager Fetching
7.2. Eager Fetching Considerations and Limitations
6. Metadata
1. Metadata Factory
2. Additional JPA Metadata
2.1. Datastore Identity
2.2. Surrogate Version
2.3. Persistent Field Values
2.4. Persistent Collection Fields
2.5. Persistent Map Fields
3. Metadata Extensions
3.1. Class Extensions
3.1.1. Fetch Groups
3.1.2. Data Cache
3.1.3. Detached State
3.2. Field Extensions
3.2.1. Dependent
3.2.2. Load Fetch Group
3.2.3. LRS
3.2.4. Inverse-Logical
3.2.5. Read-Only
3.2.6. Type
3.2.7. Externalizer
3.2.8. Factory
3.2.9. External Values
3.3. Example
7. Mapping
1. Forward Mapping
1.1. Using the Mapping Tool
1.2. Generating DDL SQL
1.3. Runtime Forward Mapping
2. Reverse Mapping
2.1. Customizing Reverse Mapping
3. Meet-in-the-Middle Mapping
4. Mapping Defaults
5. Mapping Factory
6. Non-Standard Joins
7. Additional JPA Mappings
7.1. Datastore Identity Mapping
7.2. Surrogate Version Mapping
7.3. Multi-Column Mappings
7.4. Join Column Attribute Targets
7.5. Embedded Mapping
7.6. Collections
7.6.1. Container Table
7.6.2. Element Join Columns
7.6.3. Order Column
7.7. One-Sided One-Many Mapping
7.8. Maps
7.9. Indexes and Constraints
7.9.1. Indexes
7.9.2. Foreign Keys
7.9.3. Unique Constraints
8. Mapping Limitations
8.1. Table Per Class
9. Mapping Extensions
9.1. Class Extensions
9.1.1. Subclass Fetch Mode
9.1.2. Strategy
9.1.3. Discriminator Strategy
9.1.4. Version Strategy
9.2. Field Extensions
9.2.1. Eager Fetch Mode
9.2.2. Nonpolymorphic
9.2.3. Class Criteria
9.2.4. Strategy
10. Custom Mappings
10.1. Custom Class Mapping
10.2. Custom Discriminator and Version Strategies
10.3. Custom Field Mapping
10.3.1. Value Handlers
10.3.2. Field Strategies
10.3.3. Configuration
11. Orphaned Keys
8. Deployment
1. Factory Deployment
1.1. Standalone Deployment
1.2. EntityManager Injection
2. Integrating with the Transaction Manager
3. XA Transactions
3.1. Using OpenJPA with XA Transactions
9. Runtime Extensions
1. Architecture
1.1. Broker Finalization
1.2. Broker Customization and Finalization
2. JPA Extensions
2.1. OpenJPAEntityManagerFactory
2.2. OpenJPAEntityManager
2.3. OpenJPAQuery
2.4. Extent
2.5. StoreCache
2.6. QueryResultCache
2.7. FetchPlan
2.8. OpenJPAPersistence
3. Object Locking
3.1. Configuring Default Locking
3.2. Configuring Lock Levels at Runtime
3.3. Object Locking APIs
3.4. Lock Manager
3.5. Rules for Locking Behavior
3.6. Known Issues and Limitations
4. Savepoints
4.1. Using Savepoints
4.2. Configuring Savepoints
5. MethodQL
6. Generators
6.1. Runtime Access
7. Transaction Events
8. Non-Relational Stores
10. Caching
1. Data Cache
1.1. Data Cache Configuration
1.2. Data Cache Usage
1.3. Query Cache
1.4. Cache Extension
1.5. Important Notes
1.6. Known Issues and Limitations
2. Query Compilation Cache
11. Remote and Offline Operation
1. Detach and Attach
1.1. Detach Behavior
1.2. Attach Behavior
1.3. Defining the Detached Object Graph
1.3.1. Detached State Field
2. Remote Event Notification Framework
2.1. Remote Commit Provider Configuration
2.1.1. JMS
2.1.2. TCP
2.1.3. Common Properties
2.2. Customization
12. Third Party Integration
1. Apache Ant
1.1. Common Ant Configuration Options
1.2. Enhancer Ant Task
1.3. Application Identity Tool Ant Task
1.4. Mapping Tool Ant Task
1.5. Reverse Mapping Tool Ant Task
1.6. Schema Tool Ant Task
13. Optimization Guidelines

Chapter 1.  Introduction

Table of Contents

1. Intended Audience

OpenJPA is a JDBC-based implementation of the JPA standard. This document is a reference for the configuration and use of OpenJPA.

1.  Intended Audience

This document is intended for OpenJPA developers. It assumes strong knowledge of Java, familiarity with the eXtensible Markup Language (XML), and an understanding of JPA. If you are not familiar with JPA, please read the JPA Overview before proceeding.

Certain sections of this guide cover advanced topics such as custom object-relational mapping, enterprise integration, and using OpenJPA with third-party tools. These sections assume prior experience with the relevant subject.

Chapter 2.  Configuration

Table of Contents

1. Introduction
2. Runtime Configuration
3. Command Line Configuration
3.1. Code Formatting
4. Plugin Configuration
5. OpenJPA Properties
5.1. openjpa.AutoClear
5.2. openjpa.AutoDetach
5.3. openjpa.BrokerFactory
5.4. openjpa.BrokerImpl
5.5. openjpa.ClassResolver
5.6. openjpa.Compatibility
5.7. openjpa.ConnectionDriverName
5.8. openjpa.Connection2DriverName
5.9. openjpa.ConnectionFactory
5.10. openjpa.ConnectionFactory2
5.11. openjpa.ConnectionFactoryName
5.12. openjpa.ConnectionFactory2Name
5.13. openjpa.ConnectionFactoryMode
5.14. openjpa.ConnectionFactoryProperties
5.15. openjpa.ConnectionFactory2Properties
5.16. openjpa.ConnectionPassword
5.17. openjpa.Connection2Password
5.18. openjpa.ConnectionProperties
5.19. openjpa.Connection2Properties
5.20. openjpa.ConnectionURL
5.21. openjpa.Connection2URL
5.22. openjpa.ConnectionUserName
5.23. openjpa.Connection2UserName
5.24. openjpa.ConnectionRetainMode
5.25. openjpa.DataCache
5.26. openjpa.DataCacheManager
5.27. openjpa.DataCacheTimeout
5.28. openjpa.DetachState
5.29. openjpa.DynamicDataStructs
5.30. openjpa.FetchBatchSize
5.31. openjpa.FetchGroups
5.32. openjpa.FlushBeforeQueries
5.33. openjpa.IgnoreChanges
5.34. openjpa.Id
5.35. openjpa.InverseManager
5.36. openjpa.LockManager
5.37. openjpa.LockTimeout
5.38. openjpa.Log
5.39. openjpa.ManagedRuntime
5.40. openjpa.Mapping
5.41. openjpa.MaxFetchDepth
5.42. openjpa.MetaDataFactory
5.43. openjpa.Multithreaded
5.44. openjpa.Optimistic
5.45. openjpa.OrphanedKeyAction
5.46. openjpa.NontransactionalRead
5.47. openjpa.NontransactionalWrite
5.48. openjpa.ProxyManager
5.49. openjpa.QueryCache
5.50. openjpa.QueryCompilationCache
5.51. openjpa.ReadLockLevel
5.52. openjpa.RemoteCommitProvider
5.53. openjpa.RestoreState
5.54. openjpa.RetainState
5.55. openjpa.RetryClassRegistration
5.56. openjpa.SavepointManager
5.57. openjpa.Sequence
5.58. openjpa.TransactionMode
5.59. openjpa.WriteLockLevel
6. OpenJPA JDBC Properties
6.1. openjpa.jdbc.ConnectionDecorators
6.2. openjpa.jdbc.DBDictionary
6.3. openjpa.jdbc.DriverDataSource
6.4. openjpa.jdbc.EagerFetchMode
6.5. openjpa.jdbc.FetchDirection
6.6. openjpa.jdbc.JDBCListeners
6.7. openjpa.jdbc.LRSSize
6.8. openjpa.jdbc.MappingDefaults
6.9. openjpa.jdbc.MappingFactory
6.10. openjpa.jdbc.ResultSetType
6.11. openjpa.jdbc.Schema
6.12. openjpa.jdbc.SchemaFactory
6.13. openjpa.jdbc.Schemas
6.14. openjpa.jdbc.SQLFactory
6.15. openjpa.jdbc.SubclassFetchMode
6.16. openjpa.jdbc.SynchronizeMappings
6.17. openjpa.jdbc.TransactionIsolation
6.18. openjpa.jdbc.UpdateManager

1.  Introduction

This chapter describes the OpenJPA configuration framework. It concludes with descriptions of all the configuration properties recognized by OpenJPA. You may want to browse these properties now, but it is not necessary. Most of them will be referenced later in the documentation as we explain the various features they apply to.

2.  Runtime Configuration

The OpenJPA runtime includes a comprehensive system of configuration defaults and overrides:

  • OpenJPA first looks for an optional openjpa.xml resource. OpenJPA searches for this resource in each top-level directory of your CLASSPATH. OpenJPA will also find the resource if you place it within a META-INF directory in any top-level directory of the CLASSPATH. The openjpa.xml resource contains property settings in JPA's XML format.

  • You can customize the name or location of the above resource by specifying the correct resource path in the openjpa.properties System property.

  • You can override any value defined in the above resource by setting the System property of the same name to the desired value.

  • In JPA, the values in the standard META-INF/persistence.xml bootstrapping file used by the Persistence class at runtime override the values in the above resource, as well as any System property settings. The Map passed to Persistence.createEntityManagerFactory at runtime also overrides previous settings, including properties defined in persistence.xml.

  • When using JCA deployment the config-property values in your ra.xml file override other settings.

  • All OpenJPA command-line tools accept flags that allow you to specify the configuration resource to use, and to override any property. Section 3, “ Command Line Configuration ” describes these flags.

Note

Internally, the OpenJPA runtime environment and development tools manipulate property settings through a general Configuration interface, and in particular its OpenJPAConfiguration and JDBCConfiguration subclasses. For advanced customization, OpenJPA's extended runtime interfaces and its development tools allow you to access these interfaces directly. See the Javadoc for details.

3.  Command Line Configuration

OpenJPA development tools share the same set of configuration defaults and overrides as the runtime system. They also allow you to specify property values on the command line:

  • -properties/-p <configuration file or resource>: Use the -properties flag, or its shorter -p form, to specify a configuration file to use. Note that OpenJPA always searches the default file locations described above, so this flag is only needed when you do not have a default resource in place, or when you wish to override the defaults. The given value can be the path to a file, or the resource name of a file somewhere in the CLASSPATH. OpenJPA will search the given location as well as the location prefixed by META-INF/ . Thus, to point a OpenJPA tool at META-INF/my-persistence.xml, you can use:

    <tool> -p my-persistence.xml
    
  • -<property name> <property value>: Any configuration property that you can specify in a configuration file can be overridden with a command line flag. The flag name is always the last token of the corresponding property name, with the first letter in either upper or lower case. For example, to override the openjpa.ConnectionUserName property, you could pass the -connectionUserName <value> flag to any tool. Values set this way override both the values in the configuration file and values set via System properties.

3.1.  Code Formatting

Some OpenJPA development tools generate Java code. These tools share a common set of command-line flags for formatting their output to match your coding style. All code formatting flags can begin with either the codeFormat or cf prefix.

  • -codeFormat./-cf.tabSpaces <spaces>: The number of spaces that make up a tab, or 0 to use tab characters. Defaults to using tab characters.

  • -codeFormat./-cf.spaceBeforeParen <true/t | false/f>: Whether or not to place a space before opening parentheses on method calls, if statements, loops, etc. Defaults to false.

  • -codeFormat./-cf.spaceInParen <true/t | false/f>: Whether or not to place a space within parentheses; i.e. method( arg) . Defaults to false.

  • -codeFormat./-cf.braceOnSameLine <true/t | false/f>: Whether or not to place opening braces on the same line as the declaration that begins the code block, or on the next line. Defaults to true .

  • -codeFormat./-cf.braceAtSameTabLevel <true/t | false/f> : When the braceOnSameLine option is disabled, you can choose whether to place the brace at the same tab level of the contained code. Defaults to false.

  • -codeFormat./-cf.scoreBeforeFieldName <true/t | false/f> : Whether to prefix an underscore to names of private member variables. Defaults to false.

  • -codeFormat./-cf.linesBetweenSections <lines>: The number of lines to skip between sections of code. Defaults to 1.

Example 2.1.  Code Formatting with the Application Id Tool

java org.apache.openjpa.enhance.ApplicationIdTool -cf.spaceBeforeParen true -cf.tabSpaces 4

4.  Plugin Configuration

Because OpenJPA is a highly customizable environment, many configuration properties relate to the creation and configuration of system plugins. Plugin properties have a syntax very similar to that of Java 5 annotations. They allow you to specify both what class to use for the plugin and how to configure the public fields or bean properties of the instantiated plugin instance. The easiest way to describe the plugin syntax is by example:

OpenJPA has a pluggable L2 caching mechanism that is controlled by the openjpa.DataCache configuration property. Suppose that you have created a new class, com.xyz.MyDataCache, that you want OpenJPA to use for caching. You've made instances of MyDataCache configurable via two methods, setCacheSize(int size) and setRemoteHost(String host). The sample below shows how you would tell OpenJPA to use an instance of your custom plugin with a max size of 1000 and a remote host of cacheserver .

<property name="openjpa.DataCache" 
    value="com.xyz.MyDataCache(CacheSize=1000, RemoteHost=cacheserver)"/>

As you can see, plugin properties take a class name, followed by a comma-separated list of values for the plugin's public fields or bean properties in parentheses. OpenJPA will match each named property to a field or setter method in the instantiated plugin instance, and set the field or invoke the method with the given value (after converting the value to the right type, of course). The first letter of the property names can be in either upper or lower case. The following would also have been valid:

com.xyz.MyDataCache(cacheSize=1000, remoteHost=cacheserver)

If you do not need to pass any property settings to a plugin, you can just name the class to use:

com.xyz.MyDataCache

Similarly, if the plugin has a default class that you do not want to change, you can simply specify a list of property settings, without a class name. For example, OpenJPA's query cache companion to the data cache has a default implementation suitable to most users, but you still might want to change the query cache's size. It has a CacheSize property for this purpose:

CacheSize=1000

Finally, many of OpenJPA's built-in options for plugins have short alias names that you can use in place of the full class name. The data cache property, for example, has an available alias of true for the standard cache implementation. The property value simply becomes:

true

The standard cache implementation class also has a CacheSize property, so to use the standard implementation and configure the size, specify:

true(CacheSize=1000)

The remainder of this chapter reviews the set of configuration properties OpenJPA recognizes.

5.  OpenJPA Properties

5.1. openjpa.AutoClear
5.2. openjpa.AutoDetach
5.3. openjpa.BrokerFactory
5.4. openjpa.BrokerImpl
5.5. openjpa.ClassResolver
5.6. openjpa.Compatibility
5.7. openjpa.ConnectionDriverName
5.8. openjpa.Connection2DriverName
5.9. openjpa.ConnectionFactory
5.10. openjpa.ConnectionFactory2
5.11. openjpa.ConnectionFactoryName
5.12. openjpa.ConnectionFactory2Name
5.13. openjpa.ConnectionFactoryMode
5.14. openjpa.ConnectionFactoryProperties
5.15. openjpa.ConnectionFactory2Properties
5.16. openjpa.ConnectionPassword
5.17. openjpa.Connection2Password
5.18. openjpa.ConnectionProperties
5.19. openjpa.Connection2Properties
5.20. openjpa.ConnectionURL
5.21. openjpa.Connection2URL
5.22. openjpa.ConnectionUserName
5.23. openjpa.Connection2UserName
5.24. openjpa.ConnectionRetainMode
5.25. openjpa.DataCache
5.26. openjpa.DataCacheManager
5.27. openjpa.DataCacheTimeout
5.28. openjpa.DetachState
5.29. openjpa.DynamicDataStructs
5.30. openjpa.FetchBatchSize
5.31. openjpa.FetchGroups
5.32. openjpa.FlushBeforeQueries
5.33. openjpa.IgnoreChanges
5.34. openjpa.Id
5.35. openjpa.InverseManager
5.36. openjpa.LockManager
5.37. openjpa.LockTimeout
5.38. openjpa.Log
5.39. openjpa.ManagedRuntime
5.40. openjpa.Mapping
5.41. openjpa.MaxFetchDepth
5.42. openjpa.MetaDataFactory
5.43. openjpa.Multithreaded
5.44. openjpa.Optimistic
5.45. openjpa.OrphanedKeyAction
5.46. openjpa.NontransactionalRead
5.47. openjpa.NontransactionalWrite
5.48. openjpa.ProxyManager
5.49. openjpa.QueryCache
5.50. openjpa.QueryCompilationCache
5.51. openjpa.ReadLockLevel
5.52. openjpa.RemoteCommitProvider
5.53. openjpa.RestoreState
5.54. openjpa.RetainState
5.55. openjpa.RetryClassRegistration
5.56. openjpa.SavepointManager
5.57. openjpa.Sequence
5.58. openjpa.TransactionMode
5.59. openjpa.WriteLockLevel

OpenJPA defines many configuration properties. Most of these properties are provided for advanced users who wish to customize OpenJPA's behavior; the majority of developers can omit them. The following properties apply to any OpenJPA back-end, though the given descriptions are tailored to OpenJPA's default JDBC store.

5.1.  openjpa.AutoClear

Property name: openjpa.AutoClear

Configuration API: org.apache.openjpa.conf.OpenJPAConfiguration.getAutoClear

Resource adaptor config-property: AutoClear

Default: datastore

Possible values: datastore, all

Description: When to automatically clear instance state: on entering a datastore transaction, or on entering any transaction.

5.2.  openjpa.AutoDetach

Property name: openjpa.AutoDetach

Configuration API: org.apache.openjpa.conf.OpenJPAConfiguration.getAutoDetach

Resource adaptor config-property: AutoDetach

Default: -

Possible values: close, commit, nontx-read

Description: A comma-separated list of events when managed instances will be automatically detached.

5.3.  openjpa.BrokerFactory

Property name: openjpa.BrokerFactory

Configuration API: org.apache.openjpa.conf.OpenJPAConfiguration.getBrokerFactory

Resource adaptor config-property: BrokerFactory

Default: jdbc

Possible values: jdbc, abstractstore, remote

Description: A plugin string (see Section 4, “ Plugin Configuration ”) describing the org.apache.openjpa.kernel.BrokerFactory type to use.

5.4.  openjpa.BrokerImpl

Property name: openjpa.BrokerImpl

Configuration API: org.apache.openjpa.conf.OpenJPAConfiguration.getBrokerImpl

Resource adaptor config-property: BrokerImpl

Default: default

Description: A plugin string (see Section 4, “ Plugin Configuration ”) describing the org.apache.openjpa.kernel.Broker type to use at runtime. See Section 1.2, “ Broker Customization and Finalization ” on for details.

5.5.  openjpa.ClassResolver

Property name: openjpa.ClassResolver

Configuration API: org.apache.openjpa.conf.OpenJPAConfiguration.getClassResolver

Resource adaptor config-property: ClassResolver

Default: default

Description: A plugin string (see Section 4, “ Plugin Configuration ”) describing the org.apache.openjpa.util.ClassResolver implementation to use for class name resolution. You may wish to plug in your own resolver if you have special classloading needs.

5.6.  openjpa.Compatibility

Property name: openjpa.Compatibility

Configuration API: org.apache.openjpa.conf.OpenJPAConfiguration.getCompatibility

Resource adaptor config-property: Compatibility

Default: -

Description: Encapsulates options to mimic the behavior of previous OpenJPA releases.

5.7.  openjpa.ConnectionDriverName

Property name: openjpa.ConnectionDriverName

Configuration API: org.apache.openjpa.conf.OpenJPAConfiguration.getConnectionDriverName

Resource adaptor config-property: ConnectionDriverName

Default: -

Description: The full class name of either the JDBC java.sql.Driver, or a javax.sql.DataSource implementation to use to connect to the database. See Chapter 4, JDBC for details.

5.8.  openjpa.Connection2DriverName

Property name: openjpa.Connection2DriverName

Configuration API: org.apache.openjpa.conf.OpenJPAConfiguration.getConnection2DriverName

Resource adaptor config-property: Connection2DriverName

Default: -

Description: This property is equivalent to the openjpa.ConnectionDriverName property described in Section 5.7, “ openjpa.ConnectionDriverName ”, but applies to the alternate connection factory used for unmanaged connections. See Section 2.1, “ Managed and XA DataSources ” for details.

5.9.  openjpa.ConnectionFactory

Property name: openjpa.ConnectionFactory

Configuration API: org.apache.openjpa.conf.OpenJPAConfiguration.getConnectionFactory

Resource adaptor config-property: ConnectionFactory

Default: -

Description: A javax.sql.DataSource to use to connect to the database. See Chapter 4, JDBC for details.

5.10.  openjpa.ConnectionFactory2

Property name: openjpa.ConnectionFactory2

Configuration API: org.apache.openjpa.conf.OpenJPAConfiguration.getConnectionFactory2

Resource adaptor config-property: ConnectionFactory2

Default: -

Description: An unmanaged javax.sql.DataSource to use to connect to the database. See Chapter 4, JDBC for details.

5.11.  openjpa.ConnectionFactoryName

Property name: openjpa.ConnectionFactoryName

Configuration API: org.apache.openjpa.conf.OpenJPAConfiguration.getConnectionFactoryName

Resource adaptor config-property: ConnectionFactoryName

Default: -

Description: The JNDI location of a javax.sql.DataSource to use to connect to the database. See Chapter 4, JDBC for details.

5.12.  openjpa.ConnectionFactory2Name

Property name: openjpa.ConnectionFactory2Name

Configuration API: org.apache.openjpa.conf.OpenJPAConfiguration.getConnectionFactory2Name

Resource adaptor config-property: ConnectionFactory2Name

Default: -

Description: The JNDI location of an unmanaged javax.sql.DataSource to use to connect to the database. See Section 3, “ XA Transactions ” for details.

5.13.  openjpa.ConnectionFactoryMode

Property name: openjpa.ConnectionFactoryMode

Configuration API: org.apache.openjpa.conf.OpenJPAConfiguration.getConnectionFactoryMode

Resource adaptor config-property: ConnectionFactoryMode

Default: local

Possible values: local, managed

Description: The connection factory mode to use when integrating with the application server's managed transactions. See Section 2.1, “ Managed and XA DataSources ” for details.

5.14.  openjpa.ConnectionFactoryProperties

Property name: openjpa.ConnectionFactoryProperties

Configuration API: org.apache.openjpa.conf.OpenJPAConfiguration.getConnectionFactoryProperties

Resource adaptor config-property: ConnectionFactoryProperties

Default: -

Description: A plugin string (see Section 4, “ Plugin Configuration ”) listing properties for configuration of the datasource in use. See the Chapter 4, JDBC for details.

5.15.  openjpa.ConnectionFactory2Properties

Property name: openjpa.ConnectionFactory2Properties

Configuration API: org.apache.openjpa.conf.OpenJPAConfiguration.getConnectionFactory2Properties

Resource adaptor config-property: ConnectionFactory2Properties

Default: -

Description: This property is equivalent to the openjpa.ConnectionFactoryProperties property described in Section 5.14, “ openjpa.ConnectionFactoryProperties ”, but applies to the alternate connection factory used for unmanaged connections. See Section 2.1, “ Managed and XA DataSources ” for details.

5.16.  openjpa.ConnectionPassword

Property name: openjpa.ConnectionPassword

Configuration API: org.apache.openjpa.conf.OpenJPAConfiguration.getConnectionPassword

Resource adaptor config-property: ConnectionPassword

Default: -

Description: The password for the user specified in the ConnectionUserName property. See Chapter 4, JDBC for details.

5.17.  openjpa.Connection2Password

Property name: openjpa.Connection2Password

Configuration API: org.apache.openjpa.conf.OpenJPAConfiguration.getConnection2Password

Resource adaptor config-property: Connection2Password

Default: -

Description: This property is equivalent to the openjpa.ConnectionPassword property described in Section 5.16, “ openjpa.ConnectionPassword ”, but applies to the alternate connection factory used for unmanaged connections. See Section 2.1, “ Managed and XA DataSources ” for details.

5.18.  openjpa.ConnectionProperties

Property name: openjpa.ConnectionProperties

Configuration API: org.apache.openjpa.conf.OpenJPAConfiguration.getConnectionProperties

Resource adaptor config-property: ConnectionProperties

Default: -

Description: A plugin string (see Section 4, “ Plugin Configuration ”) listing properties to configure the driver listed in the ConnectionDriverName property described below. See Chapter 4, JDBC for details.

5.19.  openjpa.Connection2Properties

Property name: openjpa.Connection2Properties

Configuration API: org.apache.openjpa.conf.OpenJPAConfiguration.getConnection2Properties

Resource adaptor config-property: Connection2Properties

Default: -

Description: This property is equivalent to the openjpa.ConnectionProperties property described in Section 5.18, “ openjpa.ConnectionProperties ”, but applies to the alternate connection factory used for unmanaged connections. See Section 2.1, “ Managed and XA DataSources ” for details.

5.20.  openjpa.ConnectionURL

Property name: openjpa.ConnectionURL

Configuration API: org.apache.openjpa.conf.OpenJPAConfiguration.getConnectionURL

Resource adaptor config-property: ConnectionURL

Default: -

Description: The JDBC URL for the database. See Chapter 4, JDBC for details.

5.21.  openjpa.Connection2URL

Property name: openjpa.Connection2URL

Configuration API: org.apache.openjpa.conf.OpenJPAConfiguration.getConnection2URL

Resource adaptor config-property: Connection2URL

Default: -

Description: This property is equivalent to the openjpa.ConnectionURL property described in Section 5.20, “ openjpa.ConnectionURL ”, but applies to the alternate connection factory used for unmanaged connections. See Section 2.1, “ Managed and XA DataSources ” for details.

5.22.  openjpa.ConnectionUserName

Property name: openjpa.ConnectionUserName

Configuration API: org.apache.openjpa.conf.OpenJPAConfiguration.getConnectionUserName

Resource adaptor config-property: ConnectionUserName

Default: -

Description: The user name to use when connecting to the database. See the Chapter 4, JDBC for details.

5.23.  openjpa.Connection2UserName

Property name: openjpa.Connection2UserName

Configuration API: org.apache.openjpa.conf.OpenJPAConfiguration.getConnection2UserName

Resource adaptor config-property: Connection2UserName

Default: -

Description: This property is equivalent to the openjpa.ConnectionUserName property described in Section 5.22, “ openjpa.ConnectionUserName ”, but applies to the alternate connection factory used for unmanaged connections. See Section 2.1, “ Managed and XA DataSources ” for details.

5.24.  openjpa.ConnectionRetainMode

Property name: openjpa.ConnectionRetainMode

Configuration API: org.apache.openjpa.conf.OpenJPAConfiguration.getConnectionRetainMode

Resource adaptor config-property: ConnectionRetainMode

Default: on-demand

Description: Controls how OpenJPA uses datastore connections. This property can also be specified for individual sessions. See Section 8, “ Configuring the Use of JDBC Connections ” for details.

5.25.  openjpa.DataCache

Property name: openjpa.DataCache

Configuration API: org.apache.openjpa.conf.OpenJPAConfiguration.getDataCache

Resource adaptor config-property: DataCache

Default: false

Description: A plugin list string (see Section 4, “ Plugin Configuration ”) describing the org.apache.openjpa.datacache.DataCaches to use for data caching. See Section 1.1, “ Data Cache Configuration ” for details.

5.26.  openjpa.DataCacheManager

Property name: openjpa.DataCacheManager

Configuration API: org.apache.openjpa.conf.OpenJPAConfiguration.getDataCacheManager

Resource adaptor config-property: DataCacheManager

Default: default

Description: A plugin string (see Section 4, “ Plugin Configuration ”) describing the openjpa.datacache.DataCacheManager that manages the system data caches. See Section 1, “ Data Cache ” for details on data caching.

5.27.  openjpa.DataCacheTimeout

Property name: openjpa.DataCacheTimeout

Configuration API: org.apache.openjpa.conf.OpenJPAConfiguration.getDataCacheTimeout

Resource adaptor config-property: DataCacheTimeout

Default: -1

Description: The number of milliseconds that data in the data cache is valid. Set this to -1 to indicate that data should not expire from the cache. This property can also be specified for individual classes. See Section 1.1, “ Data Cache Configuration ” for details.

5.28.  openjpa.DetachState

Property name: openjpa.DetachState

Configuration API: org.apache.openjpa.conf.OpenJPAConfiguration.getDetachState

Resource adaptor config-property: DetachState

Default: loaded

Possible values: loaded, fgs, all

Description: Determines which fields are part of the detached graph and related options. For more details, see Section 1.3, “ Defining the Detached Object Graph ”.

5.29.  openjpa.DynamicDataStructs

Property name: openjpa.DynamicDataStructs

Configuration API: org.apache.openjpa.conf.OpenJPAConfiguration.getDynamicDataStructs

Resource adaptor config-property: DynamicDataStructs

Default: false

Description: Whether to dynamically generate customized structs to hold persistent data. Both the OpenJPA data cache and the remote framework rely on data structs to cache and transfer persistent state. With dynamic structs, OpenJPA can customize data storage for each class, eliminating the need to generate primitive wrapper objects. This saves memory and speeds up certain runtime operations. The price is a longer warm-up time for the application - generating and loading custom classes into the JVM takes time. Therefore, only set this property to true if you have a long-running application where the initial cost of class generation is offset by memory and speed optimization over time.

5.30.  openjpa.FetchBatchSize

Property name: openjpa.FetchBatchSize

Configuration API: org.apache.openjpa.conf.OpenJPAConfiguration.getFetchBatchSize

Resource adaptor config-property: FetchBatchSize

Default: -1

Description: The number of rows to fetch at once when scrolling through a result set. The fetch size can also be set at runtime. See Section 9, “ Large Result Sets ” for details.

5.31.  openjpa.FetchGroups

Property name: openjpa.FetchGroups

Configuration API: org.apache.openjpa.conf.OpenJPAConfiguration.getFetchGroups

Resource adaptor config-property: FetchGroups

Default: -

Description: A comma-separated list of fetch group names that are to be loaded when retrieving objects from the datastore. Fetch groups can also be set at runtime. See Section 6, “ Fetch Groups ” for details.

5.32.  openjpa.FlushBeforeQueries

Property name: openjpa.FlushBeforeQueries

Property name: openjpa.FlushBeforeQueries

Configuration API: org.apache.openjpa.conf.OpenJPAConfiguration.getFlushBeforeQueries

Resource adaptor config-property: FlushBeforeQueries

Default: true

Description: Whether or not to flush any changes made in the current transaction to the datastore before executing a query. See Section 8, “ Configuring the Use of JDBC Connections ” for details.

5.33.  openjpa.IgnoreChanges

Property name: openjpa.IgnoreChanges

Configuration API: org.apache.openjpa.conf.OpenJPAConfiguration.getIgnoreChanges

Resource adaptor config-property: IgnoreChanges

Default: false

Description: Whether to consider modifications to persistent objects made in the current transaction when evaluating queries. Setting this to true allows OpenJPA to ignore changes and execute the query directly against the datastore. A value of false forces OpenJPA to consider whether the changes in the current transaction affect the query, and if so to either evaluate the query in-memory or flush before running it against the datastore.

5.34. openjpa.Id

Property name: openjpa.Id

Resource adaptor config-property: Id

Default: none

Description: An environment-specific identifier for this configuration. This might correspond to a JPA persistence-unit name, or to some other more-unique value available in the current environment.

5.35.  openjpa.InverseManager

Property name: openjpa.InverseManager

Configuration API: org.apache.openjpa.conf.OpenJPAConfiguration.getInverseManager

Resource adaptor config-property: InverseManager

Default: false

Possible values: false, true

Description: A plugin string (see Section 4, “ Plugin Configuration ”) describing a org.apache.openjpa.kernel.InverseManager to use for managing bidirectional relations upon a flush. See Section 4, “ Managed Inverses ” for usage documentation.

5.36.  openjpa.LockManager

Property name: openjpa.LockManager

Configuration API: org.apache.openjpa.conf.OpenJPAConfiguration.getLockManager

Resource adaptor config-property: LockManager

Default: version

Possible values: none, sjvm, pessimistic, version

Description: A plugin string (see Section 4, “ Plugin Configuration ”) describing a org.apache.openjpa.kernel.LockManager to use for acquiring locks on persistent instances during transactions. See Section 3.4, “ Lock Manager ” for more information.

5.37.  openjpa.LockTimeout

Property name: openjpa.LockTimeout

Configuration API: org.apache.openjpa.conf.OpenJPAConfiguration.getLockTimeout

Resource adaptor config-property: LockTimeout

Default: -1

Description: The number of milliseconds to wait for an object lock before throwing an exception, or -1 for no limit. See Section 3, “ Object Locking ” for details.

5.38.  openjpa.Log

Property name: openjpa.Log

Configuration API: org.apache.openjpa.conf.OpenJPAConfiguration.getLog

Resource adaptor config-property: Log

Default: true

Possible values: openjpa, commons, log4j, none

Description: A plugin string (see Section 4, “ Plugin Configuration ”) describing a org.apache.openjpa.lib.log.LogFactory to use for logging. For details on logging, see Chapter 3, Logging .

5.39.  openjpa.ManagedRuntime

Property name: openjpa.ManagedRuntime

Configuration API: org.apache.openjpa.conf.OpenJPAConfiguration.getManagedRuntime

Resource adaptor config-property: ManagedRuntime

Default: auto

Description: A plugin string (see Section 4, “ Plugin Configuration ”) describing the org.apache.openjpa.ee.ManagedRuntime implementation to use for obtaining a reference to the TransactionManager in an enterprise environment.

5.40.  openjpa.Mapping

Property name: openjpa.Mapping

Configuration API: org.apache.openjpa.conf.OpenJPAConfiguration.getMapping

Resource adaptor config-property: Mapping

Default: -

Description: The symbolic name of the object-to-datastore mapping to use.

5.41.  openjpa.MaxFetchDepth

Property name: openjpa.MaxFetchDepth

Configuration API: org.apache.openjpa.conf.OpenJPAConfiguration.getMaxFetchDepth

Resource adaptor config-property: MaxFetchDepth

Default: -1

Description: The maximum depth of relations to traverse when eager fetching. Use -1 for no limit. Defaults to no limit. See Section 7, “ Eager Fetching ” for details on eager fetching.

5.42.  openjpa.MetaDataFactory

Property name: openjpa.MetaDataFactory

Configuration API: org.apache.openjpa.conf.OpenJPAConfiguration.getMetaDataFactory

Resource adaptor config-property: MetaDataFactory

Default: jpa

Description: A plugin string (see Section 4, “ Plugin Configuration ”) describing the openjpa.meta.MetaDataFactory to use to store and retrieve metadata for your persistent classes. See Section 1, “ Metadata Factory ” for details.

5.43.  openjpa.Multithreaded

Property name: openjpa.Multithreaded

Configuration API: org.apache.openjpa.conf.OpenJPAConfiguration.getMultithreaded

Resource adaptor config-property: Multithreaded

Default: false

Description: Whether persistent instances and OpenJPA components other than the EntityManagerFactory will be accessed by multiple threads at once.

5.44.  openjpa.Optimistic

Property name: openjpa.Optimistic

Configuration API: org.apache.openjpa.conf.OpenJPAConfiguration.getOptimistic

Resource adaptor config-property: Optimistic

Default: true

Description: Selects between optimistic and pessimistic (datastore) transactional modes.

5.45.  openjpa.OrphanedKeyAction

Property name: openjpa.OrphanedKeyAction

Configuration API: org.apache.openjpa.conf.OpenJPAConfiguration.getOrphanedKeyAction

Resource adaptor config-property: OrphanedKeyAction

Default: log

Possible values: log, exception, none

Description: A plugin string (see Section 4, “ Plugin Configuration ”) describing a org.apache.openjpa.event.OrphanedKeyAction to invoke when OpenJPA discovers an orphaned datastore key. See Section 11, “ Orphaned Keys ” for details.

5.46.  openjpa.NontransactionalRead

Property name: openjpa.NontransactionalRead

Configuration API: org.apache.openjpa.conf.OpenJPAConfiguration.getNontransactionalRead

Resource adaptor config-property: NontransactionalRead

Default: true

Description: Whether the OpenJPA runtime will allow you to read data outside of a transaction.

5.47.  openjpa.NontransactionalWrite

Property name: openjpa.NontransactionalWrite

Configuration API: org.apache.openjpa.conf.OpenJPAConfiguration.getNontransactionalWrite

Resource adaptor config-property: NontransactionalWrite

Default: false

Description: Whether you can modify persistent objects and perform persistence operations outside of a transaction. Changes will take effect on the next transaction.

5.48.  openjpa.ProxyManager

Property name: openjpa.ProxyManager

Configuration API: org.apache.openjpa.conf.OpenJPAConfiguration.getProxyManager

Resource adaptor config-property: ProxyManager

Default: default

Description: A plugin string (see Section 4, “ Plugin Configuration ”) describing a org.apache.openjpa.util.ProxyManager to use for proxying mutable second class objects. See Section 5.4.3, “ Custom Proxies ” for details.

5.49.  openjpa.QueryCache

Property name: openjpa.QueryCache

Configuration API: org.apache.openjpa.conf.OpenJPAConfiguration.getQueryCache

Resource adaptor config-property: QueryCache

Default: true, when the data cache (see Section 5.25, “ openjpa.DataCache ”) is also enabled, false otherwise.

Description: A plugin string (see Section 4, “ Plugin Configuration ”) describing the org.apache.openjpa.datacache.QueryCache implementation to use for caching of queries loaded from the data store. See Section 1.3, “ Query Cache ” for details.

5.50.  openjpa.QueryCompilationCache

Property name: openjpa.QueryCompilationCache

Resource adaptor config-property: QueryCompilationCache

Default: true.

Description: A plugin string (see Section 4, “ Plugin Configuration ”) describing the java.util.Map to use for caching of data used during query compilation. See Section 2, “ Query Compilation Cache ” for details.

5.51.  openjpa.ReadLockLevel

Property name: openjpa.ReadLockLevel

Configuration API: org.apache.openjpa.conf.OpenJPAConfiguration.getReadLockLevel

Resource adaptor config-property: ReadLockLevel

Default: read

Possible values: none, read, write, numeric values for lock-manager specific lock levels

Description: The default level at which to lock objects retrieved during a non-optimistic transaction. Note that for the default JDBC lock manager, read and write lock levels are equivalent.

5.52.  openjpa.RemoteCommitProvider

Property name: openjpa.RemoteCommitProvider

Configuration API: org.apache.openjpa.conf.OpenJPAConfiguration.getRemoteCommitProvider

Resource adaptor config-property: RemoteCommitProvider

Default: -

Description: A plugin string (see Section 4, “ Plugin Configuration ”) describing the org.apache.openjpa.event.RemoteCommitProvider implementation to use for distributed event notification. See Section 2.1, “ Remote Commit Provider Configuration ” for more information.

5.53.  openjpa.RestoreState

Property name: openjpa.RestoreState

Configuration API: org.apache.openjpa.conf.OpenJPAConfiguration.getRestoreState

Resource adaptor config-property: RestoreState

Default: none

Possible values: none, immutable, all

Description: Whether to restore managed fields to their pre-transaction values when a rollback occurs.

5.54.  openjpa.RetainState

Property name: openjpa.RetainState

Configuration API: org.apache.openjpa.conf.OpenJPAConfiguration.getRetainState

Resource adaptor config-property: RetainState

Default: true

Description: Whether persistent fields retain their values on transaction commit.

5.55.  openjpa.RetryClassRegistration

Property name: openjpa.RetryClassRegistration

Configuration API: org.apache.openjpa.conf.OpenJPAConfiguration.getRetryClassRegistration

Resource adaptor config-property: RetryClassRegistration

Default: false

Description: Controls whether to log a warning and defer registration instead of throwing an exception when a persistent class cannot be fully processed. This property should only be used in complex classloader situations where security is preventing OpenJPA from reading registered classes. Setting this to true unnecessarily may obscure more serious problems.

5.56.  openjpa.SavepointManager

Property name: openjpa.SavepointManager

Configuration API: org.apache.openjpa.conf.OpenJPAConfiguration.getSavepointManager

Resource adaptor config-property: SavepointManager

Default: in-mem

Possible values: in-mem, jdbc, oracle

Description: A plugin string (see Section 4, “ Plugin Configuration ”) describing a org.apache.openjpa.kernel.SavepointManager to use for managing transaction savepoints. See Section 4, “ Savepoints ” for details.

5.57.  openjpa.Sequence

Property name: openjpa.Sequence

Configuration API: org.apache.openjpa.conf.OpenJPAConfiguration.getSequence

Resource adaptor config-property: Sequence

Default: table

Description: A plugin string (see Section 4, “ Plugin Configuration ”) describing the org.apache.openjpa.kernel.Seq implementation to use for the system sequence. See Section 6, “ Generators ” for more information.

5.58.  openjpa.TransactionMode

Property name: openjpa.TransactionMode

Configuration API: org.apache.openjpa.conf.OpenJPAConfiguration.getTransactionMode

Resource adaptor config-property: TransactionMode

Default: local

Possible values: local, managed

Description: The default transaction mode to use. You can override this setting per-session.

5.59.  openjpa.WriteLockLevel

Property name: openjpa.WriteLockLevel

Configuration API: org.apache.openjpa.conf.OpenJPAConfiguration.getWriteLockLevel

Resource adaptor config-property: WriteLockLevel

Default: write

Possible values: none, read, write, numeric values for lock-manager specific lock levels

Description: The default level at which to lock objects changed during a non-optimistic transaction. Note that for the default JDBC lock manager, read and write lock levels are equivalent.

6.  OpenJPA JDBC Properties

The following properties apply exclusively to the OpenJPA JDBC back-end.

6.1.  openjpa.jdbc.ConnectionDecorators

Property name: openjpa.jdbc.ConnectionDecorators

Configuration API: org.apache.openjpa.jdbc.conf.JDBCConfiguration.getConnectionDecorators

Resource adaptor config-property: ConnectionDecorators

Default: -

Description: A comma-separated list of plugin strings (see Section 4, “ Plugin Configuration ”) describing org.apache.openjpa.lib.jdbc.ConnectionDecorator instances to install on the connection factory. These decorators can wrap connections passed from the underlying DataSource to add functionality. OpenJPA will pass all connections through the list of decorators before using them. Note that by default OpenJPA employs all of the built-in decorators in the org.apache.openjpa.lib.jdbc package already; you do not need to list them here.

6.2.  openjpa.jdbc.DBDictionary

Property name: openjpa.jdbc.DBDictionary

Configuration API: org.apache.openjpa.jdbc.conf.JDBCConfiguration.getDBDictionary

Resource adaptor config-property: DBDictionary

Default: Based on the openjpa.ConnectionURL openjpa.ConnectionDriverName

Description: A plugin string (see Section 4, “ Plugin Configuration ”) describing the org.apache.openjpa.jdbc.sql.DBDictionary to use for database interaction. OpenJPA typically auto-configures the dictionary based on the JDBC URL, but you may have to set this property explicitly if you are using an unrecognized driver, or to plug in your own dictionary for a database OpenJPA does not support out-of-the-box. See Section 4, “ Database Support ” for details.

6.3.  openjpa.jdbc.DriverDataSource

Property name: openjpa.jdbc.DriverDataSource

Configuration API: org.apache.openjpa.jdbc.conf.JDBCConfiguration.getDriverDataSource

Resource adaptor config-property: DriverDataSource

Default: pooling

Description: The alias or full class name of the org.apache.openjpa.jdbc.schema.DriverDataSource implementation to use to wrap JDBC Driver classes with javax.sql.DataSource instances.

6.4.  openjpa.jdbc.EagerFetchMode

Property name: openjpa.jdbc.EagerFetchMode

Configuration API: org.apache.openjpa.jdbc.conf.JDBCConfiguration.getEagerFetchMode

Resource adaptor config-property: EagerFetchMode

Default: parallel

Possible values: parallel, join, none

Description: Optimizes how OpenJPA loads persistent relations. This setting can also be varied at runtime. See Section 7, “ Eager Fetching ” for details.

6.5.  openjpa.jdbc.FetchDirection

Property name: openjpa.jdbc.FetchDirection

Configuration API: org.apache.openjpa.jdbc.conf.JDBCConfiguration.getFetchDirection

Resource adaptor config-property: FetchDirection

Default: forward

Possible values: forward, reverse, unknown

Description: The expected order in which query result lists will be accessed. This property can also be varied at runtime. See Section 9, “ Large Result Sets ” for details.

6.6.  openjpa.jdbc.JDBCListeners

Property name: openjpa.jdbc.JDBCListeners

Configuration API: org.apache.openjpa.jdbc.conf.JDBCConfiguration.getJDBCListeners

Resource adaptor config-property: JDBCListeners

Default: -

Description: A comma-separated list of plugin strings (see Section 4, “ Plugin Configuration ”) describing org.apache.openjpa.lib.jdbc.JDBCListener event listeners to install. These listeners will be notified on various JDBC-related events.

6.7.  openjpa.jdbc.LRSSize

Property name: openjpa.jdbc.LRSSize

Configuration API: org.apache.openjpa.jdbc.conf.JDBCConfiguration.getLRSSize

Resource adaptor config-property: LRSSize

Default: query

Possible values: query, last, unknown

Description: The strategy to use to calculate the size of a result list. This property can also be varied at runtime. See Section 9, “ Large Result Sets ” for details.

6.8.  openjpa.jdbc.MappingDefaults

Property name: openjpa.jdbc.MappingDefaults

Configuration API: org.apache.openjpa.jdbc.conf.JDBCConfiguration.getMappingDefaults

Resource adaptor config-property: MappingDefaults

Default: jpa

Description: A plugin string (see Section 4, “ Plugin Configuration ”) describing the org.apache.openjpa.jdbc.meta.MappingDefaults to use to define default column names, table names, and constraints for your persistent classes. See Section 5, “ Mapping Factory ” for details.

6.9.  openjpa.jdbc.MappingFactory

Property name: openjpa.jdbc.MappingFactory

Configuration API: org.apache.openjpa.jdbc.conf.JDBCConfiguration.getMappingFactory

Resource adaptor config-property: MappingFactory

Default: -

Description: A plugin string (see Section 4, “ Plugin Configuration ”) describing the org.apache.openjpa.meta.MetaDataFactory to use to store and retrieve object-relational mapping information for your persistent classes. See Section 5, “ Mapping Factory ” for details.

6.10.  openjpa.jdbc.ResultSetType

Property name: openjpa.jdbc.ResultSetType

Configuration API: org.apache.openjpa.jdbc.conf.JDBCConfiguration.getResultSetType

Resource adaptor config-property: ResultSetType

Default: forward-only

Possible values: forward-only , scroll-sensitive, scroll-insensitive

Description: The JDBC result set type to use when fetching result lists. This property can also be varied at runtime. See Section 9, “ Large Result Sets ” for details.

6.11.  openjpa.jdbc.Schema

Property name: openjpa.jdbc.Schema

Configuration API: org.apache.openjpa.jdbc.conf.JDBCConfiguration.getSchema

Resource adaptor config-property: Schema

Default: -

Description: The default schema name to prepend to unqualified table names. Also, the schema in which OpenJPA will create new tables. See Section 10, “ Default Schema ” for details.

6.12.  openjpa.jdbc.SchemaFactory

Property name: openjpa.jdbc.SchemaFactory

Configuration API: org.apache.openjpa.jdbc.conf.JDBCConfiguration.getSchemaFactory

Resource adaptor config-property: SchemaFactory

Default: dynamic

Possible values: dynamic, native, file, table, others

Description: A plugin string (see Section 4, “ Plugin Configuration ”) describing the org.apache.openjpa.jdbc.schema.SchemaFactory to use to store and retrieve information about the database schema. See Section 11.2, “ Schema Factory ” for details.

6.13.  openjpa.jdbc.Schemas

Property name: openjpa.jdbc.Schemas

Configuration API: org.apache.openjpa.jdbc.conf.JDBCConfiguration.getSchemas

Resource adaptor config-property: Schemas

Default: -

Description: A comma-separated list of the schemas and/or tables used for your persistent data. See Section 11.1, “ Schemas List ” for details.

6.14.  openjpa.jdbc.SQLFactory

Property name: openjpa.jdbc.SQLFactory

Configuration API: org.apache.openjpa.jdbc.conf.JDBCConfiguration.getSQLFactory

Resource adaptor config-property: SQLFactory

Default: default

Description: A plugin string (see Section 4, “ Plugin Configuration ”) describing the org.apache.openjpa.jdbc.sql.SQLFactory to use to abstract common SQL constructs.

6.15.  openjpa.jdbc.SubclassFetchMode

Property name: openjpa.jdbc.SubclassFetchMode

Configuration API: org.apache.openjpa.jdbc.conf.JDBCConfiguration.getSubclassFetchMode

Resource adaptor config-property: SubclassFetchMode

Default: parallel

Possible values: parallel, join, none

Description: How to select subclass data when it is in other tables. This setting can also be varied at runtime. See Section 7, “ Eager Fetching ”.

6.16.  openjpa.jdbc.SynchronizeMappings

Property name: openjpa.jdbc.SynchronizeMappings

Configuration API: org.apache.openjpa.jdbc.conf.JDBCConfiguration.getSynchronizeMappings

Resource adaptor config-property: SynchronizeMappings

Default: -

Description: Controls whether OpenJPA will attempt to run the mapping tool on all persistent classes to synchronize their mappings and schema at runtime. Useful for rapid test/debug cycles. See Section 1.3, “ Runtime Forward Mapping ” for more information.

6.17.  openjpa.jdbc.TransactionIsolation

Property name: openjpa.jdbc.TransactionIsolation

Configuration API: org.apache.openjpa.jdbc.conf.JDBCConfiguration.getTransactionIsolation

Resource adaptor config-property: TransactionIsolation

Default: default

Possible values: default, none, read-committed, read-uncommitted, repeatable-read, serializable

Description: The JDBC transaction isolation level to use. See Section 5, “ Setting the Transaction Isolation ” for details.

6.18.  openjpa.jdbc.UpdateManager

Property name: openjpa.jdbc.UpdateManager

Configuration API: org.apache.openjpa.jdbc.conf.JDBCConfiguration.getUpdateManager

Resource adaptor config-property: UpdateManager

Default: default

Description: The full class name of the org.apache.openjpa.jdbc.kernel.UpdateManager to use to flush persistent object changes to the datastore. The provided default implementation is org.apache.openjpa.jdbc.kernel.OperationOrderUpdateManager .

Chapter 3.  Logging

Logging is an important means of gaining insight into your application's runtime behavior. OpenJPA provides a flexible logging system that integrates with many existing runtime systems, such as application servers and servlet runners.

There are four built-in logging plugins: a default logging framework that covers most needs, a Log4J delegate, an Apache Commons Logging delegate, and a no-op implementation for disabling logging.

Warning

Logging can have a negative impact on performance. Disable verbose logging (such as logging of SQL statements) before running any performance tests. It is advisable to limit or disable logging for a production system. You can disable logging altogether by setting the openjpa.Log property to none.

1.  Logging Channels

Logging is done over a number of logging channels, each of which has a logging level which controls the verbosity of log messages recorded for the channel. OpenJPA uses the following logging channels:

  • openjpa.Tool: Messages issued by the OpenJPA command line and Ant tools. Most messages are basic statements detailing which classes or files the tools are running on. Detailed output is only available via the logging category the tool belongs to, such as openjpa.Enhance for the enhancer (see Section 2, “ Enhancement ”) or openjpa.MetaData for the mapping tool (see Section 1, “ Forward Mapping ”). This logging category is provided so that you can get a general idea of what a tool is doing without having to manipulate logging settings that might also affect runtime behavior.

  • openjpa.Configuration: Messages issued by the configuration framework.

  • openjpa.Enhance: Messages pertaining to enhancement and runtime class generation.

  • openjpa.MetaData: Details about the generation of metadata and object-relational mappings.

  • openjpa.Runtime: General OpenJPA runtime messages.

  • openjpa.Query: Messages about queries. Query strings and any parameter values, if applicable, will be logged to the TRACE level at execution time. Information about possible performance concerns will be logged to the INFO level.

  • openjpa.DataCache: Messages from the L2 data cache plugins.

  • openjpa.jdbc.JDBC: JDBC connection information. General JDBC information will be logged to the TRACE level. Information about possible performance concerns will be logged to the INFO level.

  • openjpa.jdbc.SQL: This is the most common logging channel to use. Detailed information about the execution of SQL statements will be sent to the TRACE level. It is useful to enable this channel if you are curious about the exact SQL that OpenJPA issues to the datastore.

    When using the built-in OpenJPA logging facilities, you can enable SQL logging by adding SQL=TRACE to your openjpa.Log property.

    OpenJPA can optionally reformat the logged SQL to make it easier to read. To enable pretty-printing, add PrettyPrint=true to the openjpa.ConnectionFactoryProperties property. You can control how many columns wide the pretty-printed SQL will be with the PrettyPrintLineLength property. The default line length is 60 columns.

    While pretty printing makes things easier to read, it can make output harder to process with tools like grep.

    Pretty-printing properties configuration might look like so:

    <property name="openjpa.Log" value="SQL=TRACE"/>
    <property name="openjpa.ConnectionFactoryProperties" 
        value="PrettyPrint=true, PrettyPrintLineLength=72"/>
    
  • openjpa.jdbc.Schema: Details about operations on the database schema.

2.  OpenJPA Logging

By default, OpenJPA uses a basic logging framework with the following output format:

millis  diagnostic context  level  [thread name]  channel - message

For example, when loading an application that uses OpenJPA, a message like the following will be sent to the openjpa.Runtime channel:

2107  INFO   [main] openjpa.Runtime - Starting OpenJPA 0.9.7

The default logging system accepts the following parameters:

  • File: The name of the file to log to, or stdout or stderr to send messages to standard out and standard error, respectively. By default, OpenJPA sends log messages to standard error.

  • DefaultLevel: The default logging level of unconfigured channels. Recognized values are TRACE, DEBUG, INFO, WARN, and ERROR. Defaults to INFO.

  • DiagnosticContext: A string that will be prepended to all log messages. If this is not supplied and a openjpa.Id property value is available, that value will be used.

  • <channel>: Using the last token of the logging channel name, you can configure the log level to use for that channel. See the examples below.

Example 3.1.  Standard OpenJPA Log Configuration

<property name="openjpa.Log" value="DefaultLevel=WARN, Runtime=INFO, Tool=INFO"/>

Example 3.2.  Standard OpenJPA Log Configuration + All SQL Statements

<property name="openjpa.Log" value="DefaultLevel=WARN, Runtime=INFO, Tool=INFO, SQL=TRACE"/>

Example 3.3.  Logging to a File

<property name="openjpa.Log" value="File=/tmp/org.apache.openjpa.log, DefaultLevel=WARN, Runtime=INFO, Tool=INFO"/>

3.  Disabling Logging

Disabling logging can be useful to analyze performance without any I/O overhead or to reduce verbosity at the console. To do this, set the openjpa.Log property to none.

Disabling logging permanently, however, will cause all warnings to be consumed. We recommend using one of the more sophisticated mechanisms described in this chapter.

4.  Log4J

When openjpa.Log is set to log4j, OpenJPA will delegate to Log4J for logging. In a standalone application, Log4J logging levels are controlled by a resource named log4j.properties , which should be available as a top-level resource (either at the top level of a jar file, or in the root of one of the CLASSPATH directories). When deploying to a web or EJB application server, Log4J configuration is often performed in a log4j.xml file instead of a properties file. For further details on configuring Log4J, please see the Log4J Manual. We present an example log4j.properties file below.

Example 3.4.  Standard Log4J Logging

log4j.rootCategory=WARN, console
log4j.category.openjpa.Tool=INFO
log4j.category.openjpa.Runtime=INFO
log4j.category.openjpa.Remote=WARN
log4j.category.openjpa.DataCache=WARN
log4j.category.openjpa.MetaData=WARN
log4j.category.openjpa.Enhance=WARN
log4j.category.openjpa.Query=WARN
log4j.category.openjpa.jdbc.SQL=WARN
log4j.category.openjpa.jdbc.JDBC=WARN
log4j.category.openjpa.jdbc.Schema=WARN

log4j.appender.console=org.apache.log4j.ConsoleAppender

5.  Apache Commons Logging

Set the openjpa.Log property to commons to use the Apache Jakarta Commons Logging thin library for issuing log messages. The Commons Logging libraries act as a wrapper around a number of popular logging APIs, including the Jakarta Log4J project, and the native java.util.logging package in JDK 1.4. If neither of these libraries are available, then logging will fall back to using simple console logging.

When using the Commons Logging framework in conjunction with Log4J, configuration will be the same as was discussed in the Log4J section above.

5.1.  JDK 1.4 java.util.logging

When using JDK 1.4 or higher in conjunction with OpenJPA's Commons Logging support, logging will proceed through Java's built-in logging provided by the java.util.logging package. For details on configuring the built-in logging system, please see the Java Logging Overview.

By default, JDK 1.4's logging package looks in the JAVA_HOME/lib/logging.properties file for logging configuration. This can be overridden with the java.util.logging.config.file system property. For example:

java -Djava.util.logging.config.file=mylogging.properties com.company.MyClass

Example 3.5.  JDK 1.4 Log Properties

# specify the handlers to create in the root logger
# (all loggers are children of the root logger)
# the following creates two handlers
handlers=java.util.logging.ConsoleHandler, java.util.logging.FileHandler
    
# set the default logging level for the root logger
.level=ALL
    
# set the default logging level for new ConsoleHandler instances
java.util.logging.ConsoleHandler.level=INFO

# set the default logging level for new FileHandler instances
java.util.logging.FileHandler.level=ALL

# set the default formatter for new ConsoleHandler instances
java.util.logging.ConsoleHandler.formatter=java.util.logging.SimpleFormatter

# set the default logging level for all OpenJPA logs
openjpa.Tool.level=INFO
openjpa.Runtime.level=INFO
openjpa.Remote.level=INFO
openjpa.DataCache.level=INFO
openjpa.MetaData.level=INFO
openjpa.Enhance.level=INFO
openjpa.Query.level=INFO
openjpa.jdbc.SQL.level=INFO
openjpa.jdbc.JDBC.level=INFO
openjpa.jdbc.Schema.level=INFO

6.  Custom Log

If none of available logging systems meet your needs, you can configure the logging system with a custom logger. You might use custom logging to integrate with a proprietary logging framework used by some applications servers, or for logging to a graphical component for GUI applications.

A custom logging framework must include an implementation of the org.apache.openjpa.lib.log.LogFactory interface. We present a custom LogFactory below.

Example 3.6.  Custom Logging Class

package com.xyz;

import org.apache.openjpa.lib.log.*;

public class CustomLogFactory
    implements LogFactory {   

    private String _prefix = "CUSTOM LOG";

    public void setPrefix (String prefix) {
        _prefix = prefix;
    }
 
    public Log getLog(String channel) {
        // Return a simple extension of AbstractLog that will log
        // everything to the System.err stream. Note that this is 
        // roughly equivalent to OpenJPA's default logging behavior.
        return new AbstractLog() {

            protected boolean isEnabled(short logLevel) {
                // log all levels
                return true;
            }

            protected void log (short type, String message, Throwable t) {
                // just send everything to System.err
                System.err.println(_prefix + ": " + type + ": "
                    + message + ": " + t);
            }
        };
    }
}

To make OpenJPA use your custom log factory, set the openjpa.Log configuration property to your factory's full class name. Because this property is a plugin property (see Section 4, “ Plugin Configuration ” ), you can also pass parameters to your factory. For example, to use the example factory above and set its prefix to "LOG MSG", you would set the openjpa.Log property to the following string:

com.xyz.CustomLogFactory(Prefix="LOG MSG")

Chapter 4.  JDBC

OpenJPA uses a relational database for object persistence. It communicates with the database using the Java DataBase Connectivity (JDBC) APIs. This chapter describes how to configure OpenJPA to work with the JDBC driver for your database, and how to access JDBC functionality at runtime.

1.  Using the OpenJPA DataSource

OpenJPA includes its own simple javax.sql.DataSource implementation. If you choose to use OpenJPA's DataSource , then you must specify the following properties:

  • openjpa.ConnectionUserName: The JDBC user name for connecting to the database.

  • openjpa.ConnectionPassword: The JDBC password for the above user.

  • openjpa.ConnectionURL: The JDBC URL for the database.

  • openjpa.ConnectionDriverName: The JDBC driver class.

To configure advanced features, use the following optional properties. The syntax of these property strings follows the syntax of OpenJPA plugin parameters described in Section 4, “ Plugin Configuration ”.

  • openjpa.ConnectionProperties: If the listed driver is an instance of java.sql.Driver, this string will be parsed into a Properties instance, which will then be used to obtain database connections through the Driver.connect(String url, Properties props) method. If, on the other hand, the listed driver is a javax.sql.DataSource, the string will be treated as a plugin properties string, and matched to the bean setter methods of the DataSource instance.

  • openjpa.ConnectionFactoryProperties: OpenJPA's built-in DataSource allows you to set the following options via this plugin string:

    • QueryTimeout: The maximum number of seconds the JDBC driver will wait for a statement to execute.

    • PrettyPrint: Boolean indicating whether to pretty-print logged SQL statements.

    • PrettyPrintLineLength: The maximum number of characters in each pretty-printed SQL line.

Example 4.1.  Properties for the OpenJPA DataSource

<property name="openjpa.ConnectionUserName" value="user"/>
<property name="openjpa.ConnectionPassword" value="pass"/>
<property name="openjpa.ConnectionURL" value="jdbc:hsqldb:db-hypersonic"/>
<property name="openjpa.ConnectionDriverName" value="org.hsqldb.jdbcDriver"/>
<property name="openjpa.ConnectionFactoryProperties" 
    value="PrettyPrint=true, PrettyPrintLineLength=80"/>

2.  Using a Third-Party DataSource

You can use OpenJPA with any third-party javax.sql.DataSource . There are multiple ways of telling OpenJPA about a DataSource:

  • Set the DataSource into the map passed to Persistence.createEntityManagerFactory under the openjpa.ConnectionFactory key.

  • Bind the DataSource into JNDI, and then specify its location in the jta-data-source or non-jta-data-source element of the JPA XML format (depending on whether the DataSource is managed by JTA), or in the openjpa.ConnectionFactoryName property.

  • Specify the full class name of the DataSource implementation in the openjpa.ConnectionDriverName property in place of a JDBC driver. In this configuration OpenJPA will instantiate an instance of the named class via reflection. It will then configure the DataSource with the properties in the openjpa.ConnectionProperties setting.

The features of OpenJPA's own DataSource can also be used with third-party implementations. OpenJPA layers on top of the third-party DataSource to provide the extra functionality. To configure these features use the openjpa.ConnectionFactoryProperties property described in the previous section.

Example 4.2.  Properties File for a Third-Party DataSource

<property name="openjpa.ConnectionDriverName" value="oracle.jdbc.pool.OracleDataSource"/>
<property name="openjpa.ConnectionProperties" 
    value="PortNumber=1521, ServerName=saturn, DatabaseName=solarsid, DriverType=thin"/>
<property name="openjpa.ConnectionFactoryProperties" value="QueryTimeout=5000"/>

2.1.  Managed and XA DataSources

Certain application servers automatically enlist their DataSource s in global transactions. When this is the case, OpenJPA should not attempt to commit the underlying connection, leaving JDBC transaction completion to the application server. To notify OpenJPA that your third-party DataSource is managed by the application server, use the jta-data-source element of your persistence.xml file or set the openjpa.ConnectionFactoryMode property to managed.

Note that OpenJPA can only use managed DataSources when it is also integrating with the application server's managed transactions. Also note that all XA DataSources are enlisted, and you must set this property when using any XA DataSource.

When using a managed DataSource, you should also configure a second unmanaged DataSource that OpenJPA can use to perform tasks that are independent of the global transaction. The most common of these tasks is updating the sequence table OpenJPA uses to generate unique primary key values for your datastore identity objects. Configure the second DataSource using the non-jta-data-source persistence.xml element, or OpenJPA's various "2" connection properties, such as openjpa.ConnectionFactory2Name or openjpa.Connection2DriverName. These properties are outlined in Chapter 2, Configuration .

Example 4.3.  Managed DataSource Configuration

<!-- managed DataSource -->
<jta-data-source>java:/OracleXASource</jta-data-source>
<properties>
    <!-- use OpenJPA's built-in DataSource for unmanaged connections -->
    <property name="openjpa.Connection2UserName" value="scott"/>
    <property name="openjpa.Connection2Password" value="tiger"/>
    <property name="openjpa.Connection2URL" value="jdbc:oracle:thin:@CROM:1521:OpenJPADB"/>
    <property name="openjpa.Connection2DriverName" value="oracle.jdbc.driver.OracleDriver"/>
</properties>

3.  Runtime Access to DataSource

The JPA standard defines how to access JDBC connections from enterprise beans. OpenJPA also provides APIs to access an EntityManager's connection, or to retrieve a connection directly from the EntityManagerFactory's DataSource.

The OpenJPAEntityManager.getConnection method returns an EntityManager's connection. If the EntityManager does not already have a connection, it will obtain one. The returned connection is only guaranteed to be transactionally consistent with other EntityManager operations if the EntityManager is in a managed or non-optimistic transaction, if the EntityManager has flushed in the current transaction, or if you have used the OpenJPAEntityManager.beginStore method to ensure that a datastore transaction is in progress. Always close the returned connection before attempting any other EntityManager operations. OpenJPA will ensure that the underlying native connection is not released if a datastore transaction is in progress.

Example 4.4.  Using the EntityManager's Connection

import java.sql.*;
import org.apache.openjpa.persistence.*;

...

OpenJPAEntityManager kem = OpenJPAPersistence.cast(em);
Connection conn = (Connection) kem.getConnection();

// do JDBC stuff

conn.close();

The example below shows how to use a connection directly from the DataSource, rather than using an EntityManager 's connection.

Example 4.5.  Using the EntityManagerFactory's DataSource

import java.sql.*;
import javax.sql.*;
import org.apache.openjpa.conf.*;
import org.apache.openjpa.persistence.*;

...

OpenJPAEntityManagerFactory kemf = OpenJPAPersistence.cast(emf);
OpenJPAConfiguration conf = kemf.getConfiguration();
DataSource dataSource = (DataSource) conf.getConnectionFactory();
Connection conn = dataSource.getConnection();

// do JDBC stuff

conn.close();

4.  Database Support

OpenJPA can take advantage of any JDBC 2.x compliant driver, making almost any major database a candidate for use. See our officially supported database list in Appendix B, Supported Databases for more information. Typically, OpenJPA auto-configures its JDBC behavior and SQL dialect for your database, based on the values of your connection-related configuration properties.

If OpenJPA cannot detect what type of database you are using, or if you are using an unsupported database, you will have to tell OpenJPA what org.apache.openjpa.jdbc.sql.DBDictionary to use. The DBDictionary abstracts away the differences between databases. You can plug a dictionary into OpenJPA using the openjpa.jdbc.DBDictionary configuration property. The built-in dictionaries are listed below. If you are using an unsupported database, you may have to write your own DBDictionary subclass, a simple process.

The example below demonstrates how to set a dictionary and configure its properties in your configuration file. The DBDictionary property uses OpenJPA's plugin syntax .

Example 4.6.  Specifying a DBDictionary

<property name="openjpa.jdbc.DBDictionary" value="hsql(SimulateLocking=true)"/>

4.1.  DBDictionary Properties

The standard dictionaries all recognize the following properties. These properties will usually not need to be overridden, since the dictionary implementation should use the appropriate default values for your database. You typically won't use these properties unless you are designing your own DBDictionary for an unsupported database.

  • DriverVendor: The vendor of the particular JDBC driver you are using. Some dictionaries must alter their behavior depending on the driver vendor. See the VENDOR_XXX constants defined in your dictionary's Javadoc for available options.

  • CatalogSeparator: The string the database uses to delimit between the schema name and the table name. This is typically "." , which is the default.

  • CreatePrimaryKeys: If false, then do not create database primary keys for identifiers. Defaults to true .

  • ConstraintNameMode: When creating constraints, whether to put the constraint name before the definition (before), just after the constraint type name (mid), or after the constraint definition (after). Defaults to before .

  • MaxTableNameLength: The maximum number of characters in a table name. Defaults to 128.

  • MaxColumnNameLength: The maximum number of characters in a column name. Defaults to 128.

  • MaxConstraintNameLength: The maximum number of characters in a constraint name. Defaults to 128.

  • MaxIndexNameLength: The maximum number of characters in an index name. Defaults to 128.

  • MaxAutoAssignNameLength: Set this property to the maximum length of name for sequences used for auto-increment columns. Names longer than this value are truncated. Defaults to 31.

  • MaxIndexesPerTable: The maximum number of indexes that can be placed on a single table. Defaults to no limit.

  • SupportsForeignKeys: Whether the database supports foreign keys. Defaults to true.

  • SupportsTimestampNanos: Whether the database supports nanoseconds with TIMESTAMP columns. Defaults to true.

  • SupportsUniqueConstraints: Whether the database supports unique constraints. Defaults to true.

  • SupportsDeferredConstraints: Whether the database supports deferred constraints. Defaults to true.

  • SupportsRestrictDeleteAction: Whether the database supports the RESTRICT foreign key delete action. Defaults to true.

  • SupportsCascadeDeleteAction: Whether the database supports the CASCADE foreign key delete action. Defaults to true.

  • SupportsNullDeleteAction: Whether the database supports the SET NULL foreign key delete action. Defaults to true.

  • SupportsDefaultDeleteAction: Whether the database supports the SET DEFAULT foreign key delete action. Defaults to true.

  • SupportsAlterTableWithAddColumn: Whether the database supports adding a new column in an ALTER TABLE statement. Defaults to true.

  • SupportsAlterTableWithDropColumn: Whether the database supports dropping a column in an ALTER TABLE statement. Defaults to true.

  • ReservedWords: A comma-separated list of reserved words for this database, beyond the standard SQL92 keywords.

  • SystemTables: A comma-separated list of table names that should be ignored.

  • SystemSchemas: A comma-separated list of schema names that should be ignored.

  • SchemaCase: The case to use when querying the database metadata about schema components. Defaults to making all names upper case. Available values are: upper, lower, preserve.

  • ValidationSQL: The SQL used to validate that a connection is still in a valid state. For example, " SELECT SYSDATE FROM DUAL " for Oracle.

  • InitializationSQL: A piece of SQL to issue against the database whenever a connection is retrieved from the DataSource .

  • JoinSyntax: The SQL join syntax to use in select statements. See Section 6, “ Setting the SQL Join Syntax ”.

  • CrossJoinClause: The clause to use for a cross join (cartesian product). Defaults to CROSS JOIN.

  • InnerJoinClause: The clause to use for an inner join. Defaults to INNER JOIN.

  • OuterJoinClause: The clause to use for an left outer join. Defaults to LEFT OUTER JOIN.

  • RequiresConditionForCrossJoin: Some databases require that there always be a conditional statement for a cross join. If set, this parameter ensures that there will always be some condition to the join clause.

  • ToUpperCaseFunction: SQL function call for for converting a string to upper case. Use the token {0} to represent the argument.

  • ToLowerCaseFunction: Name of the SQL function for converting a string to lower case. Use the token {0} to represent the argument.

  • StringLengthFunction: Name of the SQL function for getting the length of a string. Use the token {0} to represent the argument.

  • SubstringFunctionName: Name of the SQL function for getting the substring of a string.

  • DistinctCountColumnSeparator: The string the database uses to delimit between column expressions in a SELECT COUNT(DISTINCT column-list) clause. Defaults to null for most databases, meaning that multiple columns in a distinct COUNT clause are not supported.

  • ForUpdateClause: The clause to append to SELECT statements to issue queries that obtain pessimistic locks. Defaults to FOR UPDATE.

  • TableForUpdateClause: The clause to append to the end of each table alias in queries that obtain pessimistic locks. Defaults to null.

  • SupportsSelectForUpdate: If true, then the database supports SELECT statements with a pessimistic locking clause. Defaults to true.

  • SupportsLockingWithDistinctClause: If true, then the database supports FOR UPDATE select clauses with DISTINCT clauses.

  • SupportsLockingWithOuterJoin: If true, then the database supports FOR UPDATE select clauses with outer join queries.

  • SupportsLockingWithInnerJoin: If true, then the database supports FOR UPDATE select clauses with inner join queries.

  • SupportsLockingWithMultipleTables: If true, then the database supports FOR UPDATE select clauses that select from multiple tables.

  • SupportsLockingWithOrderClause: If true, then the database supports FOR UPDATE select clauses with ORDER BY clauses.

  • SupportsLockingWithSelectRange: If true, then the database supports FOR UPDATE select clauses with queries that select a range of data using LIMIT, TOP or the database equivalent. Defaults to true.

  • SimulateLocking: Some databases do not support pessimistic locking, which will result in an exception when you attempt a pessimistic transaction. Setting this property to true bypasses the locking check to allow pessimistic transactions even on databases that do not support locking. Defaults to false.

  • SupportsQueryTimeout: If true, then the JDBC driver supports calls to java.sql.Statement.setQueryTimeout.

  • SupportsHaving: Whether this database supports HAVING clauses in selects.

  • SupportsSelectStartIndex: Whether this database can create a select that skips the first N results.

  • SupportsSelectEndIndex: Whether this database can create a select that is limited to the first N results.

  • SupportsSubselect: Whether this database supports subselects in queries.

  • RequiresAliasForSubselect: If true, then the database requires that subselects in a FROM clause be assigned an alias.

  • SupportsMultipleNontransactionalResultSets: If true, then a nontransactional connection is capable of having multiple open ResultSet instances.

  • StorageLimitationsFatal: If true, then any data truncation/rounding that is performed by the dictionary in order to store a value in the database will be treated as a fatal error, rather than just issuing a warning.

  • StoreLargeNumbersAsStrings: Many databases have limitations on the number of digits that can be stored in a numeric field (for example, Oracle can only store 38 digits). For applications that operate on very large BigInteger and BigDecimal values, it may be necessary to store these objects as string fields rather than the database's numeric type. Note that this may prevent meaningful numeric queries from being executed against the database. Defaults to false.

  • StoreCharsAsNumbers: Set this property to false to store Java char fields as CHAR values rather than numbers. Defaults to true.

  • UseGetBytesForBlobs: If true, then ResultSet.getBytes will be used to obtain blob data rather than ResultSet.getBinaryStream.

  • UseGetObjectForBlobs: If true, then ResultSet.getObject will be used to obtain blob data rather than ResultSet.getBinaryStream.

  • UseSetBytesForBlobs: If true, then PreparedStatement.setBytes will be used to set blob data, rather than PreparedStatement.setBinaryStream.

  • UseGetStringForClobs: If true, then ResultSet.getString will be used to obtain clob data rather than ResultSet.getCharacterStream.

  • UseSetStringForClobs: If true, then PreparedStatement.setString will be used to set clob data, rather than PreparedStatement.setCharacterStream.

  • CharacterColumnSize: The default size of varchar and char columns. Typically 255.

  • ArrayTypeName: The overridden default column type for java.sql.Types.ARRAY. This is only used when the schema is generated by the mappingtool.

  • BigintTypeName: The overridden default column type for java.sql.Types.BIGINT. This is only used when the schema is generated by the mappingtool.

  • BinaryTypeName: The overridden default column type for java.sql.Types.BINARY. This is only used when the schema is generated by the mappingtool.

  • BitTypeName: The overridden default column type for java.sql.Types.BIT. This is only used when the schema is generated by the mappingtool.

  • BlobTypeName: The overridden default column type for java.sql.Types.BLOB. This is only used when the schema is generated by the mappingtool.

  • CharTypeName: The overridden default column type for java.sql.Types.CHAR. This is only used when the schema is generated by the mappingtool.

  • ClobTypeName: The overridden default column type for java.sql.Types.CLOB. This is only used when the schema is generated by the mappingtool.

  • DateTypeName: The overridden default column type for java.sql.Types.DATE. This is only used when the schema is generated by the mappingtool.

  • DecimalTypeName: The overridden default column type for java.sql.Types.DECIMAL. This is only used when the schema is generated by the mappingtool.

  • DistinctTypeName: The overridden default column type for java.sql.Types.DISTINCT. This is only used when the schema is generated by the mappingtool.

  • DoubleTypeName: The overridden default column type for java.sql.Types.DOUBLE. This is only used when the schema is generated by the mappingtool.

  • FloatTypeName: The overridden default column type for java.sql.Types.FLOAT. This is only used when the schema is generated by the mappingtool.

  • IntegerTypeName: The overridden default column type for java.sql.Types.INTEGER. This is only used when the schema is generated by the mappingtool.

  • JavaObjectTypeName: The overridden default column type for java.sql.Types.JAVAOBJECT. This is only used when the schema is generated by the mappingtool.

  • LongVarbinaryTypeName: The overridden default column type for java.sql.Types.LONGVARBINARY. This is only used when the schema is generated by the mappingtool.

  • LongVarcharTypeName: The overridden default column type for java.sql.Types.LONGVARCHAR. This is only used when the schema is generated by the mappingtool.

  • NullTypeName: The overridden default column type for java.sql.Types.NULL. This is only used when the schema is generated by the mappingtool.

  • NumericTypeName: The overridden default column type for java.sql.Types.NUMERIC. This is only used when the schema is generated by the mappingtool.

  • OtherTypeName: The overridden default column type for java.sql.Types.OTHER. This is only used when the schema is generated by the mappingtool.

  • RealTypeName: The overridden default column type for java.sql.Types.REAL. This is only used when the schema is generated by the mappingtool.

  • RefTypeName: The overridden default column type for java.sql.Types.REF. This is only used when the schema is generated by the mappingtool.

  • SmallintTypeName: The overridden default column type for java.sql.Types.SMALLINT. This is only used when the schema is generated by the mappingtool.

  • StructTypeName: The overridden default column type for java.sql.Types.STRUCT. This is only used when the schema is generated by the mappingtool.

  • TimeTypeName: The overridden default column type for java.sql.Types.TIME. This is only used when the schema is generated by the mappingtool.

  • TimestampTypeName: The overridden default column type for java.sql.Types.TIMESTAMP. This is only used when the schema is generated by the mappingtool.

  • TinyintTypeName: The overridden default column type for java.sql.Types.TINYINT. This is only used when the schema is generated by the mappingtool.

  • VarbinaryTypeName: The overridden default column type for java.sql.Types.VARBINARY. This is only used when the schema is generated by the mappingtool.

  • VarcharTypeName: The overridden default column type for java.sql.Types.VARCHAR. This is only used when the schema is generated by the mappingtool.

  • UseSchemaName: If false, then avoid including the schema name in table name references. Defaults to true .

  • TableTypes: Comma-separated list of table types to use when looking for tables during schema reflection, as defined in the java.sql.DatabaseMetaData.getTableInfo JDBC method. An example is: "TABLE,VIEW,ALIAS". Defaults to "TABLE".

  • SupportsSchemaForGetTables: If false, then the database driver does not support using the schema name for schema reflection on table names.

  • SupportsSchemaForGetColumns: If false, then the database driver does not support using the schema name for schema reflection on column names.

  • SupportsNullTableForGetColumns: If true, then the database supports passing a null parameter to DatabaseMetaData.getColumns as an optimization to get information about all the tables. Defaults to true.

  • SupportsNullTableForGetPrimaryKeys: If true, then the database supports passing a null parameter to DatabaseMetaData.getPrimaryKeys as an optimization to get information about all the tables. Defaults to false.

  • SupportsNullTableForGetIndexInfo: If true, then the database supports passing a null parameter to DatabaseMetaData.getIndexInfo as an optimization to get information about all the tables. Defaults to false.

  • SupportsNullTableForGetImportedKeys: If true, then the database supports passing a null parameter to DatabaseMetaData.getImportedKeys as an optimization to get information about all the tables. Defaults to false.

  • UseGetBestRowIdentifierForPrimaryKeys: If true, then metadata queries will use DatabaseMetaData.getBestRowIdentifier to obtain information about primary keys, rather than DatabaseMetaData.getPrimaryKeys.

  • RequiresAutoCommitForMetadata: If true, then the JDBC driver requires that autocommit be enabled before any schema interrogation operations can take place.

  • AutoAssignClause: The column definition clause to append to a creation statement. For example, " AUTO_INCREMENT " for MySQL. This property is set automatically in the dictionary, and should not need to be overridden, and is only used when the schema is generated using the mappingtool.

  • AutoAssignTypeName: The column type name for auto-increment columns. For example, " SERIAL " for PostgreSQL. This property is set automatically in the dictionary, and should not need to be overridden, and is only used when the schema is generated using the mappingtool.

  • LastGeneratedKeyQuery: The query to issue to obtain the last automatically generated key for an auto-increment column. For example, " select @@identity " for Sybase. This property is set automatically in the dictionary, and should not need to be overridden.

  • NextSequenceQuery: A SQL string for obtaining a native sequence value. May use a placeholder of {0} for the variable sequence name. Defaults to a database-appropriate value.

4.2.  MySQLDictionary Properties

The mysql dictionary also understands the following properties:

  • DriverDeserializesBlobs: Many MySQL drivers automatically deserialize BLOBs on calls to ResultSet.getObject. The MySQLDictionary overrides the standard DBDictionary.getBlobObject method to take this into account. If your driver does not deserialize automatically, set this property to false.

  • TableType: The MySQL table type to use when creating tables. Defaults to innodb.

  • UseClobs: Some older versions of MySQL do not handle clobs correctly. To enable clob functionality, set this to true. Defaults to false.

  • OptimizeMultiTableDeletes: MySQL as of version 4.0.0 supports multiple tables in DELETE statements. When this option is set, OpenJPA will use that syntax when doing bulk deletes from multiple tables. This can happen when the deleteTableContents SchemaTool action is used. (See Section 12, “ Schema Tool ” for more info about deleteTableContents.) Defaults to false, since the statement may fail if using InnoDB tables and delete constraints.

4.3.  OracleDictionary Properties

The oracle dictionary understands the following additional properties:

  • UseTriggersForAutoAssign: If true, then OpenJPA will allow simulation of auto-increment columns by the use of Oracle triggers. OpenJPA will assume that the current sequence value from the sequence specified in the AutoAssignSequenceName parameter will hold the value of the new primary key for rows that have been inserted. For more details on auto-increment support, see Section 3.4, “ Autoassign / Identity Strategy Caveats ” .

  • AutoAssignSequenceName: The global name of the sequence that OpenJPA will assume to hold the value of primary key value for rows that use auto-increment. If left unset, OpenJPA will use a the sequence named "SEQ_<table name>".

  • MaxEmbeddedBlobSize: Oracle is unable to persist BLOBs using the embedded update method when BLOBs get over a certain size. The size depends on database configuration, e.g. encoding. This property defines the maximum size BLOB to persist with the embedded method. Defaults to 4000 bytes.

  • MaxEmbeddedClobSize: Oracle is unable to persist CLOBs using the embedded update method when Clobs get over a certain size. The size depends on database configuration, e.g. encoding. This property defines the maximum size CLOB to persist with the embedded method. Defaults to 4000 characters.

  • UseSetFormOfUseForUnicode: Prior to Oracle 10i, statements executed against unicode capable columns (the NCHAR, NVARCHAR, NCLOB Oracle types) required special handling to be able to store unicode values. Setting this property to true (the default) will cause OpenJPA to attempt to detect when the column of one of these types, and if so, will attempt to correctly configure the statement using the OraclePreparedStatement.setFormOfUse. For more details, see the Oracle Readme For NChar. Note that this can only work if OpenJPA is able to access the underlying OraclePreparedStatement instance, which may not be possible when using some third-party datasources. If OpenJPA detects that this is the case, a warning will be logged.

5.  Setting the Transaction Isolation

OpenJPA typically retains the default transaction isolation level of the JDBC driver. However, you can specify a transaction isolation level to use through the openjpa.jdbc.TransactionIsolation configuration property. The following is a list of standard isolation levels. Note that not all databases support all isolation levels.

  • default: Use the JDBC driver's default isolation level. OpenJPA uses this option if you do not explicitly specify any other.

  • none: No transaction isolation.

  • read-committed: Dirty reads are prevented; non-repeatable reads and phantom reads can occur.

  • read-uncommitted: Dirty reads, non-repeatable reads and phantom reads can occur.

  • repeatable-read: Dirty reads and non-repeatable reads are prevented; phantom reads can occur.

  • serializable: Dirty reads, non-repeatable reads, and phantom reads are prevented.

Example 4.7.  Specifying a Transaction Isolation

<property name="openjpa.jdbc.TransactionIsolation" value="repeatable-read"/>

6.  Setting the SQL Join Syntax

Object queries often involve using SQL joins behind the scenes. You can configure OpenJPA to use either SQL 92-style join syntax, in which joins are placed in the SQL FROM clause, the traditional join syntax, in which join criteria are part of the WHERE clause, or a database-specific join syntax mandated by the DBDictionary. OpenJPA only supports outer joins when using SQL 92 syntax or a database-specific syntax with outer join support.

The openjpa.jdbc.DBDictionary plugin accepts the the JoinSyntax property to set the system's default syntax. The available values are:

  • traditional: Traditional SQL join syntax; outer joins are not supported.

  • database: The database's native join syntax. Databases that do not have a native syntax will default to one of the other options.

  • sql92: ANSI SQL92 join syntax. Outer joins are supported. Not all databases support this syntax.

You can change the join syntax at runtime through the OpenJPA fetch configuration API, which is described in Chapter 9, Runtime Extensions .

Example 4.8.  Specifying the Join Syntax Default

<property name="openjpa.jdbc.DBDictionary" value="JoinSyntax=sql92"/>

Example 4.9.  Specifying the Join Syntax at Runtime

import org.apache.openjpa.persistence.jdbc.*; 

... 

Query q = em.createQuery("select m from Magazine m where m.title = 'JDJ'");
OpenJPAQuery kq = OpenJPAPersistence.cast(q);
JDBCFetchPlan fetch = (JDBCFetchPlan) kq.getFetchPlan ();
fetch.setJoinSyntax(JDBCFetchPlan.JOIN_SYNTAX_SQL92);
List results = q.getResultList();

7.  Accessing Multiple Databases

Through the properties we've covered thus far, you can configure each EntityManagerFactory to access a different database. If your application accesses multiple databases, we recommend that you maintain a separate persistence unit for each one. This will allow you to easily load the appropriate resource for each database at runtime, and to give the correct configuration file to OpenJPA's command-line tools during development.

8.  Configuring the Use of JDBC Connections

In its default configuration, OpenJPA obtains JDBC connections on an as-needed basis. OpenJPA EntityManagers do not retain a connection to the database unless they are in a datastore transaction or there are open Query results that are using a live JDBC result set. At all other times, including during optimistic transactions, EntityManagers request a connection for each query, then immediately release the connection back to the pool.

In some cases, it may be more efficient to retain connections for longer periods of time. You can configure OpenJPA's use of JDBC connections through the openjpa.ConnectionRetainMode configuration property. The property accepts the following values:

  • always: Each EntityManager obtains a single connection and uses it until the EntityManager closes.

  • transaction: A connection is obtained when each transaction begins (optimistic or datastore), and is released when the transaction completes. Non-transactional connections are obtained on-demand.

  • on-demand: Connections are obtained only when needed. This option is equivalent to the transaction option when datastore transactions are used. For optimistic transactions, though, it means that a connection will be retained only for the duration of the datastore flush and commit process.

You can also specify the connection retain mode of individual EntityManagers when you retrieve them from the EntityManagerFactory. See Section 2.1, “ OpenJPAEntityManagerFactory ” for details.

The openjpa.FlushBeforeQueries configuration property controls another aspect of connection usage: whether to flush transactional changes before executing object queries. This setting only applies to queries that would otherwise have to be executed in-memory because the IgnoreChanges property is set to false and the query may involve objects that have been changed in the current transaction. Legal values are:

  • true: Always flush rather than executing the query in-memory. If the current transaction is optimistic, OpenJPA will begin a non-locking datastore transaction. This is the default.

  • false: Never flush before a query.

  • with-connection: Flush only if the EntityManager has already established a dedicated connection to the datastore, otherwise execute the query in-memory. This option is useful if you use long-running optimistic transactions and want to ensure that these transactions do not consume database resources until commit. OpenJPA's behavior with this option is dependent on the transaction status and mode, as well as the configured connection retain mode described earlier in this section.

The flush mode can also be varied at runtime using the OpenJPA fetch configuration API, discussed in Chapter 9, Runtime Extensions .

The table below describes the behavior of automatic flushing in various situations. In all cases, flushing will only occur if OpenJPA detects that you have made modifications in the current transaction that may affect the query's results.

Table 4.1.  OpenJPA Automatic Flush Behavior

FlushBeforeQueries = false FlushBeforeQueries = true FlushBeforeQueries = with-connection; ConnectionRetainMode = on-demand FlushBeforeQueries = with-connection; ConnectionRetainMode = transaction or always
IgnoreChanges = true no flush no flush no flush no flush
IgnoreChanges = false; no tx active no flush no flush no flush no flush
IgnoreChanges = false; datastore tx active no flush flush flush flush
IgnoreChanges = false; optimistic tx active no flush flush no flush unless flush has already been invoked flush

Example 4.10.  Specifying Connection Usage Defaults

<property name="openjpa.ConnectionRetainMode" value="on-demand"/>
<property name="openjpa.FlushBeforeQueries" value="true"/>

Example 4.11.  Specifying Connection Usage at Runtime

import org.apache.openjpa.persistence.*;

// obtaining an em with a certain connection retain mode
Map props = new HashMap();
props.put("openjpa.ConnectionRetainMode", "always");
EntityManager em = emf.createEntityManager(props);

9.  Large Result Sets

By default, OpenJPA uses standard forward-only JDBC result sets, and completely instantiates the results of database queries on execution. When using a JDBC driver that supports version 2.0 or higher of the JDBC specification, however, you can configure OpenJPA to use scrolling result sets that may not bring all results into memory at once. You can also configure the number of result objects OpenJPA keeps references to, allowing you to traverse potentially enormous amounts of data without exhausting JVM memory.

Note

You can also configure on-demand loading for individual collection and map fields via large result set proxies. See Section 5.4.2, “ Large Result Set Proxies ”.

Use the following properties to configure OpenJPA's handling of result sets:

  • openjpa.FetchBatchSize : The number of objects to instantiate at once when traversing a result set. This number will be set as the fetch size on JDBC Statement objects used to obtain result sets. It also factors in to the number of objects OpenJPA will maintain a hard reference to when traversing a query result.

    The fetch size defaults to -1, meaning all results will be instantiated immediately on query execution. A value of 0 means to use the JDBC driver's default batch size. Thus to enable large result set handling, you must set this property to 0 or to a positive number.

  • openjpa.jdbc.ResultSetType : The type of result set to use when executing database queries. This property accepts the following values, each of which corresponds exactly to the same-named java.sql.ResultSet constant:

    • forward-only: This is the default.

    • scroll-sensitive

    • scroll-insensitive

    Different JDBC drivers treat the different result set types differently. Not all drivers support all types.

  • openjpa.jdbc.FetchDirection: The expected order in which you will access the query results. This property affects the type of datastructure OpenJPA will use to hold the results, and is also given to the JDBC driver in case it can optimize for certain access patterns. This property accepts the following values, each of which corresponds exactly to the same-named java.sql.ResultSet FETCH constant:

    • forward: This is the default.

    • reverse

    • unknown

    Not all drivers support all fetch directions.

  • openjpa.jdbc.LRSSize : The strategy OpenJPA will use to determine the size of result sets. This property is only used if you change the fetch batch size from its default of -1, so that OpenJPA begins to use on-demand result loading. Available values are:

    • query: This is the default. The first time you ask for the size of a query result, OpenJPA will perform a SELECT COUNT(*) query to determine the number of expected results. Note that depending on transaction status and settings, this can mean that the reported size is slightly different than the actual number of results available.

    • last: If you have chosen a scrollable result set type, this setting will use the ResultSet.last method to move to the last element in the result set and get its index. Unfortunately, some JDBC drivers will bring all results into memory in order to access the last one. Note that if you do not choose a scrollable result set type, then this will behave exactly like unknown. The default result set type is forward-only, so you must change the result set type in order for this property to have an effect.

    • unknown: Under this setting OpenJPA will return Integer.MAX_VALUE as the size for any query result that uses on-demand loading.

Example 4.12.  Specifying Result Set Defaults

<property name="openjpa.FetchBatchSize" value="20"/>
<property name="openjpa.jdbc.ResultSetType" value="scroll-insensitive"/>
<property name="openjpa.jdbc.FetchDirection" value="forward"/>
<property name="openjpa.jdbc.LRSSize" value="last"/>

Many OpenJPA runtime components also have methods to configure these properties on a case-by-case basis through their fetch configuration. See Chapter 9, Runtime Extensions .

Example 4.13.  Specifying Result Set Behavior at Runtime

import java.sql.*;
import org.apache.openjpa.persistence.jdbc.*;

...

Query q = em.createQuery("select m from Magazine m where m.title = 'JDJ'");
OpenJPAQuery kq = OpenJPAPersistence.cast(q);
JDBCFetchPlan fetch = (JDBCFetchPlan) kq.getFetchPlan();
fetch.setFetchSize(20);
fetch.setResultSetType(ResultSet.TYPE_SCROLL_INSENSITIVE);
fetch.setFetchDirection(ResultSet.FETCH_FORWARD);
fetch.setLRSSize(JDBCFetchPlan.SIZE_LAST);
List results = q.getResultList();

10.  Default Schema

It is common to duplicate a database model in multiple schemas. You may have one schema for development and another for production, or different database users may access different schemas. OpenJPA facilitates these patterns with the openjpa.jdbc.Schema configuration property. This property establishes a default schema for any unqualified table names, allowing you to leave schema names out of your mapping definitions.

The Schema property also establishes the default schema for new tables created through OpenJPA tools, such as the mapping tool covered in Section 1, “ Forward Mapping ”.

11.  Schema Reflection

OpenJPA needs information about your database schema for two reasons. First, it can use schema information at runtime to validate that your schema is compatible with your persistent class definitions. Second, OpenJPA requires schema information during development so that it can manipulate the schema to match your object model. OpenJPA uses the SchemaFactory interface to provide runtime mapping information, and the SchemaTool for development-time data. Each is presented below.

11.1.  Schemas List

By default, schema reflection acts on all the schemas your JDBC driver can "see". You can limit the schemas and tables OpenJPA acts on with the openjpa.jdbc.Schemas configuration property. This property accepts a comma-separated list of schemas and tables. To list a schema, list its name. To list a table, list its full name in the form <schema-name>.<table-name>. If a table does not have a schema or you do not know its schema, list its name as .<table-name> (notice the preceding '.'). For example, to list the BUSOBJS schema, the ADDRESS table in the GENERAL schema, and the SYSTEM_INFO table, regardless of what schema it is in, use the string:

BUSOBJS,GENERAL.ADDRESS,.SYSTEM_INFO

Note

Some databases are case-sensitive with respect to schema and table names. Oracle, for example, requires names in all upper case.

11.2.  Schema Factory

OpenJPA relies on the openjpa.jdbc.SchemaFactory interface for runtime schema information. You can control the schema factory OpenJPA uses through the openjpa.jdbc.SchemaFactory property. There are several built-in options to choose from:

  • dynamic: This is the default setting. It is an alias for the org.apache.openjpa.jdbc.schema.DynamicSchemaFactory. The DynamicSchemaFactory is the most performant schema factory, because it does not validate mapping information against the database. Instead, it assumes all object-relational mapping information is correct, and dynamically builds an in-memory representation of the schema from your mapping metadata. When using this factory, it is important that your mapping metadata correctly represent your database's foreign key constraints so that OpenJPA can order its SQL statements to meet them.

  • native: This is an alias for the org.apache.openjpa.jdbc.schema.LazySchemaFactory . As persistent classes are loaded by the application, OpenJPA reads their metadata and object-relational mapping information. This factory uses the java.sql.DatabaseMetaData interface to reflect on the schema and ensure that it is consistent with the mapping data being read. Use this factory if you want up-front validation that your mapping metadata is consistent with the database during development. This factory accepts the following important properties:

    • ForeignKeys: Set to true to automatically read foreign key information during schema validation.

  • table: This is an alias for the org.apache.openjpa.jdbc.schema.TableSchemaFactory . This schema factory stores schema information as an XML document in a database table it creates for this purpose. If your JDBC driver doesn't support the java.sql.DatabaseMetaData standard interface, but you still want some schema validation to occur at runtime, you might use this factory. It is not recommended for most users, though, because it is easy for the stored XML schema definition to get out-of-synch with the actual database. This factory accepts the following properties:

    • Table: The name of the table to create to store schema information. Defaults to OPENJPA_SCHEMA.

    • PrimaryKeyColumn: The name of the table's numeric primary key column. Defaults to ID.

    • SchemaColumn: The name of the table's string column for holding the schema definition as an XML string. Defaults to SCHEMA_DEF .

  • file: This is an alias for the org.apache.openjpa.jdbc.schema.FileSchemaFactory . This factory is a lot like the TableSchemaFactory, and has the same advantages and disadvantages. Instead of storing its XML schema definition in a database table, though, it stores it in a file. This factory accepts the following properties:

    • File: The resource name of the XML schema file. By default, the factory looks for a resource called package.schema, located in any top-level directory of the CLASSPATH or in the top level of any jar in your CLASSPATH.

You can switch freely between schema factories at any time. The XML file format used by some factories is detailed in Section 13, “ XML Schema Format ” . As with any OpenJPA plugin, you can can also implement your own schema factory if you have needs not met by the existing options.

12.  Schema Tool

Most users will only access the schema tool indirectly, through the interfaces provided by other tools. You may find, however, that the schema tool is a powerful utility in its own right. The schema tool has two functions:

  1. To reflect on the current database schema, optionally translating it to an XML representation for further manipulation.

  2. To take in an XML schema definition, calculate the differences between the XML and the existing database schema, and apply the necessary changes to make the database match the XML.

The XML format used by the schema tool abstracts away the differences between SQL dialects used by different database vendors. The tool also automatically adapts its SQL to meet foreign key dependencies. Thus the schema tool is useful as a general way to manipulate schemas.

You can invoke the schema tool through its Java class, org.apache.openjpa.jdbc.schema.SchemaTool. In addition to the universal flags of the configuration framework, the schema tool accepts the following command line arguments:

  • -ignoreErrors/-i <true/t | false/f>: If false , an exception will be thrown if the tool encounters any database errors. Defaults to false.

  • -file/-f <stdout | output file>: Use this option to write a SQL script for the planned schema modifications, rather them committing them to the database. When used in conjunction with the export or reflect actions, the named file will be used to write the exported schema XML. If the file names a resource in the CLASSPATH, data will be written to that resource. Use stdout to write to standard output. Defaults to stdout.

  • -openjpaTables/-ot <true/t | false/f>: When reflecting on the schema, whether to reflect on tables and sequences whose names start with OPENJPA_. Certain OpenJPA components may use such tables - for example, the table schema factory option covered in Section 11.2, “ Schema Factory ”. When using other actions, openjpaTables controls whether these tables can be dropped. Defaults to false.

  • -dropTables/-dt <true/t | false/f>: Set this option to true to drop tables that appear to be unused during retain and refresh actions. Defaults to true.

  • -dropSequences/-dsq <true/t | false/f>: Set this option to true to drop sequences that appear to be unused during retain and refresh actions. Defaults to true.

  • -sequences/-sq <true/t | false/f>: Whether to manipulate sequences. Defaults to true.

  • -indexes/-ix <true/t | false/f>: Whether to manipulate indexes on existing tables. Defaults to true.

  • -primaryKeys/-pk <true/t | false/f>: Whether to manipulate primary keys on existing tables. Defaults to true.

  • -foreignKeys/-fk <true/t | false/f>: Whether to manipulate foreign keys on existing tables. Defaults to true.

  • -record/-r <true/t | false/f>: Use false to prevent writing the schema changes made by the tool to the current schema factory. Defaults to true.

  • -schemas/-s <schema list>: A list of schema and table names that OpenJPA should access during this run of the schema tool. This is equivalent to setting the openjpa.jdbc.Schemas property for a single run.

The schema tool also accepts an -action or -a flag. Multiple actions can be composed in a comma-separated list. The available actions are:

  • add: This is the default action if you do not specify one. It brings the schema up-to-date with the given XML document by adding tables, columns, indexes, etc. This action never drops any schema components.

  • retain: Keep all schema components in the given XML definition, but drop the rest from the database. This action never adds any schema components.

  • drop: Drop all schema components in the schema XML. Tables will only be dropped if they would have 0 columns after dropping all columns listed in the XML.

  • refresh: Equivalent to retain, then add.

  • build: Generate SQL to build a schema matching the one in the given XML file. Unlike add, this option does not take into account the fact that part of the schema defined in the XML file might already exist in the database. Therefore, this action is typically used in conjunction with the -file flag to write a SQL script. This script can later be used to recreate the schema in the XML.

  • reflect: Generate an XML representation of the current database schema.

  • createDB: Generate SQL to re-create the current database. This action is typically used in conjunction with the -file flag to write a SQL script that can be used to recreate the current schema on a fresh database.

  • dropDB: Generate SQL to drop the current database. Like createDB, this action can be used with the -file flag to script a database drop rather than perform it.

  • import: Import the given XML schema definition into the current schema factory. Does nothing if the factory does not store a record of the schema.

  • export: Export the current schema factory's stored schema definition to XML. May produce an empty file if the factory does not store a record of the schema.

  • deleteTableContents: Execute SQL to delete all rows from all tables that OpenJPA knows about.

Note

The schema tool manipulates tables, columns, indexes, constraints, and sequences. It cannot create or drop the database schema objects in which the tables reside, however. If your XML documents refer to named database schemas, those schemas must exist.

We present some examples of schema tool usage below.

Example 4.14.  Schema Creation

Add the necessary schema components to the database to match the given XML document, but don't drop any data:

java org.apache.openjpa.jdbc.schema.SchemaTool targetSchema.xml

Example 4.15.  SQL Scripting

Repeat the same action as the first example, but this time don't change the database. Instead, write any planned changes to a SQL script:

java org.apache.openjpa.jdbc.schema.SchemaTool -f script.sql targetSchema.xml

Write a SQL script that will re-create the current database:

java org.apache.openjpa.jdbc.schema.SchemaTool -a createDB -f script.sql

Example 4.16.  Table Cleanup

Refresh the schema and delete all contents of all tables that OpenJPA knows about:

java org.apache.openjpa.jdbc.schema.SchemaTool -a refresh,deleteTableContents

Example 4.17.  Schema Drop

Drop the current database:

java org.apache.openjpa.jdbc.schema.SchemaTool -a dropDB

Example 4.18.  Schema Reflection

Write an XML representation of the current schema to file schema.xml .

java org.apache.openjpa.jdbc.schema.SchemaTool -a reflect -f schema.xml

13.  XML Schema Format

The schema tool and schema factories all use the same XML format to represent database schema. The Document Type Definition (DTD) for schema information is presented below, followed by examples of schema definitions in XML.

<!ELEMENT schemas (schema)+>
<!ELEMENT schema (table|sequence)+>
<!ATTLIST schema name CDATA #IMPLIED>

<!ELEMENT sequence EMPTY>
<!ATTLIST sequence name CDATA #REQUIRED>
<!ATTLIST sequence initial-value CDATA #IMPLIED>
<!ATTLIST sequence increment CDATA #IMPLIED>
<!ATTLIST sequence allocate CDATA #IMPLIED>

<!ELEMENT table (column|index|pk|fk|unique)+>
<!ATTLIST table name CDATA #REQUIRED>

<!ELEMENT column EMPTY>
<!ATTLIST column name CDATA #REQUIRED> 
<!ATTLIST column type (array | bigint | binary | bit | blob | char | clob 
    | date | decimal | distinct | double | float | integer | java_object 
    | longvarbinary | longvarchar | null | numeric | other | real | ref 
    | smallint | struct | time | timestamp | tinyint | varbinary | varchar) 
    #REQUIRED>
<!ATTLIST column not-null (true|false) "false">
<!ATTLIST column auto-assign (true|false) "false">
<!ATTLIST column default CDATA #IMPLIED>
<!ATTLIST column size CDATA #IMPLIED>
<!ATTLIST column decimal-digits CDATA #IMPLIED>

<!-- the type-name attribute can be used when you want OpenJPA to   -->
<!-- use a particular SQL type declaration when creating the     -->
<!-- column. It is up to you to ensure that this type is         -->
<!-- compatible with the JDBC type used in the type attribute.   -->
<!ATTLIST column type-name CDATA #IMPLIED>

<!-- the 'column' attribute of indexes, pks, and fks can be used -->
<!-- when the element has only one column (or for foreign keys,  -->
<!-- only one local column); in these cases the on/join child    -->
<!-- elements can be omitted                                     -->
<!ELEMENT index (on)*>
<!ATTLIST index name CDATA #REQUIRED>
<!ATTLIST index column CDATA #IMPLIED>
<!ATTLIST index unique (true|false) "false">

<!-- the 'logical' attribute of pks should be set to 'true' if   --> 
<!-- the primary key does not actually exist in the database,    --> 
<!-- but the given column should be used as a primary key for    -->
<!-- O-R purposes                                                -->
<!ELEMENT pk (on)*>
<!ATTLIST pk name CDATA #IMPLIED>
<!ATTLIST pk column CDATA #IMPLIED>
<!ATTLIST pk logical (true|false) "false">

<!ELEMENT on EMPTY>
<!ATTLIST on column CDATA #REQUIRED>

<!-- fks with a delete-action of 'none' are similar to logical   -->
<!-- pks; they do not actually exist in the database, but        -->
<!-- represent a logical relation between tables (or their       -->
<!-- corresponding Java classes)                                 -->
<!ELEMENT fk (join)*>
<!ATTLIST fk name CDATA #IMPLIED>
<!ATTLIST fk deferred (true|false) "false">
<!ATTLIST fk to-table CDATA #REQUIRED>
<!ATTLIST fk column CDATA #IMPLIED>
<!ATTLIST fk delete-action (cascade|default|exception|none|null) "none">

<!ELEMENT join EMPTY>
<!ATTLIST join column CDATA #REQUIRED>
<!ATTLIST join to-column CDATA #REQUIRED>
<!ATTLIST join value CDATA #IMPLIED>

<!-- unique constraint -->
<!ELEMENT unique (on)*>
<!ATTLIST unique name CDATA #IMPLIED>
<!ATTLIST unique column CDATA #IMPLIED>
<!ATTLIST unique deferred (true|false) "false">

Example 4.19.  Basic Schema

A very basic schema definition.

<schemas>
    <schema>
        <sequence name="S_ARTS"/>
        <table name="ARTICLE">
            <column name="TITLE" type="varchar" size="255" not-null="true"/>
            <column name="AUTHOR_FNAME" type="varchar" size="28"> 
            <column name="AUTHOR_LNAME" type="varchar" size="28"> 
            <column name="CONTENT" type="clob"> 
        </table>
        <table name="AUTHOR">
            <column name="FIRST_NAME" type="varchar" size="28" not-null="true"> 
            <column name="LAST_NAME" type="varchar" size="28" not-null="true"> 
        </table>
    </schema>
</schemas>

Example 4.20.  Full Schema

Expansion of the above schema with primary keys, constraints, and indexes, some of which span multiple columns.

<schemas>
    <schema>
        <sequence name="S_ARTS"/>
        <table name="ARTICLE">
            <column name="TITLE" type="varchar" size="255" not-null="true"/>
            <column name="AUTHOR_FNAME" type="varchar" size="28"> 
            <column name="AUTHOR_LNAME" type="varchar" size="28"> 
            <column name="CONTENT" type="clob"> 
            <pk column="TITLE"/>
            <fk to-table="AUTHOR" delete-action="exception">
                <join column="AUTHOR_FNAME" to-column="FIRST_NAME"/>
                <join column="AUTHOR_LNAME" to-column="LAST_NAME"/>
            </fk>
            <index name="ARTICLE_AUTHOR">
                <on column="AUTHOR_FNAME"/>
                <on column="AUTHOR_LNAME"/>
            </index>
        </table>
        <table name="AUTHOR">
            <column name="FIRST_NAME" type="varchar" size="28" not-null="true"> 
            <column name="LAST_NAME" type="varchar" size="28" not-null="true"> 
            <pk>
                <on column="FIRST_NAME"/>
                <on column="LAST_NAME"/>
            </pk>
        </table>
    </schema>
</schemas>

Chapter 5.  Persistent Classes

Persistent class basics are covered in Chapter 4, Entity of the JPA Overview. This chapter details the persistent class features OpenJPA offers beyond the core JPA specification.

1.  Persistent Class List

Unlike many ORM products, OpenJPA does not need to know about all of your persistent classes at startup. OpenJPA discovers new persistent classes automatically as they are loaded into the JVM; in fact you can introduce new persistent classes into running applications under OpenJPA. However, there are certain situations in which providing OpenJPA with a persistent class list is helpful:

  • OpenJPA must be able to match entity names in JPQL queries to persistent classes. OpenJPA automatically knows the entity names of any persistent classes already loaded into the JVM. To match entity names to classes that have not been loaded, however, you must supply a persistent class list.

  • When OpenJPA manipulates classes in a persistent inheritance hierarchy, OpenJPA must be aware of all the classes in the hierarchy. If some of the classes have not been loaded into the JVM yet, OpenJPA may not know about them, and queries may return incorrect results.

  • If you configure OpenJPA to create the needed database schema on startup (see Section 1.3, “ Runtime Forward Mapping ”), OpenJPA must know all of your persistent classes up-front.

When any of these conditions are a factor in your JPA application, use the class, mapping-file, and jar-file elements of JPA's standard XML format to list your persistent classes. See Section 1, “ persistence.xml ” for details.

Note

Listing persistent classes (or their metadata or jar files) is an all-or-nothing endeavor. If your persistent class list is non-empty, OpenJPA will assume that any unlisted class is not persistent.

2.  Enhancement

In order to provide optimal runtime performance, flexible lazy loading, and efficient, immediate dirty tracking, OpenJPA uses an enhancer . An enhancer is a tool that automatically adds code to your persistent classes after you have written them. The enhancer post-processes the bytecode generated by your Java compiler, adding the necessary fields and methods to implement the required persistence features. This bytecode modification perfectly preserves the line numbers in stack traces and is compatible with Java debuggers. In fact, the only change to debugging is that the persistent setter and getter methods of entity classes using property access will be prefixed with pc in stack traces and step-throughs. For example, if your entity has a getId method for persistent property id, and that method throws an exception, the stack trace will report the exception from method pcgetId. The line numbers, however, will correctly correspond to the getId method in your source file.

The diagram above illustrates the compilation of a persistent class.

You can add the OpenJPA enhancer to your build process, or use Java 1.5's new instrumentation features to transparently enhance persistent classes when they are loaded into the JVM. The following sections describe each option.

2.1.  Enhancing at Build Time

The enhancer can be invoked at build time via its Java class, org.apache.openjpa.enhance.PCEnhancer.

Note

You can also enhance via Ant; see Section 1.2, “ Enhancer Ant Task ”.

Example 5.1.  Using the OpenJPA Enhancer

java org.apache.openjpa.enhance.PCEnhancer Magazine.java

The enhancer accepts the standard set of command-line arguments defined by the configuration framework (see Section 3, “ Command Line Configuration ” ), along with the following flags:

  • -directory/-d <output directory>: Path to the output directory. If the directory does not match the enhanced class' package, the package structure will be created beneath the directory. By default, the enhancer overwrites the original .class file.

  • -enforcePropertyRestrictions/-epr <true/t | false/f>: Whether to throw an exception when it appears that a property access entity is not obeying the restrictions placed on property access. Defaults to false.

  • -addDefaultConstructor/-adc <true/t | false/f>: The spec requires that all persistent classes define a no-arg constructor. This flag tells the enhancer whether to add a protected no-arg constructor to any persistent classes that don't already have one. Defaults to true.

  • -tmpClassLoader/-tcl <true/t | false/f>: Whether to load persistent classes with a temporary class loader. This allows other code to then load the enhanced version of the class within the same JVM. Defaults to true. Try setting this flag to false as a debugging step if you run into class loading problems when running the enhancer.

Each additional argument to the enhancer must be one of the following:

  • The full name of a class.

  • The .java file for a class.

  • The .class file of a class.

If you do not supply any arguments to the enhancer, it will run on the classes in your persistent class list (see Section 1, “ Persistent Class List ”).

You can run the enhancer over classes that have already been enhanced, in which case it will not further modify the class. You can also run it over classes that are not persistence-capable, in which case it will treat the class as persistence-aware. Persistence-aware classes can directly manipulate the persistent fields of persistence-capable classes.

Note that the enhancement process for subclasses introduces dependencies on the persistent parent class being enhanced. This is normally not problematic; however, when running the enhancer multiple times over a subclass whose parent class is not yet enhanced, class loading errors can occur. In the event of a class load error, simply re-compile and re-enhance the offending classes.

2.2.  Enhancing JPA Entities on Deployment

The JEE 5 specification includes hooks to automatically enhance JPA entities when they are deployed into a container. Thus, if you are using a JEE 5-compliant application server, OpenJPA will enhance your entities automatically at runtime. Note that if you prefer build-time enhancement, OpenJPA's runtime enhancer will correctly recognize and skip pre-enhanced classes.

If your application server does not support the JEE 5 enhancement hooks, consider using the build-time enhancement described above, or the more general runtime enhancement described in the next section.

2.3.  Enhancing at Runtime

OpenJPA includes a Java agent for automatically enhancing persistent classes as they are loaded into the JVM. Java agents are classes that are invoked prior to your application's main method. OpenJPA's agent uses JVM hooks to intercept all class loading to enhance classes that have persistence metadata before the JVM loads them.

Searching for metadata for every class loaded by the JVM can slow application initialization. One way to speed things up is to take advantage of the optional persistent class list described in Section 1, “ Persistent Class List ”. If you declare a persistent class list, OpenJPA will only search for metadata for classes in that list.

To employ the OpenJPA agent, invoke java with the -javaagent set to the path to your OpenJPA jar file.

Example 5.2.  Using the OpenJPA Agent for Runtime Enhancement

java -javaagent:/home/dev/openjpa/lib/openjpa.jar com.xyz.Main

You can pass settings to the agent using OpenJPA's plugin syntax (see Section 4, “ Plugin Configuration ”). The agent accepts the long form of any of the standard configuration options (Section 3, “ Command Line Configuration ” ). It also accepts the following options, the first three of which correspond exactly to to the same-named options of the enhancer tool described in Section 2.1, “ Enhancing at Build Time ”:

  • addDefaultConstructor

  • enforcePropertyRestrictions

  • scanDevPath: Boolean indicating whether to scan the classpath for persistent types if none have been configured. If you do not specify a persistent types list and do not set this option to true, OpenJPA will check whether each class loaded into the JVM is persistent, and enhance it accordingly. This may slow down class load times significantly.

Example 5.3.  Passing Options to the OpenJPA Agent

java -javaagent:/home/dev/openjpa/lib/openjpa.jar=addDefaultConstructor=false com.xyz.Main

2.4.  Serializing Enhanced Types

By default, OpenJPA maintains serialization compatibility between the enhanced and unenhanced versions of a class. This allows you to serialize instances between a server using OpenJPA and a client that does not have access to enhanced classes or OpenJPA libraries. In some cases, however, you can make the persist and attach processes more robust and efficient by allowing breaks in serialization compatibility. See Section 1.3, “ Defining the Detached Object Graph ” for details.

3.  Object Identity

OpenJPA makes several enhancements to JPA's standard entity identity.

3.1.  Datastore Identity

The JPA specification requires you to declare one or more identity fields in your persistent classes. OpenJPA fully supports this form of object identity, called application identity. OpenJPA, however, also supports datastore identity. In datastore identity, you do not declare any primary key fields. OpenJPA manages the identity of your persistent objects for you through a surrogate key in the database.

You can control how your JPA datastore identity value is generated through OpenJPA's org.apache.openjpa.persistence.DataStoreId class annotation. This annotation has strategy and generator properties that mirror the same-named properties on the standard javax.persistence.GeneratedValue annotation described in Section 2.2, “ Id ” of the JPA Overview.

To retrieve the identity value of a datastore identity entity, use the OpenJPAEntityManager.getObjectId(Object entity) method. See Section 2.2, “ OpenJPAEntityManager ” for more information on the OpenJPAEntityManager.

Example 5.4.  JPA Datastore Identity Metadata

import org.apache.openjpa.persistence.*;

@Entity
@DataStoreId
public class LineItem {

    ... no @Id fields declared ...
}

Internally, OpenJPA uses the public org.apache.openjpa.util.Id class for datastore identity objects. When writing OpenJPA plugins, you can manipulate datastore identity objects by casting them to this class. You can also create your own Id instances and pass them to any internal OpenJPA method that expects an identity object.

In JPA, you will never see Id instances directly. Instead, calling OpenJPAEntityManager.getObjectId on a datastore identity object will return the Long surrogate primary key value for that object. You can then use this value in calls to EntityManager.find for subsequent lookups of the same record.

3.2.  Entities as Identity Fields

The JPA specification limits identity fields to simple types. OpenJPA, however, also allows ManyToOne and OneToOne relations to be identity fields. To identify a relation field as an identity field, simply annotate it with both the @ManyToOne or @OneToOne relation annotation and the @Id identity annotation.

When finding an entity identified by a relation, pass the id of the relation to the EntityManager.find method:

Example 5.5.  Finding an Entity with an Entity Identity Field

public Delivery createDelivery(EntityManager em, Order order) {
    Delivery delivery = new Delivery();
    delivery.setId(o);
    delivery.setDelivered(new Date());
    return delivery;
}

public Delivery findDelivery(EntityManager em, Order order) {
    // use the identity of the related instance 
    return em.find(Delivery.class, order.getId());
}

When your entity has multiple identity fields, at least one of which is a relation to another entity, you must use an identity class (see Section 2.1, “ Identity Class ” in the JPA Overview). You cannot use an embedded identity object. Identity class fields corresponding to entity identity fields should be of the same type as the related entity's identity.

Example 5.6.  Id Class for Entity Identity Fields

@Entity
public class Order {

    @Id
    private long id;

    ... 
}

@Entity
@IdClass(LineItemId.class)
public class LineItem {
    
    @Id
    private int index;

    @Id
    @ManyToOne
    private Order order;

    ...
}

public class LineItemId {

    public int index;
    public long order; // same type as order's identity

    ...
}

In the example above, if Order had used an identity class OrderId in place of a simple long value, then the LineItemId.order field would have been of type OrderId.

3.3.  Application Identity Tool

If you choose to use application identity, you may want to take advantage of OpenJPA's application identity tool. The application identity tool generates Java code implementing the identity class for any persistent type using application identity. The code satisfies all the requirements the specification places on identity classes. You can use it as-is, or simply use it as a starting point, editing it to meet your needs.

Before you can run the application identity tool on a persistent class, the class must be compiled and must have complete metadata. All primary key fields must be marked as such in the metadata.

In JPA metadata, do not attempt to specify the @IdClass annotation unless you are using the application identity tool to overwrite an existing identity class. Attempting to set the value of the @IdClass to a non-existent class will prevent your persistent class from compiling. Instead, use the -name or -suffix options described below to tell OpenJPA what name to give your generated identity class. Once the application identity tool has generated the class code, you can set the @IdClass annotation.

The application identity tool can be invoked via its Java class, org.apache.openjpa.enhance.ApplicationIdTool.

Note

Section 1.3, “ Application Identity Tool Ant Task ” describes the application identity tool's Ant task.

Example 5.7.  Using the Application Identity Tool

java org.apache.openjpa.enhance.ApplicationIdTool -s Id Magazine.java

The application identity tool accepts the standard set of command-line arguments defined by the configuration framework (see Section 3, “ Command Line Configuration ”), including code formatting flags described in Section 3.1, “ Code Formatting ”. It also accepts the following arguments:

  • -directory/-d <output directory>: Path to the output directory. If the directory does not match the generated oid class' package, the package structure will be created beneath the directory. If not specified, the tool will first try to find the directory of the .java file for the persistence-capable class, and failing that will use the current directory.

  • -ignoreErrors/-i <true/t | false/f>: If false , an exception will be thrown if the tool is run on any class that does not use application identity, or is not the base class in the inheritance hierarchy (recall that subclasses never define the application identity class; they inherit it from their persistent superclass).

  • -token/-t <token>: The token to use to separate stringified primary key values in the string form of the object id. This option is only used if you have multiple primary key fields. It defaults to "::".

  • -name/-n <id class name>: The name of the identity class to generate. If this option is specified, you must run the tool on exactly one class. If the class metadata already names an object id class, this option is ignored. If the name is not fully qualified, the persistent class' package is prepended to form the qualified name.

  • -suffix/-s <id class suffix>: A string to suffix each persistent class name with to form the identity class name. This option is overridden by -name or by any object id class specified in metadata.

Each additional argument to the tool must be one of the following:

  • The full name of a persistent class.

  • The .java file for a persistent class.

  • The .class file of a persistent class.

If you do not supply any arguments to the tool, it will act on the classes in your persistent classes list (see Section 1, “ Persistent Class List ”).

3.4.  Autoassign / Identity Strategy Caveats

Section 2.3, “ Generated Value ” explains how to use JPA's IDENTITY generation type to automatically assign field values. However, here are some additional caveats you should be aware of when using IDENTITY generation:

  1. Your database must support auto-increment / identity columns, or some equivalent (see Section 4.3, “ OracleDictionary Properties ” for how to configure a combination of triggers and sequences to fake auto-increment support in Oracle).

  2. Auto-increment / identity columns must be an integer or long integer type.

  3. Databases support auto-increment / identity columns to varying degrees. Some do not support them at all. Others only allow a single such column per table, and require that it be the primary key column. More lenient databases may allow non-primary key auto-increment columns, and may allow more than one per table. See your database documentation for details.

4.  Managed Inverses

Bidirectional relations are an essential part of data modeling. Chapter 12, Mapping Metadata in the JPA Overview explains how to use the mappedBy annotation attribute to form bidirectional relations that also share datastore storage in JPA.

OpenJPA also allows you to define purely logical bidirectional relations. The org.apache.openjpa.persistence.InverseLogical annotation names a logical inverse in JPA metadata.

Example 5.8.  Specifying Logical Inverses

Magazine.coverPhoto and Photograph.mag are each mapped to different foreign keys in their respective tables, but form a logical bidirectional relation. Only one of the fields needs to declare the other as its logical inverse, though it is not an error to set the logical inverse of both fields.

import org.apache.openjpa.persistence.*;

@Entity
public class Magazine {

    @OneToOne
    private Photograph coverPhoto;

    ...
}

@Entity
public class Photograph {

    @OneToOne
    @InverseLogical("coverPhoto")
    private Magazine mag;

    ...
}

Java does not provide any native facilities to ensure that both sides of a bidirectional relation remain consistent. Whenever you set one side of the relation, you must manually set the other side as well.

By default, OpenJPA behaves the same way. OpenJPA does not automatically propagate changes from one field in bidirectional relation to the other field. This is in keeping with the philosophy of transparency, and also provides higher performance, as OpenJPA does not need to analyze your object graph to correct inconsistent relations.

If convenience is more important to you than strict transparency, however, you can enable inverse relation management in OpenJPA. Set the openjpa.InverseManager plugin property to true for standard management. Under this setting, OpenJPA detects changes to either side of a bidirectional relation (logical or physical), and automatically sets the other side appropriately on flush.

Example 5.9.  Enabling Managed Inverses

<property name="openjpa.InverseManager" value="true"/>

The inverse manager has options to log a warning or throw an exception when it detects an inconsistent bidirectional relation, rather than correcting it. To use these modes, set the manager's Action property to warn or exception, respectively.

By default, OpenJPA excludes large result set fields from management. You can force large result set fields to be included by setting the ManageLRS plugin property to true.

Example 5.10.  Log Inconsistencies

<property name="openjpa.InverseManager" value="true(Action=warn)"/>

5.  Persistent Fields

OpenJPA enhances the specification's support for persistent fields in many ways. This section documents aspects of OpenJPA's persistent field handling that may affect the way you design your persistent classes.

5.1.  Restoring State

While the JPA specification says that you should not use rolled back objects, such objects are perfectly valid in OpenJPA. You can control whether the objects' managed state is rolled back to its pre-transaction values with the openjpa.RestoreState configuration property. none does not roll back state (the object becomes hollow, and will re-load its state the next time it is accessed), immutable restores immutable values (primitives, primitive wrappers, strings) and clears mutable values so that they are reloaded on next access, and all restores all managed values to their pre-transaction state.

5.2.  Typing and Ordering

When loading data into a field, OpenJPA examines the value you assign the field in your declaration code or in your no-args constructor. If the field value's type is more specific than the field's declared type, OpenJPA uses the value type to hold the loaded data. OpenJPA also uses the comparator you've initialized the field with, if any. Therefore, you can use custom comparators on your persistent field simply by setting up the comparator and using it in your field's initial value.

Example 5.11.  Using Initial Field Values

Though the annotations are left out for simplicity, assume employeesBySal and departments are persistent fields in the class below.

public class Company {

    // OpenJPA will detect the custom comparator in the initial field value
    // and use it whenever loading data from the database into this field
    private Collection employeesBySal = new TreeSet(new SalaryComparator());
    private Map departments;

    public Company {
        // or we can initialize fields in our no-args constructor; even though
        // this field is declared type Map, OpenJPA will detect that it's 
        // actually a TreeMap and use natural ordering for loaded data
        departments = new TreeMap();
    }

    // rest of class definition...
}

5.3.  Calendar Fields and TimeZones

OpenJPA's support for the java.util.Calendar type will store only the Date part of the field, not the TimeZone associated with the field. When loading the date into the Calendar field, OpenJPA will use the TimeZone that was used to initialize the field.

Note

OpenJPA will automatically track changes made via modification methods in fields of type Calendar, with one exception: when using Java version 1.3, the set() method cannot be overridden, so when altering the calendar using that method, the field must be explicitly marked as dirty. This limitation does not apply when running with Java version 1.4 and higer.

5.4.  Proxies

At runtime, the values of all mutable second class object fields in persistent and transactional objects are replaced with implementation-specific proxies. On modification, these proxies notify their owning instance that they have been changed, so that the appropriate updates can be made on the datastore.

5.4.1.  Smart Proxies

Most proxies only track whether or not they have been modified. Smart proxies for collection and map fields, however, keep a record of which elements have been added, removed, and changed. This record enables the OpenJPA runtime to make more efficient database updates on these fields.

When designing your persistent classes, keep in mind that you can optimize for OpenJPA smart proxies by using fields of type java.util.Set , java.util.TreeSet, and java.util.HashSet for your collections whenever possible. Smart proxies for these types are more efficient than proxies for Lists. You can also design your own smart proxies to further optimize OpenJPA for your usage patterns. See the section on custom proxies for details.

5.4.2.  Large Result Set Proxies

Under standard ORM behavior, traversing a persistent collection or map field brings the entire contents of that field into memory. Some persistent fields, however, might represent huge amounts of data, to the point that attempting to fully instantiate them can overwhelm the JVM or seriously degrade performance.

OpenJPA uses special proxy types to represent these "large result set" fields. OpenJPA's large result set proxies do not cache any data in memory. Instead, each operation on the proxy offloads the work to the database and returns the proper result. For example, the contains method of a large result set collection will perform a SELECT COUNT(*) query with the proper WHERE conditions to find out if the given element exists in the database's record of the collection. Similarly, each time you obtain an iterator OpenJPA performs the proper query using the current large result set settings, as discussed in the JDBC chapter. As you invoke Iterator.next, OpenJPA instantiates the result objects on-demand.

You can free the resources used by a large result set iterator by passing it to the static OpenJPAPersistence.close method.

Example 5.12.  Using a Large Result Set Iterator

import org.apache.openjpa.persistence.*;

...

Collection employees = company.getEmployees(); // employees is a lrs collection
Iterator itr = employees.iterator();
while (itr.hasNext())
    process((Employee) itr.next());
OpenJPAPersistence.close(itr);

You can also add and remove from large result set proxies, just as with standard fields. OpenJPA keeps a record of all changes to the elements of the proxy, which it uses to make sure the proper results are always returned from collection and map methods, and to update the field's database record on commit.

In order to use large result set proxies in JPA, add the org.apache.openjpa.persistence.LRS annotation to the persistent field.

The following restrictions apply to large result set fields:

  • The field must be declared as either a java.util.Collection or java.util.Map. It cannot be declared as any other type, including any sub-interface of collection or map, or any concrete collection or map class.

  • The field cannot have an externalizer (see Section 5.5, “ Externalization ”).

  • Because they rely on their owning object for context, large result set proxies cannot be transferred from one persistent field to another. The following code would result in an error on commit:

    Collection employees = company.getEmployees()  // employees is a lrs collection
    company.setEmployees(null);
    anotherCompany.setEmployees(employees);
    

Example 5.13.  Marking a Large Result Set Field

import org.apache.openjpa.persistence.*;

@Entity
public class Company {

    @ManyToMany
    @LRS private Collection<Employee> employees;

    ...
}

5.4.3.  Custom Proxies

OpenJPA manages proxies through the org.apache.openjpa.util.ProxyManager interface. OpenJPA includes a default proxy manager, the org.apache.openjpa.util.ProxyManagerImpl (with a plugin alias name of default), that will meet the needs of most users. The default proxy manager understands the following configuration properties:

  • TrackChanges: Whether to use smart proxies. Defaults to true.

  • AssertAllowedType: Whether to immediately throw an exception if you attempt to add an element to a collection or map that is not assignable to the element type declared in metadata. Defaults to false.

The default proxy manager can proxy the standard methods of any Collection, List, Map, Queue, Date, or Calendar class, including custom implementations. It can also proxy custom classes whose accessor and mutator methods follow JavaBean naming conventions. Your custom types must, however, meet the following criteria:

  • Custom container types must have a public no-arg constructor or a public constructor that takes a single Comparator parameter.

  • Custom date types must have a public no-arg constructor or a public constructor that takes a single long parameter representing the current time.

  • Other custom types must have a public no-arg constructor or a public copy constructor. If a custom types does not have a copy constructor, it must be possible to fully copy an instance A by creating a new instance B and calling each of B's setters with the value from the corresponding getter on A.

If you have custom classes that must be proxied and do not meet these requirements, OpenJPA allows you to define your own proxy classes and your own proxy manager. See the openjpa.util package Javadoc for details on the interfaces involved, and the utility classes OpenJPA provides to assist you.

You can plug your custom proxy manager into the OpenJPA runtime through the openjpa.ProxyManager configuration property.

Example 5.14.  Configuring the Proxy Manager

<property name="openjpa.ProxyManager" value="TrackChanges=false"/>

5.5.  Externalization

OpenJPA offers the ability to write custom field mappings in order to have complete control over the mechanism with which fields are stored, queried, and loaded from the datastore. Often, however, a custom mapping is overkill. There is often a simple transformation from a Java field value to its database representation. Thus, OpenJPA provides the externalization service. Externalization allows you to specify methods that will externalize a field value to its database equivalent on store and then rebuild the value from its externalized form on load.

Note

Fields of embeddable classes used for @EmbeddedId values in JPA cannot have externalizers.

The OpenJPA org.apache.openjpa.persistence.Externalizer annotation sets the name of a method that will be invoked to convert the field into its external form for database storage. You can specify either the name of a non-static method, which will be invoked on the field value, or a static method, which will be invoked with the field value as a parameter. Each method can also take an optional StoreContext parameter for access to a persistence context. The return value of the method is the field's external form. By default, OpenJPA assumes that all named methods belong to the field value's class (or its superclasses). You can, however, specify static methods of other classes using the format <class-name>.<method-name>.

Given a field of type CustomType that externalizes to a string, the table below demonstrates several possible externalizer methods and their corresponding metadata extensions.

Table 5.1.  Externalizer Options

Method Extension
public String CustomType.toString() @Externalizer("toString")
public String CustomType.toString(StoreContext ctx) @Externalizer("toString")
public static String AnyClass.toString(CustomType ct) @Externalizer("AnyClass.toString")
public static String AnyClass.toString(CustomType ct, StoreContext ctx) @Externalizer("AnyClass.toString")

The OpenJPA org.apache.openjpa.persistence.Factory annotation contains the name of a method that will be invoked to instantiate the field from the external form stored in the database. Specify a static method name. The method will will be invoked with the externalized value and must return an instance of the field type. The method can also take an optional StoreContext parameter for access to a persistence context. If a factory is not specified, OpenJPA will use the constructor of the field type that takes a single argument of the external type, or will throw an exception if no constructor with that signature exists.

Given a field of type CustomType that externalizes to a string, the table below demonstrates several possible factory methods and their corresponding metadata extensions.

Table 5.2.  Factory Options

Method Extension
public CustomType(String str) none
public static CustomType CustomType.fromString(String str) @Factory("fromString")
public static CustomType CustomType.fromString(String str, StoreContext ctx) @Factory("fromString")
public static CustomType AnyClass.fromString(String str) @Factory("AnyClass.fromString")
public static CustomType AnyClass.fromString(String str, StoreContext ctx) @Factory("AnyClass.fromString")

If your externalized field is not a standard persistent type, you must explicitly mark it persistent. In OpenJPA, you can force a persistent field by annotating it with org.apache.openjpa.persistence.Persistent annotation.

Note

If your custom field type is mutable and is not a standard collection, map, or date class, OpenJPA will not be able to detect changes to the field. You must mark the field dirty manually, or create a custom field proxy. See OpenJPAEntityManager.dirty for how to mark a field dirty manually in JPA. See Section 5.4, “ Proxies ” for a discussion of proxies.

You can externalize a field to virtually any value that is supported by OpenJPA's field mappings (embedded relations are the exception; you must declare your field to be a persistence-capable type in order to embed it). This means that a field can externalize to something as simple as a primitive, something as complex as a collection or map of entities, or anything in between. If you do choose to externalize to a collection or map, OpenJPA recognizes a family of metadata extensions for specying type information for the externalized form of your fields - see Section 3.2.6, “ Type ”. If the external form of your field is an entity object or contains entities, OpenJPA will correctly include the objects in its persistence-by-reachability algorithms and its delete-dependent algorithms.

The example below demonstrates a few forms of externalization.

Example 5.15.  Using Externalization

import org.apache.openjpa.persistence.*;

@Entity
public class Magazine {

    // use Class.getName and Class.forName to go to/from strings
    @Persistent
    @Externalizer("getName")
    @Factory("forName")
    private Class cls;

    // use URL.getExternalForm for externalization. no factory;
    // we can rely on the URL string constructor
    @Persistent
    @Externalizer("toExternalForm")
    private URL url;

    // use our custom methods
    @Persistent
    @Externalizer("Magazine.authorsFromCustomType")
    @Factory("Magazine.authorsToCustomType")
    @ElementType(Author.class)
    private CustomType customType;

    public static Collection authorsFromCustomType(CustomType customType) {
        ... logic to pack custom type into a list of authors ...
    }

    public static CustomType authorsToCustomType (Collection authors) {
        ... logic to create custom type from a collection of authors ...
    }

    ...
}

You can query externalized fields using parameters. Pass in a value of the field type when executing the query. OpenJPA will externalize the parameter using the externalizer method named in your metadata, and compare the externalized parameter with the value stored in the database. As a shortcut, OpenJPA also allows you to use parameters or literals of the field's externalized type in queries, as demonstrated in the example below.

Note

Currently, queries are limited to fields that externalize to a primitive, primitive wrapper, string, or date types, due to constraints on query syntax.

Example 5.16.  Querying Externalization Fields

Assume the Magazine class has the same fields as in the previous example.

// you can query using parameters
Query q = em.createQuery("select m from Magazine m where m.url = :u");
q.setParameter("u", new URL("http://www.solarmetric.com"));
List results = q.getResultList();

// or as a shortcut, you can use the externalized form directly
q = em.createQuery("select m from Magazine m where m.url = 'http://www.solarmetric.com'");
results = q.getResultList();

5.5.1.  External Values

Externalization often takes simple constant values and transforms them to constant values of a different type. An example would be storing a boolean field as a char, where true and false would be stored in the database as 'T' and 'F' respectively.

OpenJPA allows you to define these simple translations in metadata, so that the field behaves as in full-fledged externalization without requiring externalizer and factory methods. External values supports translation of pre-defined simple types (primitives, primitive wrappers, and Strings), to other pre-defined simple values.

Use the OpenJPA org.apache.openjpa.persistence.ExternalValues annotation to define external value translations. The values are defined in a format similar to that of configuration plugins, except that the value pairs represent Java and datastore values. To convert the Java boolean values of true and false to the character values T and F, for example, you would use the extension value: true=T,false=F.

If the type of the datastore value is different from the field's type, use the org.apache.openjpa.persistence.Type annotation to define the datastore type.

Example 5.17.  Using External Values

This example uses external value translation to transform a string field to an integer in the database.

public class Magazine {

    @ExternalValues({"SMALL=5", "MEDIUM=8", "LARGE=10"})
    @Type(int.class)
    private String sizeWidth;

    ...
}

6.  Fetch Groups

Fetch groups are sets of fields that load together. They can be used to to pool together associated fields in order to provide performance improvements over standard data fetching. Specifying fetch groups allows for tuning of lazy loading and eager fetching behavior.

The JPA Overview's Section 2.6.1, “ Fetch Type ” describes how to use JPA metadata annotations to control whether a field is fetched eagerly or lazily. Fetch groups add a dynamic aspect to this standard ability. As you will see, OpenJPA's JPA extensions allow you can add and remove fetch groups at runtime to vary the sets of fields that are eagerly loaded.

6.1.  Custom Fetch Groups

OpenJPA places any field that is eagerly loaded according to the JPA metadata rules into the built-in default fetch group. As its name implies, the default fetch group is active by default. You may also define your own named fetch groups and activate or deactivate them at runtime, as described later in this chapter. OpenJPA will eagerly-load the fields in all active fetch groups when loading objects from the datastore.

You create fetch groups with the org.apache.openjpa.persistence.FetchGroup annotation. If your class only has one custom fetch group, you can place this annotation directly on the class declaration. Otherwise, use the org.apache.openjpa.persistence.FetchGroups annotation to declare an array of individual FetchGroup values. The FetchGroup annotation has the following properties:

  • String name: The name of the fetch group. Fetch group names are global, and are expected to be shared among classes. For example, a shopping website may use a detail fetch group in each product class to efficiently load all the data needed to display a product's "detail" page. The website might also define a sparse list fetch group containing only the fields needed to display a table of products, as in a search result.

    The following names are reserved for use by OpenJPA: default , values, all, none, and any name beginning with jdo, jpa, or openjpa.

  • FetchAttribute[] attributes: The set of persistent fields or properties in the fetch group.

  • String[] fetchGroups: Other fetch groups whose fields to include in this group.

As you might expect, listing a org.apache.openjpa.persistence.FetchAttribute within a FetchGroup includes the corresponding persistent field or property in the fetch group. Each FetchAttribute has the following properties:

  • String name: The name of the persistent field or property to include in the fetch group.

  • recursionDepth: If the attribute represents a relation, the maximum number of same-typed relations to eager-fetch from this field. Defaults to 1. For example, consider an Employee class with a manager field, also of type Employee. When we load an Employee and the manager field is in an active fetch group, the recursion depth (along with the max fetch depth setting, described below) determines whether we only retrieve the target Employee and his manager (depth 1), or whether we also retrieve the manager's manager (depth 2), or the manager's manager's manager (depth 3), etc. Use -1 for unlimited depth.

Example 5.18.  Custom Fetch Group Metadata

Creates a detail fetch group consisting of the publisher and articles relations.

import org.apache.openjpa.persistence.*;

@Entity
@FetchGroups({
    @FetchGroup(name="detail", attributes={
        @FetchAttribute(name="publisher"),
        @FetchAttribute(name="articles")
    }),
    ...
})
public class Magazine {
   ...
}

A field can be a member of any number of fetch groups. A field can also declare a load fetch group. When you access a lazy-loaded field for the first time, OpenJPA makes a datastore trip to fetch that field's data. Sometimes, however, you know that whenever you access a lazy field A, you're likely to access lazy fields B and C as well. Therefore, it would be more efficient to fetch the data for A, B, and C in the same datastore trip. By setting A's load fetch group to the name of a fetch group containing B and C, you can tell OpenJPA to load all of these fields together when A is first accessed.

Use OpenJPA's org.apache.openjpa.persistence.LoadFetchGroup annotation to specify the load fetch group of any persistent field. The value of the annotation is the name of a declared fetch group whose members should be loaded along with the annotated field.

Example 5.19.  Load Fetch Group Metadata

import org.apache.openjpa.persistence.*;

@Entity
@FetchGroups({
    @FetchGroup(name="detail", attributes={
        @FetchAttribute(name="publisher"),
        @FetchAttribute(name="articles")
    }),
    ...
})
public class Magazine {

   @ManyToOne(fetch=FetchType.LAZY)
   @LoadFetchGroup("detail")
   private Publisher publisher;

   ...
}

6.2.  Custom Fetch Group Configuration

You can control the default set of fetch groups with the openjpa.FetchGroups configuration property. Set this property to a comma-separated list of fetch group names.

You can also set the system's default maximum fetch depth with the openjpa.MaxFetchDepth configuration property. The maximum fetch depth determines how "deep" into the object graph to traverse when loading an instance. For example, with a MaxFetchDepth of 1, OpenJPA will load at most the target instance and its immediate relations. With a MaxFetchDepth of 2, OpenJPA may load the target instance, its immediate relations, and the relations of those relations. This works to arbitrary depth. In fact, the default MaxFetchDepth value is -1, which symbolizes infinite depth. Under this setting, OpenJPA will fetch configured relations until it reaches the edges of the object graph. Of course, which relation fields are loaded depends on whether the fields are eager or lazy, and on the active fetch groups. A fetch group member's recursion depth may also limit the fetch depth to something less than the configured maximum.

OpenJPA's OpenJPAEntityManager and OpenJPAQuery extensions to the standard EntityManager and Query interfaces provide access to a org.apache.openjpa.persistence.FetchPlan object. The FetchPlan maintains the set of active fetch groups and the maximum fetch depth. It begins with the groups and depth defined in the openjpa.FetchGroups and openjpa.MaxFetchDepth properties, but allows you to add or remove groups and change the maximum fetch depth for an individual EntityManager or Query through the methods below.

public FetchPlan addFetchGroup(String group);
public FetchPlan addFetchGroups(String... groups);
public FetchPlan addFetchGroups(Collection groups);
public FetchPlan removeFetchGrop(String group);
public FetchPlan removeFetchGroups(String... groups);
public FetchPlan removeFetchGroups(Collection groups);
public FetchPlan resetFetchGroups();
public Collection<String> getFetchGroups();
public void clearFetchGroups();
public FetchPlan setMaxFetchDepth(int depth);
public int getMaxFetchDepth();

Chapter 9, Runtime Extensions details the OpenJPAEntityManager, OpenJPAQuery, and FetchPlan interfaces.

Example 5.20.  Using the FetchPlan

import org.apache.openjpa.persistence.*;

...

OpenJPAQuery kq = OpenJPAPersistence.cast(em.createQuery(...));
kq.getFetchPlan().setMaxFetchDepth(3).addFetchGroup("detail");
List results = kq.getResultList();

6.3.  Per-field Fetch Configuration

In addition to controlling fetch configuration on a per-fetch-group basis, you can configure OpenJPA to include particular fields in the current fetch plan. This allows you to add individual fields that are not in the default fetch group or in any other active fetch groups to the set of fields that will be eagerly loaded from the database.

JPA FetchPlan methods:

public FetchPlan addField(String field);
public FetchPlan addFields(String... fields);
public FetchPlan addFields(Class cls, String... fields);
public FetchPlan addFields(Collection fields);
public FetchPlan addFields(Class cls, Collection fields);
public FetchPlan removeField(String field);
public FetchPlan removeFields(String... fields);
public FetchPlan removeFields(Class cls, String... fields);
public FetchPlan removeFields(Collection fields);
public FetchPlan removeFields(Class cls, Collection fields);
public Collection<String> getFields();
public void clearFields();

The methods that take only string arguments use the fully-qualified field name, such as org.mag.Magazine.publisher. Similarly, getFields returns the set of fully-qualified field names. In all methods, the named field must be defined in the class specified in the invocation, not a superclass. So, if the field publisher is defined in base class Publication rather than subclass Magazine, you must invoke addField (Publication.class, "publisher") and not addField (Magazine.class, "publisher"). This is stricter than Java's default field-masking algorithms, which would allow the latter method behavior if Magazine did not also define a field called publisher.

In order to avoid the cost of reflection, OpenJPA does not perform any validation of the field name / class name pairs that you put into the fetch configuration. If you specify non-existent class / field pairs, nothing adverse will happen, but you will receive no notification of the fact that the specified configuration is not being used.

Example 5.21.  Adding an Eager Field

import org.apache.openjpa.persistence.*;

...

OpenJPAEntityManager kem = OpenJPAPersistence.cast(em);
kem.getFetchPlan().addField(Magazine.class, "publisher");
Magazine mag = em.find(Magazine.class, magId);

6.4.  Implementation Notes

  • Even when a direct relation is not eagerly fetched, OpenJPA selects the foreign key columns and caches the values. This way when you do traverse the relation, OpenJPA can often find the related object in its cache, or at least avoid joins when loading the related object from the database.

  • The above implicit foreign key-selecting behavior does not always apply when the relation is in a subclass table. If the subclass table would not otherwise be joined into the select, OpenJPA avoids the extra join just to select the foreign key values.

7.  Eager Fetching

Eager fetching is the ability to efficiently load subclass data and related objects along with the base instances being queried. Typically, OpenJPA has to make a trip to the database whenever a relation is loaded, or when you first access data that is mapped to a table other than the least-derived superclass table. If you perform a query that returns 100 Person objects, and then you have to retrieve the Address for each person, OpenJPA may make as many as 101 queries (the initial query, plus one for the address of each person returned). Or if some of the Person instances turn out to be Employees, where Employee has additional data in its own joined table, OpenJPA once again might need to make extra database trips to access the additional employee data. With eager fetching, OpenJPA can reduce these cases to a single query.

Eager fetching only affects relations in the active fetch groups, and is limited by the declared maximum fetch depth and field recursion depth (see Section 6, “ Fetch Groups ”). In other words, relations that would not normally be loaded immediately when retrieving an object or accessing a field are not affected by eager fetching. In our example above, the address of each person would only be eagerly fetched if the query were configured to include the address field or its fetch group, or if the address were in the default fetch group. This allows you to control exactly which fields are eagerly fetched in different situations. Similarly, queries that exclude subclasses aren't affected by eager subclass fetching, described below.

Eager fetching has three modes:

  • none: No eager fetching is performed. Related objects are always loaded in an independent select statement. No joined subclass data is loaded unless it is in the table(s) for the base type being queried. Unjoined subclass data is loaded using separate select statements rather than a SQL UNION operation.

  • join: In this mode, OpenJPA joins to to-one relations in the configured fetch groups. If OpenJPA is loading data for a single instance, then OpenJPA will also join to any collection field in the configured fetch groups. When loading data for multiple instances, though, (such as when executing a Query) OpenJPA will not join to collections by default. Instead, OpenJPA defaults to parallel mode for collections, as described below. You can force OpenJPA use a join rather than parallel mode for a collection field using the metadata extension described in Section 9.2.1, “ Eager Fetch Mode ”.

    Under join mode, OpenJPA uses a left outer join (or inner join, if the relations' field metadata declares the relation non-nullable) to select the related data along with the data for the target objects. This process works recursively for to-one joins, so that if Person has an Address, and Address has a TelephoneNumber, and the fetch groups are configured correctly, OpenJPA might issue a single select that joins across the tables for all three classes. To-many joins can not recursively spawn other to-many joins, but they can spawn recursive to-one joins.

    Under the join subclass fetch mode, subclass data in joined tables is selected by outer joining to all possible subclass tables of the type being queried. As you'll see below, subclass data fetching is configured separately from relation fetching, and can be disabled for specific classes.

    Note

    Some databases may not support outer joins. Also, OpenJPA can not use outer joins if you have set the DBDictionary's JoinSyntax to traditional. See Section 6, “ Setting the SQL Join Syntax ”.

  • parallel: Under this mode, OpenJPA selects to-one relations and joined collections as outlined in the join mode description above. Unjoined collection fields, however, are eagerly fetched using a separate select statement for each collection, executed in parallel with the select statement for the target objects. The parallel selects use the WHERE conditions from the primary select, but add their own joins to reach the related data. Thus, if you perform a query that returns 100 Company objects, where each company has a list of Employee objects and Department objects, OpenJPA will make 3 queries. The first will select the company objects, the second will select the employees for those companies, and the third will select the departments for the same companies. Just as for joins, this process can be recursively applied to the objects in the relations being eagerly fetched. Continuing our example, if the Employee class had a list of Projects in one of the fetch groups being loaded, OpenJPA would execute a single additional select in parallel to load the projects of all employees of the matching companies.

    Using an additional select to load each collection avoids transferring more data than necessary from the database to the application. If eager joins were used instead of parallel select statements, each collection added to the configured fetch groups would cause the amount of data being transferred to rise dangerously, to the point that you could easily overwhelm the network.

    Polymorphic to-one relations to table-per-class mappings use parallel eager fetching because proper joins are impossible. You can force other to-one relations to use parallel rather than join mode eager fetching using the metadata extension described in Section 9.2.1, “ Eager Fetch Mode ”.

    Parallel subclass fetch mode only applies to queries on joined inheritance hierarchies. Rather than outer-joining to subclass tables, OpenJPA will issue the query separately for each subclass. In all other situations, parallel subclass fetch mode acts just like join mode in regards to vertically-mapped subclasses.

    When OpenJPA knows that it is selecting for a single object only, it never uses parallel mode, because the additional selects can be made lazily just as efficiently. This mode only increases efficiency over join mode when multiple objects with eager relations are being loaded, or when multiple selects might be faster than joining to all possible subclasses.

7.1.  Configuring Eager Fetching

You can control OpenJPA's default eager fetch mode through the openjpa.jdbc.EagerFetchMode and openjpa.jdbc.SubclassFetchMode configuration properties. Set each of these properties to one of the mode names described in the previous section: none, join, parallel. If left unset, the eager fetch mode defaults to parallel and the subclass fetch mode defaults to join These are generally the most robust and performant strategies.

You can easily override the default fetch modes at runtime for any lookup or query through OpenJPA's fetch configuration APIs. See Chapter 9, Runtime Extensions for details.

Example 5.22.  Setting the Default Eager Fetch Mode

<property name="openjpa.jdbc.EagerFetchMode" value="parallel"/>
<property name="openjpa.jdbc.SubclassFetchMode" value="join"/>

Example 5.23.  Setting the Eager Fetch Mode at Runtime

import org.apache.openjpa.persistence.*;
import org.apache.openjpa.persistence.jdbc.*;

...

Query q = em.createQuery("select p from Person p where p.address.state = 'TX'");
OpenJPAQuery kq = OpenJPAPersistence.cast(q);
JDBCFetchPlan fetch = (JDBCFetchPlan) kq.getFetchPlan();
fetch.setEagerFetchMode(JDBCFetchPlan.EAGER_PARALLEL);
fetch.setSubclassFetchMode(JDBCFetchPlan.EAGER_JOIN);
List results = q.getResultList();

You can specify a default subclass fetch mode for an individual class with the metadata extension described in Section 9.1.1, “ Subclass Fetch Mode ”. Note, however, that you cannot "upgrade" the runtime fetch mode with your class setting. If the runtime fetch mode is none, no eager subclass data fetching will take place, regardless of your metadata setting.

This applies to the eager fetch mode metadata extension as well (see Section 9.2.1, “ Eager Fetch Mode ”). You can use this extension to disable eager fetching on a field or to declare that a collection would rather use joins than parallel selects or vice versa. But an extension value of join won't cause any eager joining if the fetch configuration's setting is none.

7.2.  Eager Fetching Considerations and Limitations

There are several important points that you should consider when using eager fetching:

  • When you are using parallel eager fetch mode and you have large result sets enabled (see Section 9, “ Large Result Sets ”) or you place a range on a query, OpenJPA performs the needed parallel selects on one page of results at a time. For example, suppose your FetchBatchSize is set to 20, and you perform a large result set query on a class that has collection fields in the configured fetch groups. OpenJPA will immediately cache the first 20 results of the query using join mode eager fetching only. Then, it will issue the extra selects needed to eager fetch your collection fields according to parallel mode. Each select will use a SQL IN clause (or multiple OR clauses if your class has a compound primary key) to limit the selected collection elements to those owned by the 20 cached results.

    Once you iterate past the first 20 results, OpenJPA will cache the next 20 and again issue any needed extra selects for collection fields, and so on. This pattern ensures that you get the benefits of eager fetching without bringing more data into memory than anticipated.

  • Once OpenJPA eager-joins into a class, it cannot issue any further eager to-many joins or parallel selects from that class in the same query. To-one joins, however, can recurse to any level.

  • Using a to-many join makes it impossible to determine the number of instances the result set contains without traversing the entire set. This is because each result object might be represented by multiple rows. Thus, queries with a range specification or queries configured for lazy result set traversal automatically turn off eager to-many joining.

  • OpenJPA cannot eagerly join to polymorphic relations to non-leaf classes in a table-per-class inheritance hierarchy. You can work around this restriction using the mapping extensions described in Section 9.2.2, “ Nonpolymorphic ”.

Chapter 6.  Metadata

The JPA Overview covers JPA metadata in Chapter 5, Metadata . This chapter discusses OpenJPA's extensions to standard JPA metadata.

1.  Metadata Factory

The openjpa.MetaDataFactory configuration property controls metadata loading and storing. This property takes a plugin string (see Section 4, “ Plugin Configuration ”) describing a concrete org.apache.openjpa.meta.MetaDataFactory implementation. A metadata factory can load mapping information as well as persistence metadata, or it can leave mapping information to a separate mapping factory (see Section 5, “ Mapping Factory ”). OpenJPA recognizes the following built-in metadata factories:

JPA has built-in settings for listing your persistent classes, which the JPA Overview describes. OpenJPA supports these JPA standard settings by translating them into its own internal metadata factory properties. Each internal property represents a different mechanism for locating persistent types; you can choose the mechanism or combination of mechanisms that are most convenient. See Section 1, “ Persistent Class List ” for a discussion of when it is necessary to list your persistent classes.

  • Types: A semicolon-separated list of fully-qualified persistent class names.

  • Resources: A semicolon-separated list of resource paths to metadata files or jar archives. Each jar archive will be scanned for annotated JPA entities.

  • URLs: A semicolon-separated list of URLs of metadata files or jar archives. Each jar archive will be scanned for annotated JPA entities.

  • ClasspathScan: A semicolon-separated list of directories or jar archives listed in your classpath. Each directory and jar archive will be scanned for annotated JPA entities.

Example 6.1.  Setting a Standard Metadata Factory

<property name="openjpa.MetaDataFactory" value="jpa(ClasspathScan=build;lib.jar)"/>

Example 6.2.  Setting a Custom Metadata Factory

<property name="openjpa.MetaDataFactory" value="com.xyz.CustomMetaDataFactory"/>

2.  Additional JPA Metadata

This section describes OpenJPA's core additions to standard entity metadata. We present the object-relational mapping syntax to support these additions in Section 7, “ Additional JPA Mappings ”. Finally, Section 3, “ Metadata Extensions ” covers additional extensions to JPA metadata that allow you to access auxiliary OpenJPA features.

2.1.  Datastore Identity

JPA typically requires you to declare one or more Id fields to act as primary keys. OpenJPA, however, can create and maintain a surrogate primary key value when you do not declare any Id fields. This form of persistent identity is called datastore identity. Section 3, “ Object Identity ” discusses OpenJPA's support for datastore identity in JPA. We cover how to map your datastore identity primary key column in Section 7.1, “ Datastore Identity Mapping ”

2.2.  Surrogate Version

Just as OpenJPA can maintain your entity's identity without any Id fields, OpenJPA can maintain your entity's optimistic version without any Version fields. Section 7.2, “ Surrogate Version Mapping ” shows you how to map surrogate version columns.

2.3.  Persistent Field Values

JPA defines Basic, Lob, Embedded , ManyToOne, and OneToOne persistence strategies for direct field values. OpenJPA supports all of these standard strategies, but adds one of its own: Persistent. The org.apache.openjpa.persistence.Persistent metadata annotation can represent any direct field value, including custom types. It has the following properties:

  • FetchType fetch: Whether to load the field eagerly or lazily. Corresponds exactly to the same-named property of standard JPA annotations such as Basic . Defaults to FetchType.EAGER.

  • CascadeType[] cascade: Array of enum values defining cascade behavior for this field. Corresponds exactly to the same-named property of standard JPA annotations such as ManyToOne. Defaults to empty array.

  • String mappedBy: Names the field in the related entity that maps this bidirectional relation. Corresponds to the same-named property of standard JPA annotations such as OneToOne.

  • boolean optional: Whether the value can be null. Corresponds to the same-named property of standard JPA annotations such as ManyToOne , but can apply to non-entity object values as well. Defaults to true.

  • boolean embedded: Set this property to true if the field value is stored as an embedded object.

Though you can use the Persistent annotation in place of most of the standard direct field annotations mentioned above, we recommend primarily using it for non-standard and custom types for which no standard JPA annotation exists. For example, Section 7.3, “ Multi-Column Mappings ” demonstrates the use of the Persistent annotation to denote a persistent java.awt.Point field.

2.4. Persistent Collection Fields

JPA standardizes support for collections of entities with the OneToMany and ManyToMany persistence strategies. OpenJPA supports these strategies, and may be extended for other strategies as well. For extended strategies, use the org.apache.openjpa.persistence.PersistentCollection metadata annotation to represents any persistent collection field. It has the following properties:

  • Class elementType: The class of the collection elements. This information is usually taken from the parameterized collection element type. You must supply it explicitly, however, if your field isn't a parameterized type.

  • FetchType fetch: Whether to load the collection eagerly or lazily. Corresponds exactly to the same-named property of standard JPA annotations such as Basic. Defaults to FetchType.LAZY.

  • String mappedBy: Names the field in the related entity that maps this bidirectional relation. Corresponds to the same-named property of standard JPA annotations such as ManyToMany.

  • CascadeType[] elementCascade: Array of enum values defining cascade behavior for the collection elements. Corresponds exactly to the cascade property of standard JPA annotations such as ManyToMany. Defaults to empty array.

  • boolean elementEmbedded: Set this property to true if the elements are stored as embedded objects.

2.5. Persistent Map Fields

JPA has limited support for maps. If you extend JPA's standard map support to encompass new mappings, use the org.apache.openjpa.persistence.PersistentMap metadata annotation to represent your custom persistent map fields. It has the following properties:

  • Class keyType: The class of the map keys. This information is usually taken from the parameterized map key type. You must supply it explicitly, however, if your field isn't a parameterized type.

  • Class elementType: The class of the map values. This information is usually taken from the parameterized map value type. You must supply it explicitly, however, if your field isn't a parameterized type.

  • FetchType fetch: Whether to load the collection eagerly or lazily. Corresponds exactly to the same-named property of standard JPA annotations such as Basic. Defaults to FetchType.LAZY.

  • CascadeType[] keyCascade: Array of enum values defining cascade behavior for the map keys. Corresponds exactly to the cascade property of standard JPA annotations such as ManyToOne. Defaults to empty array.

  • CascadeType[] elementCascade: Array of enum values defining cascade behavior for the map values. Corresponds exactly to the cascade property of standard JPA annotations such as ManyToOne. Defaults to empty array.

  • boolean keyEmbedded: Set this property to true if the map keys are stored as embedded objects.

  • boolean elementEmbedded: Set this property to true if the map values are stored as embedded objects.

3.  Metadata Extensions

OpenJPA extends standard metadata to allow you to access advanced OpenJPA functionality. This section covers persistence metadata extensions; we discuss mapping metadata extensions in Section 9, “ Mapping Extensions ”. All metadata extensions are optional; OpenJPA will rely on its defaults when no explicit data is provided.

3.1.  Class Extensions

OpenJPA recognizes the following class extensions:

3.1.1.  Fetch Groups

The org.apache.openjpa.persistence.FetchGroups and org.apache.openjpa.persistence.FetchGroup annotations allow you to define fetch groups in your JPA entities. Section 6, “ Fetch Groups ” discusses OpenJPA's support for fetch groups in general; see Section 6.1, “ Custom Fetch Groups ” for how to use these annotations in particular.

3.1.2.  Data Cache

Section 1, “ Data Cache ” examines caching in OpenJPA. Metadata extensions allow individual classes to override system caching defaults.

OpenJPA defines the org.apache.openjpa.persistence.DataCache annotation for caching information. This annotation has the following properties:

  • boolean enabled: Whether to cache data for instances of the class. Defaults to true for base classes, or the superclass value for subclasses. If you set this property to false, all other properties are ignored.

  • int timeout: The number of milliseconds data for the class remains valid. Use -1 for no timeout. Defaults to the openjpa.DataCacheTimeout property value.

3.1.3.  Detached State

The OpenJPA enhancer may add a synthetic field to detachable classes to hold detached state (see Section 1.3, “ Defining the Detached Object Graph ” for details). You can instead declare your own detached state field or supress the creation of a detached state field altogether. In the latter case, your class must not use datastore identity, and should declare a version field to detect optimistic concurrency errors during detached modifications.

OpenJPA defines the org.apache.openjpa.persistence.DetachedState annotation for controlling detached state. When used to annotate a class, DetachedState recognizes the following properties:

  • boolean enabled: Set to false to suppress the use of detached state.

  • String fieldName: Use this property to declare your own detached state field. The field must be of type Object. Typically this property is only used if the field is inherited from a non-persisted superclass. If the field is declared in your entity class, you will typically annotate the field directly, as described below.

If you declare your own detached state field, you can annotate that field with DetachedState directly, rather than placing the annotation at the class level and using the fieldName property. When placed on a field, DetachedState acts as a marker annotation; it does not recognize any properties. Your annotated field must be of type Object.

3.2.  Field Extensions

OpenJPA recognizes the following field extensions:

3.2.1.  Dependent

In a dependent relation, the referenced object is deleted whenever the owning object is deleted, or whenever the relation is severed by nulling or resetting the owning field. For example, if the Magazine.coverArticle field is marked dependent, then setting Magazine.coverArticle to a new Article instance will automatically delete the old Article stored in the field. Similarly, deleting a Magazine object will automatically delete its current cover Article. (This latter processing is analogous to using JPA's CascadeType.REMOVE functionality as described in Section 2.8.1, “ Cascade Type ”.) You can prevent an orphaned dependent object from being automatically deleted by assigning it to another relation in the same transaction.

OpenJPA offers a family of marker annotations to denote dependent relations in JPA entities:

3.2.2.  Load Fetch Group

The org.apache.openjpa.persistence.LoadFetchGroup annotation specifies a field's load fetch group. Section 6, “ Fetch Groups ” discusses OpenJPA's support for fetch groups in general; see Section 6.1, “ Custom Fetch Groups ” for how to use this annotation in particular.

3.2.3.  LRS

This boolean extension, denoted by the OpenJPA org.apache.openjpa.persistence.LRS annotation, indicates that a field should use OpenJPA's special large result set collection or map proxies. A complete description of large result set proxies is available in Section 5.4.2, “ Large Result Set Proxies ”.

3.2.4.  Inverse-Logical

This extension names the inverse field in a logical bidirectional relation. To create a logical bidrectional relation in OpenJPA, use the org.apache.openjpa.persistence.InverseLogical annotation. We discuss logical bidirectional relations and this extension in detail in Section 4, “ Managed Inverses ”.

3.2.5.  Read-Only

The read-only extension makes a field unwritable. The extension only applies to existing persistent objects; new object fields are always writeable.

To mark a field read-only in JPA metadata, set the org.apache.openjpa.persistence.ReadOnly annotation to an org.apache.openjpa.persistence.UpdateAction enum value. The UpdateAction enum includes:

  • UpdateAction.IGNORE: Updates to the field are completely ignored. The field is not considered dirty. The new value will not even get stored in the OpenJPA data cache.

  • UpdateAction.RESTRICT: Any attempt to change the field will result in an immediate exception.

3.2.6.  Type

OpenJPA has three levels of support for relations:

  1. Relations that hold a reference to an object of a concrete persistent class are supported by storing the primary key values of the related instance in the database.

  2. Relations that hold a reference to an object of an unknown persistent class are supported by storing the stringified identity value of the related instance. This level of support does not allow queries across the relation.

  3. Relations that hold an unknown object or interface. The only way to support these relations is to serialize their value to the database. This does not allow you to query the field, and is not very efficient.

Clearly, when you declare a field's type to be another persistence-capable class, OpenJPA uses level 1 support. By default, OpenJPA assumes that any interface-typed fields you declare will be implemented only by other persistent classes, and assigns interfaces level 2 support. The exception to this rule is the java.io.Serializable interface. If you declare a field to be of type Serializable, OpenJPA lumps it together with java.lang.Object fields and other non-interface, unrecognized field types, which are all assigned level 3 support.

With OpenJPA's type family of metadata extensions, you can control the level of support given to your unknown/interface-typed fields. Setting the value of this extension to Entity indicates that the field value will always be some persistent object, and gives level 2 support. Setting the value of this extension to the class of a concrete persistent type is even better; it gives you level 1 support (just as if you had declared your field to be of that type in the first place). Setting this extension to Object uses level 3 support. This is useful when you have an interface relation that may not hold other persistent objects (recall that OpenJPA assumes interface fields will always hold persistent instances by default).

This extension is also used with OpenJPA's externalization feature, described in Section 5.5, “ Externalization ”.

OpenJPA defines the following type annotations for field values, collection, array, and map elements, and map keys, respectively:

3.2.7.  Externalizer

The OpenJPA org.apache.openjpa.persistence.Externalizer annotation names a method to transform a field value into a value of another type. See Section 5.5, “ Externalization ” for details.

3.2.8.  Factory

The OpenJPA org.apache.openjpa.persistence.Factory annotation names a method to re-create a field value from its externalized form. See Section 5.5, “ Externalization ” for details.

3.2.9.  External Values

The OpenJPA org.apache.openjpa.persistence.ExternalValues annotation declares values for transformation of simple fields to different constant values in the datastore. See Section 5.5.1, “ External Values ” for details.

3.3.  Example

The following example shows you how to specify extensions in metadata.

Example 6.3.  OpenJPA Metadata Extensions

import org.apache.openjpa.persistence.*;

@Entity
@DataCache(enabled=false)
public class Magazine
{
    @ManyToMany
    @LRS
    private Collection<Subscriber> subscribers;

    @ExternalValues({"true=1", "false=2"})
    @Type(int.class)
    private boolean weekly;

    ...
}

Chapter 7.  Mapping

The JPA Overview's Chapter 12, Mapping Metadata explains object-relational mapping under JPA. This chapter reviews the mapping utilities OpenJPA provides and examines OpenJPA features that go beyond the JPA specification.

1.  Forward Mapping

Forward mapping is the process of creating mappings and their corresponding database schema from your object model. OpenJPA supports forward mapping through the mapping tool. The next section presents several common mapping tool use cases. You can invoke the tool through its Java class, org.apache.openjpa.jdbc.meta.MappingTool.

Note

Section 1.4, “ Mapping Tool Ant Task ” describes the mapping tool Ant task.

Example 7.1.  Using the Mapping Tool

java org.apache.openjpa.jdbc.meta.MappingTool Magazine.java

In addition to the universal flags of the configuration framework, the mapping tool accepts the following command line arguments:

  • -schemaAction/-sa <add | refresh | drop | build | retain | reflect | createDB | dropDB | import | export | none> : The action to take on the schema. These options correspond to the same-named actions on the schema tool described in Section 12, “ Schema Tool ”. Actions can be composed in a comma-separated list. Unless you are running the mapping tool on all of your persistent types at once or dropping a mapping, we strongly recommend you use the default add action or the build action. Otherwise you may end up inadvertently dropping schema components that are used by classes you are not currently running the tool over.

  • -schemaFile/-sf <stdout | output file>: Use this option to write the planned schema to an XML document rather than modify the database. The document can then be manipulated and committed to the database with the schema tool.

  • -sqlFile/-sql <stdout | output file>: Use this option to write the planned schema modifications to a SQL script rather than modify the database. Combine this with a schemaAction of build to generate a script that recreates the schema for the current mappings, even if the schema already exists.

  • -dropTables/-dt <true/t | false/f>: Corresponds to the same-named option on the schema tool.

  • -dropSequences/-dsq <true/t | false/f>: Corresponds to the same-named option on the schema tool.

  • -openjpaTables/-ot <true/t | false/f>: Corresponds to the same-named option on the schema tool.

  • -ignoreErrors/-i <true/t | false/f>: Corresponds to the same-named option on the schema tool.

  • -schemas/-s <schema and table names>: Corresponds to the same-named option on the schema tool. This option is ignored if readSchema is not set to true.

  • -readSchema/-rs <true/t | false/f>: Set this option to true to read the entire existing schema when the tool runs. Reading the existing schema ensures that OpenJPA does not generate any mappings that use table, index, primary key, or foreign key names that conflict with existing names. Depending on the JDBC driver, though, it can be a slow process for large schemas.

  • -primaryKeys/-pk <true/t | false/f>: Whether to read and manipulate primary key information of existing tables. Defaults to false.

  • -foreignKeys/-fk <true/t | false/f>: Whether to read and manipulate foreign key information of existing tables. Defaults to false. This means that to add any new foreign keys to a class that has already been mapped, you must explicitly set this flag to true.

  • -indexes/-ix <true/t | false/f>: Whether to read and manipulate index information of existing tables. Defaults to false. This means that to add any new indexes to a class that has already been mapped once, you must explicitly set this flag to true.

  • -sequences/-sq <true/t | false/f>: Whether to manipulate sequences. Defaults to true.

  • -meta/-m <true/t | false/f>: Whether the given action applies to metadata rather than or in addition to mappings.

The mapping tool also uses an -action/-a argument to specify the action to take on each class. The available actions are:

  • buildSchema: This is the default action. It makes the database schema match your existing mappings. If your provided mappings conflict with your class definitions, OpenJPA will fail with an informative exception.

  • validate: Ensure that the mappings for the given classes are valid and that they match the schema. No mappings or tables will be changed. An exception is thrown if any mappings are invalid.

Each additional argument to the tool should be one of:

  • The full name of a persistent class.

  • The .java file for a persistent class.

  • The .class file of a persistent class.

If you do not supply any arguments to the mapping tool, it will run on the classes in your persistent classes list (see Section 1, “ Persistent Class List ”).

The mappings generated by the mapping tool are stored by the system mapping factory. Section 5, “ Mapping Factory ” discusses your mapping factory options.

1.1.  Using the Mapping Tool

The JPA specification defines a comprehensive set of defaults for missing mapping information. Thus, forward mapping in JPA is virtually automatic. After using the mapping annotations covered in Chapter 12, Mapping Metadata of the JPA Overview to override any unsatisfactory defaults, run the mapping tool on your persistent classes. The default buildSchema mapping tool action manipulates the database schema to match your mappings. It fails if any of your mappings don't match your object model.

Example 7.2.  Creating the Relational Schema from Mappings

java org.apache.openjpa.jdbc.meta.MappingTool Magazine.java

To drop the schema for a persistent class, set the mapping tool's schemaAction to drop.

Example 7.3.  Refreshing entire schema and cleaning out tables

java org.apache.openjpa.jdbc.meta.MappingTool -sa add,deleteTableContents

Example 7.4.  Dropping Mappings and Association Schema

java org.apache.openjpa.jdbc.meta.MappingTool -sa drop Magazine.java

1.2.  Generating DDL SQL

The examples below show how to use the mapping tool to generate DDL SQL scripts, rather than modifying the database directly.

Example 7.5.  Create DDL for Current Mappings

This example uses your existing mappings to determine the needed schema, then writes the SQL to create that schema to create.sql.

java org.apache.openjpa.jdbc.meta.MappingTool -sa build -sql create.sql Magazine.java

Example 7.6.  Create DDL to Update Database for Current Mappings

This example uses your existing mappings to determine the needed schema. It then writes the SQL to add any missing tables and columns to the current schema to update.sql.

java org.apache.openjpa.jdbc.meta.MappingTool -sql update.sql Magazine.java

1.3.  Runtime Forward Mapping

You can configure OpenJPA to automatically run the mapping tool at runtime through the openjpa.jdbc.SynchronizeMappings configuration property. Using this property saves you the trouble of running the mapping tool manually, and is meant for use during rapid test/debug cycles.

In order to enable automatic runtime mapping, you must first list all your persistent classes as described in Section 1, “ Persistent Class List ”.

OpenJPA will run the mapping tool on these classes when your application obtains its first EntityManager.

The openjpa.jdbc.SynchronizeMappings property is a plugin string (see Section 4, “ Plugin Configuration ”) where the class name is the mapping tool action to invoke, and the properties are the MappingTool class' JavaBean properties. These properties correspond go the long versions of the tool's command line flags.

Example 7.7.  Configuring Runtime Forward Mapping

<property name="openjpa.jdbc.SynchronizeMappings" value="buildSchema(ForeignKeys=true)"/>

The setting above corresponds to running the following command:

java org.apache.openjpa.jdbc.meta.MappingTool -a buildSchema -fk true

2.  Reverse Mapping

OpenJPA includes a reverse mapping tool for generating persistent class definitions, complete with metadata, from an existing database schema. You do not have to use the reverse mapping tool to access an existing schema; you are free to write your classes and mappings yourself, as described in Section 3, “ Meet-in-the-Middle Mapping ”. The reverse mapping tool, however, can give you an excellent starting point from which to grow your persistent classes.

To use the reverse mapping tool, follow the steps below:

  1. Use the schema tool to export your current schema to an XML schema file. You can skip this step and the next step if you want to run the reverse mapping tool directly against the database.

    Example 7.8.  Reflection with the Schema Tool

    java org.apache.openjpa.jdbc.schema.SchemaTool -a reflect -f schema.xml
    
  2. Examine the generated schema file. JDBC drivers often provide incomplete or faulty metadata, in which case the file will not exactly match the actual schema. Alter the XML file to match the true schema. The XML format for the schema file is described in Section 13, “ XML Schema Format ”.

    After fixing any errors in the schema file, modify the XML to include foreign keys between all relations. The schema tool will have automatically detected existing foreign key constraints; many schemas, however, do not employ database foreign keys for every relation. By manually adding any missing foreign keys, you will give the reverse mapping tool the information it needs to generate the proper relations between the persistent classes it creates.

  3. Run the reverse mapping tool on the finished schema file. If you do not supply the schema file to reverse map, the tool will run directly against the schema in the database. The tool can be run via its Java class, org.apache.openjpa.jdbc.meta.ReverseMappingTool.

    Example 7.9.  Using the Reverse Mapping Tool

    java org.apache.openjpa.jdbc.meta.ReverseMappingTool -pkg com.xyz -d ~/src -cp customizer.properties schema.xml
    

    In addition to OpenJPA's standard configuration flags, including code formatting options, the reverse mapping tool recognizes the following command line arguments:

    • -schemas/-s <schema and table names>: A comma-separated list of schema and table names to reverse map, if no XML schema file is supplied. Each element of the list must follow the naming conventions for the openjpa.jdbc.Schemas property described in Section 11.1, “ Schemas List ”. In fact, if this flag is omitted, it defaults to the value of the Schemas property. If the Schemas property is not defined, all schemas will be reverse-mapped.

    • -package/-pkg <package name>: The package name of the generated classes. If no package name is given, the generated code will not contain package declarations.

    • -directory/-d <output directory>: All generated code and metadata will be written to the directory at this path. If the path does not match the package of a class, the package structure will be created beneath this directory. Defaults to the current directory.

    • -useSchemaName/-sn <true/t | false/f>: Set this flag to true to include the schema as well as table name in the name of each generated class. This can be useful when dealing with multiple schemas with same-named tables.

    • -useForeignKeyName/-fkn <true/t | false/f>: Set this flag to true if you would like field names for relations to be based on the database foreign key name. By default, relation field names are derived from the name of the related class.

    • -nullableAsObject/-no <true/t | false/f>: By default, all non-foreign key columns are mapped to primitives. Set this flag to true to generate primitive wrapper fields instead for columns that allow null values.

    • -blobAsObject/-bo <true/t | false/f>: By default, all binary columns are mapped to byte[] fields. Set this flag to true to map them to Object fields instead. Note that when mapped this way, the column is presumed to contain a serialized Java object.

    • -primaryKeyOnJoin/-pkj <true/t | false/f>: The standard reverse mapping tool behavior is to map all tables with primary keys to persistent classes. If your schema has primary keys on many-many join tables as well, set this flag to true to avoid creating classes for those tables.

    • -inverseRelations/-ir <true/t | false/f>: Set to false to prevent the creation of inverse 1-many/1-1 relations for every many-1/1-1 relation detected.

    • -useDatastoreIdentity/-ds <true/t | false/f>: Set to true to use datastore identity for tables that have single numeric primary key columns. The tool typically uses application identity for all generated classes.

    • -useBuiltinIdentityClass/-bic <true/t | false/f>: Set to false to prevent the tool from using built-in application identity classes when possible. This will force the tool to to create custom application identity classes even when there is only one primary key column.

    • -innerIdentityClasses/-inn <true/t | false/f>: Set to true to have any generated application identity classes be created as static inner classes within the persistent classes. Defaults to false.

    • -identityClassSuffix/-is <suffix>: Suffix to append to class names to form application identity class names, or for inner identity classes, the inner class name. Defaults to Id.

    • -typeMap/-typ <type mapping>: A string that specifies the default Java classes to generate for each SQL type that is seen in the schema. The format is SQLTYPE1=JavaClass1,SQLTYPE2=JavaClass2 . The SQL type name first looks for a customization based on SQLTYPE(SIZE,PRECISION), then SQLTYPE(SIZE), then SQLTYPE(SIZE,PRECISION). So if a column whose type name is CHAR is found, it will first look for the CHAR(50,0) type name specification, then it will look for CHAR(50), and finally it will just look for CHAR. For example, to generate a char array for every CHAR column whose size is exactly 50, and to generate a short for every type name of INTEGER, you might specify: CHAR(50)=char[],INTEGER=short. Note that since various databases report different type names differently, one database's type name specification might not work for another database. Enable TRACE level logging on the MetaData channel to track which type names OpenJPA is examining.

    • -customizerClass/-cc <class name>: The full class name of a org.apache.openjpa.jdbc.meta.ReverseCustomizer customization plugin. If you do not specify a reverse customizer of your own, the system defaults to a PropertiesReverseCustomizer. This customizer allows you to specify simple customization options in the properties file given with the -customizerProperties flag below. We present the available property keys below.

    • -customizerProperties/-cp <properties file or resource> : The path or resource name of a properties file to pass to the reverse customizer on initialization.

    • -customizer./-c.<property name> <property value> : The given property name will be matched with the corresponding Java bean property in the specified reverse customizer, and set to the given value.

    Running the tool will generate .java files for each generated class (and its application identity class, if applicable), along with an orm.xml file containing the corresponding persistence metadata.

  4. Examine the generated class, metadata, and mapping information, and modify it as necessary. Remember that the reverse mapping tool only provides a starting point, and you are free to make whatever modifications you like to the code it generates.

    After you are satisfied with the generated classes and their mappings, you should first compile the classes with javac, jikes, or your favorite Java compiler. Make sure the classes are located in the directory corresponding to the -package flag you gave the reverse mapping tool. Next, move the generated orm.xml file to a META-INF directory within a directory in your classpath. Finally, enhance the classes if necessary (see Section 2, “ Enhancement ” ).

Your persistent classes are now ready to access your existing schema.

2.1.  Customizing Reverse Mapping

The org.apache.openjpa.jdbc.meta.ReverseCustomizer plugin interface allows you to customze the reverse mapping process. See the class Javadoc for details on the hooks that this interface provides. Specify the concrete plugin implementation to use with the -customizerClass/-cc command-line flag, described in the preceding section.

By default, the reverse mapping tool uses a org.apache.openjpa.jdbc.meta.PropertiesReverseCustomizer . This customizer allows you to perform relatively simple customizations through the properties file named with the -customizerProperties tool flag. The customizer recognizes the following properties:

  • <table name>.table-type <type>: Override the default type of the table with name <table name>. Legal values are:

    • base: Primary table for a base class.

    • secondary: Secondary table for a class. The table must have a foreign key joining to a class table.

    • secondary-outer: Outer-joined secondary table for a class. The table must have a foreign key joining to a class table.

    • association: Association table. The table must have two foreign keys to class tables.

    • collection: Collection table. The table must have one foreign key to a class table and one data column.

    • subclass: A joined subclass table. The table must have a foreign key to the superclass' table.

    • none: The table should not be reverse-mapped.

  • <class name>.rename <new class name>: Override the given tool-generated name <class name> with a new value. Use full class names, including package. You are free to rename a class to a new package. Specify a value of none to reject the class and leave the corresponding table unmapped.

  • <table name>.class-name <new class name>: Assign the given fully-qualified class name to the type created from the table with name <table name>. Use a value of none to prevent reverse mapping this table. This property can be used in place of the rename property.

  • <class name>.identity <datastore | builtin | identity class name>: Set this property to datastore to use datastore identity for the class <class name>, builtin to use a built-in identity class, or the desired application identity class name. Give full class names, including package. You are free to change the package of the identity class this way. If the persistent class has been renamed, use the new class name for this property key. Remember that datastore identity requires a table with a single numeric primary key column, and built-in identity requires a single primary key column of any type.

  • <class name>.<field name>.rename <new field name> : Override the tool-generated <field name> in class <class name> with the given name. Use the field owner's full class name in the property key. If the field owner's class was renamed, use the new class name. The property value should be the new field name, without the preceding class name. Use a value of none to reject the generated mapping and remove the field from the class.

  • <table name>.<column name>.field-name <new field name>: Set the generated field name for the <table name> table's <column name> column. If this is a multi-column mapping, any of the columns can be used. Use a value of none to prevent the column and its associated columns from being reverse-mapped.

  • <class name>.<field name>.type <field type> : The type to give the named field. Use full class names. If the field or the field's owner class has been renamed, use the new name.

  • <class name>.<field name>.value: The initial value for the named field. The given string will be placed as-is in the generated Java code, so be sure it is valid Java. If the field or the field's owner class has been renamed, use the new name.

All property keys are optional; if not specified, the customizer keeps the default value generated by the reverse mapping tool.

Example 7.10.  Customizing Reverse Mapping with Properties

java org.apache.openjpa.jdbc.meta.ReverseMappingTool -pkg com.xyz -cp custom.properties schema.xml

Example custom.properties:

com.xyz.TblMagazine.rename:             com.xyz.Magazine
com.xyz.TblArticle.rename:              com.xyz.Article
com.xyz.TblPubCompany.rename:           com.xyz.pub.Company
com.xyz.TblSysInfo.rename:              none

com.xyz.Magazine.allArticles.rename:    articles
com.xyz.Magazine.articles.type:         java.util.Collection
com.xyz.Magazine.articles.value:        new TreeSet()
com.xyz.Magazine.identity:              datastore

com.xyz.pub.Company.identity:           com.xyz.pub.CompanyId

3.  Meet-in-the-Middle Mapping

In the meet-in-the-middle mapping approach, you control both the relational model and the object model. It is up to you to define the mappings between these models. The mapping tool's validate action is useful to meet-in-the-middle mappers. This action verifies that the mapping information for a class matches the class definition and the existing schema. It throws an informative exception when your mappings are incorrect.

Example 7.11.  Validating Mappings

java org.apache.openjpa.jdbc.meta.MappingTool -a validate Magazine.java

The buildSchema action we discussed in Section 1, “ Forward Mapping ” is also somewhat useful during meet-in-the-middle mapping. Unlike the validate action, which throws an exception if your mapping data does not match the existing schema, the buildSchema action assumes your mapping data is correct, and modifies the schema to match your mappings. This lets you modify your mapping data manually, but saves you the hassle of using your database's tools to bring the schema up-to-date.

4.  Mapping Defaults

The previous sections showed how to use the mapping tool to generate default mappings. But how does the mapping tool know what mappings to generate? The answer lies in the org.apache.openjpa.jdbc.meta.MappingDefaults interface. OpenJPA uses an instance of this interface to decide how to name tables and columns, where to put foreign keys, and generally how to create a schema that matches your object model.

Important

OpenJPA relies on foreign key constraint information at runtime to order SQL appropriately. Be sure to set your mapping defaults to reflect your existing database constraints, set the schema factory to reflect on the database for constraint information (see Section 11.2, “ Schema Factory ”), or use explicit foreign key mappings as described in Section 7.9.2, “ Foreign Keys ”.

The openjpa.jdbc.MappingDefaults configuration property controls the MappingDefaults interface implementation in use. This is a plugin property (see Section 4, “ Plugin Configuration ”), so you can substitute your own implementation or configure the existing ones. OpenJPA includes the following standard implementations:

  • jpa: Provides defaults in compliance with the JPA standard. This is an alias for the org.apache.openjpa.persistence.jdbc.PersistenceMappingDefaults class. This class extends the MappingDefaultsImpl class described below, so it has all the same properties (though with different default values), as well as:

    • PrependFieldNameToJoinTableInverseJoinColumns: Whether to prepend the owning field name to the names of inverse join columns in join tables. Defaults to true per the JPA specification. Set to false for compatibility with older OpenJPA versions which did not prepend the field name.

  • default: This is an alias for the org.apache.openjpa.jdbc.meta.MappingDefaultsImpl class. This default implementation is highly configurable. It has the following properties:

    • DefaultMissingInfo: Whether to default missing column and table names rather than throw an exception. If set to false, full explicit mappings are required at runtime and when using mapping tool actions like buildSchema and validate.

    • BaseClassStrategy: The default mapping strategy for base classes. You can specify a built-in strategy alias or the full class name of a custom class strategy. You can also use OpenJPA's plugin format (see Section 4, “ Plugin Configuration ”) to pass arguments to the strategy instance. See the org.apache.openjpa.jdbc.meta.strats package for available strategies.

    • SubclassStrategy: The default mapping strategy for subclasses. You can specify a builtin strategy alias or the full class name of a custom class strategy. You can also use OpenJPA's plugin format (see Section 4, “ Plugin Configuration ”) to pass arguments to the strategy instance. Common strategies are vertical and flat, the default. See the org.apache.openjpa.jdbc.meta.strats package for all available strategies.

    • VersionStrategy: The default version strategy for classes without a version field. You can specify a builtin strategy alias or the full class name of a custom version strategy. You can also use OpenJPA's plugin format (see Section 4, “ Plugin Configuration ”) to pass arguments to the strategy instance. Common strategies are none, state-comparison, timestamp, and version-number, the default. See the org.apache.openjpa.jdbc.meta.strats package for all available strategies.

    • DiscriminatorStrategy: The default discriminator strategy when no discriminator value is given. You can specify a builtin strategy alias or the full class name of a custom discriminator strategy. You can also use OpenJPA's plugin format (see Section 4, “ Plugin Configuration ”) to pass arguments to the strategy instance. Common strategies are final for a base class without subclasses, none to use joins to subclass tables rather than a discriminator column, and class-name, the default. See the org.apache.openjpa.jdbc.meta.strats package for all available strategies.

    • FieldStrategies: This property associates field types with custom strategies. The format of this property is similar to that of plugin strings (see Section 4, “ Plugin Configuration ” ), without the class name. It is a comma-separated list of key/value pairs, where each key is a possible field type, and each value is itself a plugin string describing the strategy for that type. We present an example below. See Section 10.3, “ Custom Field Mapping ” for information on custum field strategies.

    • ForeignKeyDeleteAction: The default delete action of foreign keys representing relations to other objects. Recognized values include restrict, cascade, null , default. These values correspond exactly to the standard database foreign key actions of the same names.

      The value none tells OpenJPA not to create database foreign keys on relation columns. This is the default.

    • JoinForeignKeyDeleteAction: The defualt delete action of foreign keys that join join secondary, collection, map, or subclass tables to the primary table. Accepts the same values as the ForeignKeyDeleteAction property above.

    • DeferConstraints: Whether to use deferred database constraints if possible. Defaults to false.

    • IndexLogicalForeignKeys: Boolean property controlling whether to create indexes on logical foreign keys. Logical foreign keys are columns that represent a link between tables, but have been configured through the ForeignKey properties above not to use a physical database foreign key. Defaults to true.

    • DataStoreIdColumnName: The default name of datastore identity columns.

    • DiscriminatorColumnName: The default name of discriminator columns.

    • IndexDiscriminator: Whether to index the discriminator column. Defaults to true.

    • VersionColumnName: The default name of version columns.

    • IndexVersion: Whether to index the version column. Defaults to false.

    • AddNullIndicator: Whether to create a synthetic null indicator column for embedded mappings. The null indicator column allows OpenJPA to distinguish between a null embedded object and one with default values for all persistent fields.

    • NullIndicatorColumnName: The default name of synthetic null indicator columns for embedded objects.

    • OrderLists: Whether to create a database ordering column for maintaining the order of persistent lists and arrays.

    • OrderColumnName: The default name of collection and array ordering columns.

    • StoreEnumOrdinal: Set to true to store enum fields as numeric ordinal values in the database. The default is to store the enum value name as a string, which is more robust if the Java enum declaration might be rearranged.

    • StoreUnmappedObjectIdString: Set to true to store the stringified identity of related objects when the declared related type is unmapped. By default, OpenJPA stores the related object's primary key value(s). However, this breaks down if different subclasses of the related type use incompatible primary key structures. In that case, stringifying the identity value is the better choice.

The example below turns on foreign key generation during schema creation and associates the org.mag.data.InfoStruct field type with the custom org.mag.mapping.InfoStructHandler value handler.

Example 7.12.  Configuring Mapping Defaults

<property name="openjpa.jdbc.MappingDefaults"
    value="ForeignKeyDeleteAction=restrict,
    FieldStrategies='org.mag.data.InfoStruct=org.mag.mapping.InfoStructHandler'"/>

5.  Mapping Factory

An important decision in the object-relational mapping process is how and where to store the data necessary to map your persistent classes to the database schema.

Section 1, “ Metadata Factory ” introduced OpenJPA's MetaDataFactory interface. OpenJPA uses this same interface to abstract the storage and retrieval of mapping information. OpenJPA includes the built-in mapping factories below, and you can create your own factory if you have custom needs. You control which mapping factory OpenJPA uses with the openjpa.jdbc.MappingFactory configuration property.

The bundled mapping factories are:

  • -: Leaving the openjpa.jdbc.MappingFactory property unset allows your metadata factory to take over mappings as well. If you are using the default jpa metadata factory, OpenJPA will read mapping information from your annotations and orm.xml when you leave the mapping factory unspecified.

Example 7.13.  Standard JPA Configuration

In the standard JPA configuration, the mapping factory is left unset.

<property name="openjpa.MetaDataFactory" value="jpa"/>

6.  Non-Standard Joins

The JPA Overview's Chapter 12, Mapping Metadata explains join mapping. All of the examples in that document, however, use "standard" joins, in that there is one foreign key column for each primary key column in the target table. OpenJPA supports additional join patterns, including partial primary key joins, non-primary key joins, and joins using constant values.

In a partial primary key join, the source table only has foreign key columns for a subset of the primary key columns in the target table. So long as this subset of columns correctly identifies the proper row(s) in the referenced table, OpenJPA will function properly. There is no special syntax for expressing a partial primary key join - just do not include column definitions for missing foreign key columns.

In a non-primary key join, at least one of the target columns is not a primary key. Once again, OpenJPA supports this join type with the same syntax as a primary key join. There is one restriction, however: each non-primary key column you are joining to must be controlled by a field mapping that implements the org.apache.openjpa.jdbc.meta.Joinable interface. All built in basic mappings implement this interface, including basic fields of embedded objects. OpenJPA will also respect any custom mappings that implement this interface. See Section 10, “ Custom Mappings ” for an examination of custom mappings.

Not all joins consist of only links between columns. In some cases you might have a schema in which one of the join criteria is that a column in the source or target table must have some constant value. OpenJPA calls joins involving constant values constant joins.

To form a constant join in JPA mapping, first set the JoinColumn 's name attribute to the name of the column. If the column with the constant value is the target of the join, give its fully qualified name in the form <table name>.<column name> . Next, set the referencedColumnName attribute to the constant value. If the constant value is a string, place it in single quotes to differentiate it from a column name.

Consider the tables above. First, we want to join row T1.R1 to row T2.R1. If we just join column T1.FK to T2.PK1, we will wind up matching both T2.R1 and T2.R2. So in addition to joining T1.FK to T2.PK1, we also have to specify that T2.PK2 has the value a. Here is how we'd accomplish this in mapping metadata.

@Entity
@Table(name="T1")
public class ...  {

    @ManyToOne
    @JoinColumns({
        @JoinColumn(name="FK" referencedColumnName="PK1"),
        @JoinColumn(name="T2.PK2" referencedColumnName="'a'")
    });
    private ...;
}

Notice that we had to fully qualify the name of column PK2 because it is in the target table. Also notice that we put single quotes around the constant value so that it won't be confused with a column name. You do not need single quotes for numeric constants. For example, the syntax to join T1.R2 to T2.R4 is:

@Entity
@Table(name="T1")
public class ...  {

    @ManyToOne
    @JoinColumns({
        @JoinColumn(name="FK" referencedColumnName="PK2"),
        @JoinColumn(name="T2.PK1" referencedColumnName="2")
    });
    private ...;
}

Finally, from the inverse direction, these joins would look like this:

@Entity
@Table(name="T2")
public class ...  {

    @ManyToOne
    @JoinColumns({
        @JoinColumn(name="T1.FK" referencedColumnName="PK1"),
        @JoinColumn(name="PK2" referencedColumnName="'a'")
    });
    private ...;

    @ManyToOne
    @JoinColumns({
        @JoinColumn(name="T1.FK" referencedColumnName="PK2"),
        @JoinColumn(name="PK1" referencedColumnName="2")
    });
    private ...;
}

7.  Additional JPA Mappings

OpenJPA supports many persistence strategies beyond those of the JPA specification. Section 2, “ Additional JPA Metadata ” covered the logical metadata for OpenJPA's additional persistence strategies. We now demonstrate how to map entities using these strategies to the database.

7.1.  Datastore Identity Mapping

Section 3, “ Object Identity ” describes how to use datastore identity in JPA. OpenJPA requires a single numeric primary key column to hold datastore identity values. The org.apache.openjpa.persistence.jdbc.DataStoreIdColumn annotation customizes the datastore identity column. This annotation has the following properties:

  • String name: Defaults to ID.

  • int precision

  • String columnDefinition

  • boolean insertable

  • boolean updatable

All properties correspond exactly to the same-named properties on the standard Column annotation, described in Section 3, “ Column ”.

Example 7.14.  Datastore Identity Mapping

import org.apache.openjpa.persistence.*;
import org.apache.openjpa.persistence.jdbc.*;

@Entity
@Table(name="LOGS")
@DataStoreIdColumn(name="ENTRY")
public class LogEntry {

    @Lob
    private String content;

    ...
}

7.2.  Surrogate Version Mapping

OpenJPA supports version fields as defined by the JPA specification, but allows you to use a surrogate version column in place of a version field if you like. You map the surrogate version column with the org.apache.openjpa.persistence.jdbc.VersionColumn annotation. You can also use the org.apache.openjpa.persistence.jdbc.VersionColumns annotation to declare an array of VersionColumn values. Each VersionColumn has the following properties:

  • String name: Defaults to VERSN.

  • int length

  • int precision

  • int scale

  • String columnDefinition

  • boolean nullable

  • boolean insertable

  • boolean updatable

All properties correspond exactly to the same-named properties on the standard Column annotation, described in Section 3, “ Column ”.

By default, OpenJPA assumes that surrogate versioning uses a version number strategy. You can choose a different strategy with the VersionStrategy annotation described in Section 9.1.4, “ Version Strategy ”.

7.3.  Multi-Column Mappings

OpenJPA makes it easy to create multi-column custom mappings. The JPA specification includes a Column annotation, but is missing a way to declare multiple columns for a single field. OpenJPA remedies this with the org.apache.openjpa.persistence.jdbc.Columns annotation, which contains an array of Column values.

Remember to annotate custom field types with Persistent, as described in Section 2.3, “ Persistent Field Values ”.

7.4.  Join Column Attribute Targets

Section 8.4, “ Direct Relations ” in the JPA Overview introduced you to the JoinColumn annotation. A JoinColumn's referencedColumnName property declares which column in the table of the related type this join column links to. Suppose, however, that the related type is unmapped, or that it is part of a table-per-class inheritance hierarchy. Each subclass that might be assigned to the field could reside in a different table, and could use entirely different names for its primary key columns. It becomes impossible to supply a single referencedColumnName that works for all subclasses.

OpenJPA rectifies this by allowing you to declare which attribute in the related type each join column links to, rather than which column. If the attribute is mapped differently in various subclass tables, OpenJPA automatically forms the proper join for the subclass record at hand. The org.apache.openjpa.persistence.jdbc.XJoinColumn annotation has all the same properties as the standard JoinColumn annotation, but adds an additional referencedAttributeName property for this purpose. Simply use a XJoinColumn in place of a JoinColumn whenever you need to access this added functionality.

For compound keys, use the org.apache.openjpa.persistence.jdbc.XJoinColumns annotation. The value of this annotation is an array of individual XJoinColumns.

7.5.  Embedded Mapping

JPA uses the AttributeOverride annotation to override the default mappings of an embeddable class. The JPA Overview details this process in Section 8.3, “ Embedded Mapping ”. AttributeOverrides suffice for simple mappings, but do not allow you to override complex mappings. Also, JPA has no way to differentitate between a null embedded object and one with default values for all of its fields.

OpenJPA overcomes these shortcomings with the org.apache.openjpa.persistence.jdbc.EmbeddedMapping annotation. This annotation has the following properties:

  • String nullIndicatorColumnName: If the named column's value is NULL, then the embedded object is assumed to be null. If the named column has a non- NULL value, then the embedded object will get loaded and populated with data from the other embedded fields. This property is entirely optional. By default, OpenJPA always assumes the embedded object is non-null, just as in standard JPA mapping.

    If the column you name does not belong to any fields of the embedded object, OpenJPA will create a synthetic null-indicator column with this name. In fact, you can specify a value of true to simply indicate that you want a synthetic null-indicator column, without having to come up with a name for it. A value of false signals that you explicitly do not want a null-indicator column created for this mapping (in case you have configured your mapping defaults to create one by default).

  • String nullIndicatorFieldName: Rather than name a null indicator column, you can name a field of the embedded type. OpenJPA will use the column of this field as the null-indicator column.

  • MappingOverride[] overrides: This array allows you to override any mapping of the embedded object.

The EmbeddedMapping's overrides array serves the same purpose as standard JPA's AttributeOverride s and AssociationOverride s. In fact, you can also use the MappingOverride annotation on an entity class to override a complex mapping of its mapped superclass, just as you can with AttributeOverride and AssociationOverride s. The MappingOverrides annotation, whose value is an array of MappingOverride s, allows you to overide multiple mapped superclass mappings.

Each org.apache.openjpa.persistence.jdbc.MappingOverride annotation has the following properties:

  • String name: The name of the field that is being overridden.

  • Column[] columns: Columns for the new field mapping.

  • XJoinColumn[] joinColumns: Join columns for the new field mapping, if it is a relation field.

  • ContainerTable containerTable: Table for the new collection or map field mapping. We cover collection mappings in Section 7.6, “ Collections ”, and map mappings in Section 7.8, “ Maps ”.

  • ElementJoinColumn[] elementJoinColumns: Element join columns for the new collection or map field mapping. You will see how to use element join columns in Section 7.6.2, “ Element Join Columns ”.

The following example defines an embeddable PathCoordinate class with a custom mapping of a java.awt.Point field to two columns. It then defines an entity which embeds a PointCoordinate and overrides the default mapping for the point field. The entity also declares that if the PathCoordinate 's siteName field column is null, it means that no PathCoordinate is stored in the embedded record; the owning field will load as null.

Example 7.15.  Overriding Complex Mappings

import org.apache.openjpa.persistence.jdbc.*;

@Embeddable
public class PathCoordinate {

    private String siteName;

    @Persistent
    @Strategy("com.xyz.openjpa.PointValueHandler")
    private Point point;

    ...
}

@Entity
public class Path {

    @Embedded
    @EmbeddedMapping(nullIndicatorFieldName="siteName", overrides={
        @MappingOverride(name="siteName", columns=@Column(name="START_SITE")),
        @MappingOverride(name="point", columns={
            @Column(name="START_X"),
            @Column(name="START_Y")
        })
    })
    private PathCoordinate start;

    ...
}

7.6.  Collections

In Section 2.4, “Persistent Collection Fields”, we explored the PersistentCollection annotation for persistent collection fields that aren't a standard OneToMany or ManyToMany relation. To map these non-standard collections, combine OpenJPA's ContainerTable annotation with ElementJoinColumns. We explore the annotations below.

7.6.1.  Container Table

The org.apache.openjpa.persistence.jdbc.ContainerTable annotation describes a database table that holds collection (or map) elements. This annotation has the following properties:

  • String name

  • String catalog

  • String schema

  • XJoinColumn[] joinColumns

  • ForeignKey joinForeignKey

  • Index joinIndex

The name, catalog, schema , and joinColumns properties describe the container table and how it joins to the owning entity's table. These properties correspond to the same-named properties on the standard JoinTable annotation, described in Section 8.5, “ Join Table ” . If left unspecified, the name of the table defaults to the first five characters of the entity table name, plus an underscore, plus the field name. The joinForeignKey and joinIndex properties override default foreign key and index generation for the join columns. We explore foreign keys and indexes later in this chapter.

You may notice that the container table does not define how to store the collection elements. That is left to separate annotations, which are the subject of the next sections.

7.6.2.  Element Join Columns

Element join columns are equivalent to standard JPA join columns, except that they represent a join to a collection or map element entity rather than a direct relation. You represent an element join column with OpenJPA's org.apache.openjpa.persistence.jdbc.ElementJoinColumn annotation. To declare a compound join, enclose an array of ElementJoinColumns in the org.apache.openjpa.persistence.jdbc.ElementJoinColumns annotation.

An ElementJoinColumn always resides in a container table, so it does not have the table property of a standard JoinColumn. Like XJoinColumns above, ElementJoinColumns can reference a linked attribute rather than a static linked column. Otherwise, the ElementJoinColumn and standard JoinColumn annotations are equivalent. See Section 8.4, “ Direct Relations ” in the JPA Overview for a review of the JoinColumn annotation.

7.6.3.  Order Column

Relational databases do not guarantee that records are returned in insertion order. If you want to make sure that your collection elements are loaded in the same order they were in when last stored, you must declare an order column. OpenJPA's org.apache.openjpa.persistence.jdbc.OrderColumn annotation has the following properties:

  • String name: Defaults to ORDR.

  • boolean enabled

  • int precision

  • String columnDefinition

  • boolean insertable

  • boolean updatable

Order columns are always in the container table. You can explicitly turn off ordering (if you have enabled it by default via your mapping defaults) by setting the enabled property to false. All other properties correspond exactly to the same-named properties on the standard Column annotation, described in Section 3, “ Column ”.

7.7.  One-Sided One-Many Mapping

The previous section covered the use of ElementJoinColumn annotations in conjunction with a ContainerTable for mapping collections to dedicate tables. ElementJoinColumn s, however, have one additional use: to create a one-sided one-many mapping. Standard JPA supports OneToMany fields without a mappedBy inverse, but only by mapping these fields to a JoinTable (see Section 8.5, “ Join Table ” in the JPA Overview for details). Often, you'd like to create a one-many association based on an inverse foreign key (logical or actual) in the table of the related type.

Consider the model above. Subscription has a collection of LineItem s, but LineItem has no inverse relation to Subscription. To retrieve all of the LineItem records for a Subscription , we join the SUB_ID inverse foreign key column in the LINE_ITEM table to the primary key column of the SUB table. The example below shows how to represent this model in mapping annotations. Note that OpenJPA automatically assumes an inverse foreign key mapping when element join columns are given, but no container or join table is given.

Example 7.16.  One-Sided One-Many Mapping

package org.mag.subscribe;

import org.apache.openjpa.persistence.jdbc.*;

@Entity
@Table(name="LINE_ITEM", schema="CNTRCT")
public class LineItem {
    ...
}

@Entity
@Table(name="SUB", schema="CNTRCT")
public class Subscription {

    @Id private long id;

    @OneToMany
    @ElementJoinColumn(name="SUB_ID", target="ID")
    private Collection<LineItem> items;

    ...
}

7.8.  Maps

We detailed the ContainerTable annotation in Section 7.6.1, “ Container Table ”. Custom map mappings may also use this annotation to represent a map table.

7.9.  Indexes and Constraints

OpenJPA uses index information during schema generation to index the proper columns. OpenJPA uses foreign key and unique constraint information during schema creation to generate the proper database constraints, and also at runtime to order SQL statements to avoid constraint violations while maximizing SQL batch size.

OpenJPA assumes certain columns have indexes or constraints based on your mapping defaults, as detailed in Section 4, “ Mapping Defaults ”. You can override the configured defaults on individual joins, field values, collection elements, map keys, or map values using the annotations presented in the following sections.

7.9.1.  Indexes

The org.apache.openjpa.persistence.jdbc.Index annotation represents an index on the columns of a field. It is also used within the ContainerTable annotation to index join columns. To index the columns of a collection element, use the org.apache.openjpa.persistence.jdbc.ElementIndex annotation. These annotations have the following properties:

  • boolean enabled: Set this property to false to explicitly tell OpenJPA not to index these columns, when OpenJPA would otherwise do so.

  • String name: The name of the index. OpenJPA will choose a name if you do not provide one.

  • boolean unique: Whether to create a unique index. Defaults to false.

7.9.2.  Foreign Keys

The org.apache.openjpa.persistence.jdbc.ForeignKey annotation represents a foreign key on the columns of a field. It is also used within the ContainerTable annotation to set a database foreign key on join columns. To set a constraint to the columns of a collection element, use the org.apache.openjpa.persistence.jdbc.ElementForeignKey annotation. These annotations have the following properties:

  • boolean enabled: Set this property to false to explicitly tell OpenJPA not to set a foreign key on these columns, when OpenJPA would otherwise do so.

  • String name: The name of the foreign key. OpenJPA will choose a name if you do not provide one, or will create an anonymous key.

  • boolean deferred: Whether to create a deferred key if supported by the database.

  • ForeignKeyAction deleteAction: Value from the org.apache.openjpa.persistence.jdbc.ForeignKeyAction enum identifying the desired delete action. Defaults to RESTRICT.

  • ForeignKeyAction updateAction: Value from the org.apache.openjpa.persistence.jdbc.ForeignKeyAction enum identifying the desired update action. Defaults to RESTRICT.

Keep in mind that OpenJPA uses foreign key information at runtime to avoid constraint violations; it is important, therefore, that your mapping defaults and foreign key annotations combine to accurately reflect your existing database constraints, or that you configure OpenJPA to reflect on your database schema to discover existing foreign keys (see Section 11.2, “ Schema Factory ”).

7.9.3.  Unique Constraints

The org.apache.openjpa.persistence.jdbc.Unique annotation represents a unqiue constraint on the columns of a field. It is more convenient than using the uniqueConstraints property of standard JPA Table and SecondaryTable annotations, because you can apply it directly to the constrained field. The Unique annotation has the following properties:

  • boolean enabled: Set this property to false to explicitly tell OpenJPA not to constrain these columns, when OpenJPA would otherwise do so.

  • String name: The name of the constraint. OpenJPA will choose a name if you do not provide one, or will create an anonymous constraint.

  • boolean deferred: Whether to create a deferred constraint if supported by the database.

8.  Mapping Limitations

The following sections outline the limitations OpenJPA places on specific mapping strategies.

8.1.  Table Per Class

Table-per-class inheritance mapping has the following limitations:

  • You cannot traverse polymorphic relations to non-leaf classes in a table-per-class inheritance hierarchy in queries.

  • You cannot map a one-sided polymorphic relation to a non-leaf class in a table-per-class inheritance hierarchy using an inverse foreign key.

  • You cannot use an order column in a polymorphic relation to a non-leaf class in a table-per-class inheritance hierarchy mapped with an inverse foreign key.

  • Table-per-class hierarchies impose limitations on eager fetching. See Section 7.2, “ Eager Fetching Considerations and Limitations ”.

Note

Non-polymorphic relations do not suffer from these limitations. You can declare a non-polymorphic relation using the extensions described in Section 9.2.2, “ Nonpolymorphic ”.

9.  Mapping Extensions

Mapping extensions allow you to access OpenJPA-specific functionality from your mappings. Note that all extensions below are specific to mappings. If you store your mappings separately from your persistence metadata, these extensions must be specified along with the mapping information, not the persistence metadata information.

9.1.  Class Extensions

OpenJPA recognizes the following class extensions.

9.1.1.  Subclass Fetch Mode

This extension specifies how to eagerly fetch subclass state. It overrides the global openjpa.jdbc.SubclassFetchMode property. Set the JPA org.apache.openjpa.persistence.jdbc.SubclassFetchMode annotation to a value from the org.apache.openjpa.persistence.jdbc.EagerFetchType enum: JOIN, PARALLEL, or NONE. See Section 7, “ Eager Fetching ” for a discussion of eager fetching.

9.1.2.  Strategy

The org.apache.openjpa.persistence.jdbc.Strategy class annotation allows you to specify a custom mapping strategy for your class. See Section 10, “ Custom Mappings ” for information on custom mappings.

9.1.3.  Discriminator Strategy

The org.apache.openjpa.persistence.jdbc.DiscriminatorStrategy class annotation allows you to specify a custom discriminator strategy. See Section 10, “ Custom Mappings ” for information on custom mappings.

9.1.4.  Version Strategy

The org.apache.openjpa.persistence.jdbc.VersionStrategy class annotation allows you to specify a custom version strategy. See Section 10, “ Custom Mappings ” for information on custom mappings.

9.2.  Field Extensions

OpenJPA recognizes the following field extensions.

9.2.1.  Eager Fetch Mode

This extension specifies how to eagerly fetch related objects. It overrides the global openjpa.jdbc.EagerFetchMode property. Set the JPA org.apache.openjpa.persistence.jdbc.EagerFetchMode annotation to a value from the org.apache.openjpa.persistence.jdbc.EagerFetchType enum: JOIN, PARALLEL, or NONE. See Section 7, “ Eager Fetching ” for a discussion of eager fetching.

9.2.2.  Nonpolymorphic

All fields in Java are polymorphic. If you declare a field of type T , you can assign any subclass of T to the field as well. This is very convenient, but can make relation traversal very inefficient under some inheritance strategies. It can even make querying across the field impossible. Often, you know that certain fields do not need to be entirely polymorphic. By telling OpenJPA about such fields, you can improve the efficiency of your relations.

Note

OpenJPA also includes the type metadata extension for narrowing the declared type of a field. See Section 3.2.6, “ Type ”.

OpenJPA defines the following extensions for nonpolymorphic values:

The value of these extensions is a constant from the org.apache.openjpa.persistence.jdbc.NonpolymorphicType enumeration. The default value, EXACT, indicates that the relation will always be of the exact declared type. A value of JOINABLE, on the other hand, means that the relation might be to any joinable subclass of the declared type. This value only excludes table-per-class subclasses.

9.2.3.  Class Criteria

This family of boolean extensions determines whether OpenJPA will use the expected class of related objects as criteria in the SQL it issues to load a relation field. Typically, this is not needed. The foreign key values uniquely identify the record for the related object. Under some rare mappings, however, you may need to consider both foreign key values and the expected class of the related object - for example, if you have an inverse relation that shares the foreign key with another inverse relation to an object of a different subclass. In these cases, set the proper class critera extension to true to force OpenJPA to append class criteria to its select SQL.

OpenJPA defines the following class criteria annotations for field relations and array or collection element relations, respectively:

9.2.4.  Strategy

OpenJPA's org.apache.openjpa.persistence.jdbc.Strategy extension allows you to specify a custom mapping strategy or value handler for a field. See Section 10, “ Custom Mappings ” for information on custom mappings.

10.  Custom Mappings

In OpenJPA, you are not limited to the set of standard mappings defined by the specification. OpenJPA allows you to define custom class, discriminator, version, and field mapping strategies with all the power of OpenJPA's built-in strategies.

10.1.  Custom Class Mapping

To create a custom class mapping, write an implementation of the org.apache.openjpa.jdbc.meta.ClassStrategy interface. You will probably want to extend one of the existing abstract or concrete strategies in the org.apache.openjpa.jdbc.meta.strats package.

The org.apache.openjpa.persistence.jdbc.Strategy annotation allows you to declare a custom class mapping strategy in JPA mapping metadata. Set the value of the annotation to the full class name of your custom strategy. You can configure your strategy class' bean properties using OpenJPA's plugin syntax, detailed in Section 4, “ Plugin Configuration ”.

10.2.  Custom Discriminator and Version Strategies

To define a custom discriminator or version strategy, implement the org.apache.openjpa.jdbc.meta.DiscriminatorStrategy or org.apache.openjpa.jdbc.meta.VersionStrategy interface, respectively. You might extend one of the existing abstract or concrete strategies in the org.apache.openjpa.jdbc.meta.strats package.

OpenJPA includes the org.apache.openjpa.persistence.jdbc.DiscriminatorStrategy and org.apache.openjpa.persistence.jdbc.VersionStrategy class annotations for declaring a custom discriminator or version strategy in JPA mapping metadata. Set the string value of these annotations to the full class name of your implementation, or to the class name or alias of an existing OpenJPA implementation.

As with custom class mappings, you can configure your strategy class' bean properties using OpenJPA's plugin syntax, detailed in Section 4, “ Plugin Configuration ”.

10.3.  Custom Field Mapping

While custom class, discriminator, and version mapping can be useful, custom field mappings are far more common. OpenJPA offers two types of custom field mappings: value handlers, and full custom field strategies. The following sections examine each.

10.3.1.  Value Handlers

Value handlers make it trivial to map any type that you can break down into one or more simple values. All value handlers implement the org.apache.openjpa.jdbc.meta.ValueHandler interface; see its Javadoc for details. Also, examine the built-in handlers in the src/openjpa/jdbc/meta/strats directory of your OpenJPA source distribution. Use these functional implementations as examples when you create your own value handlers.

10.3.2.  Field Strategies

OpenJPA interacts with persistent fields through the org.apache.openjpa.jdbc.meta.FieldStrategy interface. You can implement this interface yourself to create a custom field strategy, or extend one of the existing abstract or concrete strategies in the org.apache.openjpa.jdbc.meta.strats package. Creating a custom field strategy is more difficult than writing a custom value handler, but gives you more freedom in how you interact with the database.

10.3.3.  Configuration

OpenJPA gives you two ways to configure your custom field mappings. The FieldStrategies property of the built-in MappingDefaults implementations allows you to globally associate field types with their corresponding custom value handler or strategy. OpenJPA will automatically use your custom strategies when it encounters a field of the associated type. OpenJPA will use your custom value handlers whenever it encounters a field of the associated type. Section 4, “ Mapping Defaults ” described mapping defaults in detail.

Your other option is to explicitly install a custom value handler or strategy on a particular field. To do so, specify the full name of your implementation class in the proper mapping metadata extension. OpenJPA includes the org.apache.openjpa.persistence.jdbc.Strategy annotation. You can configure the named strategy or handler's bean properties in these extensions using OpenJPA's plugin format (see Section 4, “ Plugin Configuration ”).

11.  Orphaned Keys

Unless you apply database foreign key constraints extensively, it is possible to end up with orphaned keys in your database. For example, suppose Magazinem has a reference to Article a. If you delete a without nulling m's reference, m's database record will wind up with an orphaned key to the non-existent a record.

Note

One way of avoiding orphaned keys is to use dependent fields.

OpenJPA's openjpa.OrphanedKeyAction configuration property controls what action to take when OpenJPA encounters an orphaned key. You can set this plugin string (see Section 4, “ Plugin Configuration ”) to a custom implementation of the org.apache.openjpa.event.OrphanedKeyAction interface, or use one of the built-in options:

Example 7.17.  Custom Logging Orphaned Keys

<property name="openjpa.OrphanedKeyAction" value="log(Channel=Orphans, Level=DEBUG)"/>

Chapter 8.  Deployment

OpenJPA deployment includes choosing a factory deployment strategy, and in a managed environment, optionally integrating with your application server's managed and XA transactions. This chapter examines each aspect of deployment in turn.

1.  Factory Deployment

OpenJPA offers two EntityManagerFactory deployment options.

1.1.  Standalone Deployment

The JPA Overview describes the javax.persistence.Persistence class. You can use Persistence to obtain EntityManagerFactory instances, as demonstrated in Chapter 6, Persistence . OpenJPA also extends Persistence to add additional EntityManagerFactory creation methods. The org.apache.openjpa.persistence.OpenJPAPersistence class Javadoc details these extensions.

After obtaining the factory, you can cache it for all EntityManager creation duties. OpenJPA factories support being bound to JNDI as well.

1.2.  EntityManager Injection

Java EE 5 application servers allow you to inject entity managers into your session beans using the PersistenceContext annotation. See your application server documentation for details.

2.  Integrating with the Transaction Manager

OpenJPA EntityManagers have the ability to automatically synchronize their transactions with an external transaction manager. Whether or not EntityManagers from a given EntityManagerFactory exhibit this behavior by default depends on the transaction type you set for the factory's persistence unit in your persistence.xml file. OpenJPA uses the given transaction type internally to set the openjpa.TransactionMode configuration property. This property accepts the following modes:

  • local: Perform transaction operations locally.

  • managed: Integrate with the application server's managed global transactions.

You can override the global transaction mode setting when you obtain an EntityManager using the EntityManagerFactory's createEntityManager(Map props) method. Simply set the openjpa.TransactionMode key of the given Map to the desired value.

Note

You can also override the openjpa.ConnectionUserName, openjpa.ConnectionPassword, and openjpa.ConnectionRetainMode settings using the given Map.

In order to use global transactions, OpenJPA must be able to access the application server's javax.transaction.TransactionManager. OpenJPA can automatically discover the transaction manager for most major application servers. Occasionally, however, you might have to point OpenJPA to the transaction manager for an unrecognized or non-standard application server setup. This is accomplished through the openjpa.ManagedRuntime configuration property. This property describes an org.apache.openjpa.ee.ManagedRuntime implementation to use for transaction manager discovery. You can specify your own implementation, or use one of the built-ins:

See the Javadoc for of each class for details on the bean properties you can pass to these plugins in your configuration string.

Example 8.1. Configuring Transaction Manager Integration

<property name="openjpa.TransactionMode" value="managed"/>
<property name="openjpa.ManagedRuntime" value="jndi(TransactionManagerName=java:/TransactionManager)"/>

3.  XA Transactions

The X/Open Distributed Transaction Processing (X/Open DTP) model, designed by Open Group (a vendor consortium), defines a standard communication architecture that provides the following:

  • Concurrent execution of applications on shared resources.

  • Coordination of transactions across applications.

  • Components, interfaces, and protocols that define the architecture and provide portability of applications.

  • Atomicity of transaction systems.

  • Single-thread control and sequential function-calling.

The X/Open DTP XA standard defines the application programming interfaces that a resource manager uses to communicate with a transaction manager. The XA interfaces enable resource managers to join transactions, to perform two-phase commit, and to recover in-doubt transactions following a failure.

3.1.  Using OpenJPA with XA Transactions

OpenJPA supports XA-compliant transactions when used in a properly configured managed environment. The following components are required:

  • A managed environment that provides an XA compliant transaction manager. Examples of this are application servers such as WebLogic or JBoss.

  • Instances of a javax.sql.XADataSource for each of the DataSources that OpenJPA will use.

Given these components, setting up OpenJPA to participate in distributed transactions is a simple two-step process:

  1. Integrate OpenJPA with your application server's transaction manager, as detailed in Section 2, “ Integrating with the Transaction Manager ” above.

  2. Point OpenJPA at an enlisted XADataSource, and configure a second non-enlisted data source. See Section 2.1, “ Managed and XA DataSources ”.

Chapter 9.  Runtime Extensions

This chapter describes OpenJPA extensions to the standard JPA interfaces, and outlines some additional features of the OpenJPA runtime.

1.  Architecture

Internally, OpenJPA does not adhere to any persistence specification. The OpenJPA kernel has its own set of APIs and components. Specifications like JPA and JDO are simply different "personalities" that can OpenJPA's native kernel can adopt.

As a OpenJPA user, you will not normally see beneath OpenJPA's JPA personality. OpenJPA allows you to access its feature set without leaving the comfort of JPA. Where OpenJPA goes beyond standard JPA functionality, we have crafted JPA-specific APIs to each OpenJPA extension for as seamless an experience as possible.

When writing OpenJPA plugins or otherwise extending the OpenJPA runtime, however, you will use OpenJPA's native APIs. So that you won't feel lost, the list below associates each specification interface with its backing native OpenJPA component:

  • javax.persistence.EntityManagerFactory: org.apache.openjpa.kernel.BrokerFactory

  • javax.persistence.EntityManager: org.apache.openjpa.kernel.Broker

  • javax.persistence.Query: org.apache.openjpa.kernel.Query

  • org.apache.openjpa.persistence.Extent: org.apache.openjpa.kernel.Extent

  • org.apache.openjpa.persistence.StoreCache: org.apache.openjpa.datacache.DataCache

  • org.apache.openjpa.persistence.QueryResultCache: org.apache.openjpa.datacache.QueryCache

  • org.apache.openjpa.persistence.FetchPlan: org.apache.openjpa.kernel.FetchConfiguration

  • org.apache.openjpa.persistence.Generator: org.apache.openjpa.kernel.Seq

The org.apache.openjpa.persistence.OpenJPAPersistence helper allows you to convert between EntityManagerFactories and BrokerFactories, EntityManagers and Brokers.

1.1.  Broker Finalization

Outside of a Java EE 5 application server or other JPA persistence container environment, the default OpenJPAEntityManager implementation automatically closes itself during instance finalization. This guards against accidental resource leaks that may occur if a developer fails to explicitly close EntityManagers when finished with them, but it also incurs a scalability bottleneck, since the JVM must perform synchronization during instance creation, and since the finalizer thread will have more instances to monitor. To avoid this overhead, set the openjpa.BrokerImpl configuration property to non-finalizing.

1.2.  Broker Customization and Finalization

As a plugin string, this property can be used to configure the BrokerImpl with the following properties:

  • EvictFromDataCache: When evicting an object through the OpenJPAEntityManager.evict methods, whether to also evict it from the OpenJPA's data cache. Defaults to false.

Example 9.1.  Evict from Data Cache

<property name="openjpa.BrokerImpl" value="EvictFromDataCache=true"/>

Additionally, some advanced users may want to add capabilities to OpenJPA's internal org.apache.openjpa.kernel.BrokerImpl. You can configure OpenJPA to use a custom subclass of BrokerImpl with the openjpa.BrokerImpl configuration property. Set this property to the full class name of your custom subclass. When implementing your subclass, consider the finalization issues mentioned in ???. It may be appropriate to create a subtype of both org.apache.openjpa.kernel.BrokerImpl and org.apache.openjpa.kernel.FinalizingBrokerImpl.

2.  JPA Extensions

The following sections outline the runtime interfaces you can use to access OpenJPA-specific functionality from JPA. Each interface contains services and convenience methods missing from the JPA specification. OpenJPA strives to use the same naming conventions and API patterns as standard JPA methods in all extensions, so that OpenJPA JDO APIs feel as much as possible like standard JPA.

You may have noticed the examples throughout this document using the OpenJPAPersistence.cast methods to cast from standard JPA interfaces to OpenJPA extended interfaces. This is the recommended practice. Some application server vendors may proxy OpenJPA's JPA implementation, preventing a straight cast. OpenJPAPersistence's cast methods work around these proxies.

 
public static OpenJPAEntityManagerFactory cast(EntityManagerFactory emf);
public static OpenJPAEntityManager cast(EntityManager em);
public static OpenJPAQuery cast(Query q);

We provide additional information on the OpenJPAPersistence helper below.

2.1.  OpenJPAEntityManagerFactory

The org.apache.openjpa.persistence.OpenJPAEntityManagerFactory interface extends the basic javax.persistence.EntityManagerFactory with OpenJPA-specific features. The OpenJPAEntityManagerFactory offers APIs to access the OpenJPA data and query caches and to perform other OpenJPA-specific operations. See the interface Javadoc for details.

2.2.  OpenJPAEntityManager

All OpenJPA EntityManagers implement the org.apache.openjpa.persistence.OpenJPAEntityManager interface. This interface extends the standard javax.persistence.EntityManager. Just as the standard EntityManager is the primary window into JPA services, the OpenJPAEntityManager is the primary window from JPA into OpenJPA-specific functionality. We strongly encourage you to investigate the API extensions this interface contains.

2.3.  OpenJPAQuery

OpenJPA extends JPA's standard query functionality with the org.apache.openjpa.persistence.OpenJPAQuery interface. See its Javadoc for details on the convenience methods it provides.

2.4.  Extent

An Extent is a logical view of all persistent instances of a given entity class, possibly including subclasses. OpenJPA adds the org.apache.openjpa.persistence.Extent class to the set of Java Persistence APIs. The following code illustrates iterating over all instances of the Magazine entity, without subclasses:

Example 9.2.  Using a JPA Extent

import org.apache.openjpa.persistence.*;

...

OpenJPAEntityManager kem = OpenJPAPersistence.cast(em);
Extent<Magazine> mags = kem.getExtent(Magazine.class, false);
for (Magazine m : mags)
    processMagazine (m);

2.5.  StoreCache

In addition to the EntityManager object cache mandated by the JPA specification, OpenJPA includes a flexible datastore-level cache. You can access this cache from your JPA code using the org.apache.openjpa.persistence.StoreCache facade. Section 1, “ Data Cache ” has detailed information on OpenJPA's data caching system, including the StoreCache facade.

2.6.  QueryResultCache

OpenJPA can cache query results as well as persistent object data. The org.apache.openjpa.persistence.QueryResultCache is an JPA-flavored facade to OpenJPA's internal query cache. See Section 1.3, “ Query Cache ” for details on query caching in OpenJPA.

2.7.  FetchPlan

Many of the aforementioned OpenJPA interfaces give you access to an org.apache.openjpa.persistence.FetchPlan instance. The FetchPlan allows you to exercise some control over how objects are fetched from the datastore, including large result set support, custom fetch groups, and lock levels.

OpenJPA goes one step further, extending FetchPlan with org.apache.openjpa.persistence.jdbc.JDBCFetchPlan to add additional JDBC-specific tuning methods. Unless you have customized OpenJPA to use a non-relational back-end (see Section 8, “ Non-Relational Stores ” ), all FetchPlans in OpenJPA implement JDBCFetchPlan, so feel free to cast to this interface.

Fetch plans pass on from parent components to child components. The EntityManagerFactory settings (via your configuration properties) for things like the fetch size, result set type, and custom fetch groups are passed on to the fetch plan of the EntityManagers it produces. The settings of each EntityManager, in turn, are passed on to each Query and Extent it returns. Note that the opposite, however, is not true. Modifying the fetch plan of a Query or Extent does not affect the EntityManager's configuration. Likewise, modifying an EntityManager's configuration does not affect the EntityManagerFactory.

Section 6, “ Fetch Groups ” includes examples using FetchPlans.

2.8.  OpenJPAPersistence

org.apache.openjpa.persistence.OpenJPAPersistence is a static helper class that adds OpenJPA-specific utility methods to javax.persistence.Persistence.

3.  Object Locking

Controlling how and when objects are locked is an important part of maximizing the performance of your application under load. This section describes OpenJPA's APIs for explicit locking, as well as its rules for implicit locking.

3.1.  Configuring Default Locking

You can control OpenJPA's default transactional read and write lock levels through the openjpa.ReadLockLevel and openjpa.WriteLockLevel configuration properties. Each property accepts a value of none, read, write, or a number corresponding to a lock level defined by the lock manager in use. These properties apply only to non-optimistic transactions; during optimistic transactions, OpenJPA never locks objects by default.

You can control the default amount of time OpenJPA will wait when trying to obtain locks through the openjpa.LockTimeout configuration property. Set this property to the number of milliseconds you are willing to wait for a lock before OpenJPA will throw an exception, or to -1 for no limit. It defaults to -1.

Example 9.3.  Setting Default Lock Levels

<property name="openjpa.ReadLockLevel" value="none"/>
<property name="openjpa.WriteLockLevel" value="write"/>
<property name="openjpa.LockTimeout" value="30000"/>

3.2.  Configuring Lock Levels at Runtime

At runtime, you can override the default lock levels through the FetchPlan interface described above. At the beginning of each datastore transaction, OpenJPA initializes the EntityManager 's fetch plan with the default lock levels and timeouts described in the previous section. By changing the fetch plan's locking properties, you can control how objects loaded at different points in the transaction are locked. You can also use the fetch plan of an individual Query to apply your locking changes only to objects loaded through that Query.

public LockModeType getReadLockMode();
public FetchPlan setReadLockMode(LockModeType mode);
public LockModeType getWriteLockMode();
public FetchPlan setWriteLockMode(LockModeType mode);
long getLockTimeout();
FetchPlan setLockTimeout(long timeout);

Controlling locking through these runtime APIs works even during optimistic transactions. At the end of the transaction, OpenJPA resets the fetch plan's lock levels to none. You cannot lock objects outside of a transaction.

Example 9.4.  Setting Runtime Lock Levels

import org.apache.openjpa.persistence.*;

...

EntityManager em = ...;
em.getTransaction().begin();

// load stock we know we're going to update at write lock mode
Query q = em.createQuery("select s from Stock s where symbol = :s");
q.setParameter("s", symbol);
OpenJPAQuery oq = OpenJPAPersistence.cast(q);
FetchPlan fetch = oq.getFetchPlan ();
fetch.setReadLockMode(LockModeType.WRITE);
fetch.setLockTimeout(3000); // 3 seconds
Stock stock = (Stock) q.getSingleResult();

// load an object we don't need locked at none lock mode
fetch = OpenJPAPersistence.cast(em).getFetchPlan();
fetch.setReadLockMode(null);
Market market = em.find(Market.class, marketId);

stock.setPrice(market.calculatePrice(stock));
em.getTransaction().commit();

3.3.  Object Locking APIs

In addition to allowing you to control implicit locking levels, OpenJPA provides explicit APIs to lock objects and to retrieve their current lock level.

public LockModeType OpenJPAEntityManager.getLockMode(Object pc);

Returns the level at which the given object is currently locked.

In addition to the standard EntityManager.lock(Object, LockModeType) method, the OpenJPAEntityManager exposes the following methods to lock objects explicitly:

public void lock(Object pc);
public void lock(Object pc, LockModeType mode, long timeout);
public void lockAll(Object... pcs);
public void lockAll(Object... pcs, LockModeType mode, long timeout);
public void lockAll(Collection pcs);
public void lockAll(Collection pcs, LockModeType mode, long timeout);

Methods that do not take a lock level or timeout parameter default to the current fetch plan. The example below demonstrates these methods in action.

Example 9.5.  Locking APIs

import org.apache.openjpa.persistence.*;

// retrieve the lock level of an object
OpenJPAEntityManager oem = OpenJPAPersistence.cast(em);
Stock stock = ...;
LockModeType level = oem.getLockMode(stock);
if (level == OpenJPAModeType.WRITE) ...

...

oem.setOptimistic(true);
oem.getTransaction().begin ();

// override default of not locking during an opt trans to lock stock object
oem.lock(stock, LockModeType.WRITE, 1000);
stock.setPrice(market.calculatePrice(stock));

oem.getTransaction().commit();

3.4.  Lock Manager

OpenJPA delegates the actual work of locking objects to the system's org.apache.openjpa.kernel.LockManager. This plugin is controlled by the openjpa.LockManager configuration property. You can write your own lock manager, or use one of the bundled options:

  • pessimistic: This is an alias for the o rg.apache.openjpa.jdbc.kernel.PessimisticLockManager, which uses SELECT FOR UPDATE statements (or the database's equivalent) to lock the database rows corresponding to locked objects. This lock manager does not distinguish between read locks and write locks; all locks are write locks.

    The pessimistic LockManager can be configued to additionally perform the version checking and incrementing behavior of the version lock manager described below by setting its VersionCheckOnReadLock and VersionUpdateOnWriteLock properties:

    <property name="openjpa.LockManager" value="pessimistic(VersionCheckOnReadLock=true,VersionUpdateOnWriteLock=true)"/>
    
  • none: An alias for the org.apache.openjpa.kernel.NoneLockManager, which does not perform any locking at all.

  • version: An alias for the org.apache.openjpa.kernel.VersionLockManager. This lock manager does not perform any exclusive locking, but instead ensures read consistency by verifying that the version of all read-locked instances is unchanged at the end of the transaction. Furthermore, a write lock will force an increment to the version at the end of the transaction, even if the object is not otherwise modified. This ensures read consistency with non-blocking behavior.

    This is the default openjpa.LockManager setting in JPA.

Note

In order for the version lock manager to prevent the dirty read phenomenon, the underlying data store's transaction isolation level must be set to the equivalent of "read committed" or higher.

Example 9.6.  Disabling Locking

<property name="openjpa.LockManager" value="none"/>

3.5.  Rules for Locking Behavior

Advanced persistence concepts like lazy-loading and object uniquing create several locking corner-cases. The rules below outline OpenJPA's implicit locking behavior in these cases.

  1. When an object's state is first read within a transaction, the object is locked at the fetch plan's current read lock level. Future reads of additional lazy state for the object will use the same read lock level, even if the fetch plan's level has changed.

  2. When an object's state is first modified within a transaction, the object is locked at the write lock level in effect when the object was first read, even if the fetch plan's level has changed. If the object was not read previously, the current write lock level is used.

  3. When objects are accessed through a persistent relation field, the related objects are loaded with the fetch plan's current lock levels, not the lock levels of the object owning the field.

  4. Whenever an object is accessed within a transaction, the object is re-locked at the current read lock level. The current read and write lock levels become those that the object "remembers" according to rules one and two above.

  5. If you lock an object explicitly through the APIs demonstrated above, it is re-locked at the specified level. This level also becomes both the read and write level that the object "remembers" according to rules one and two above.

  6. When an object is already locked at a given lock level, re-locking at a lower level has no effect. Locks cannot be downgraded during a transaction.

3.6.  Known Issues and Limitations

Due to performance concerns and database limitations, locking cannot be perfect. You should be aware of the issues outlined in this section, as they may affect your application.

  • Typically, during optimistic transactions OpenJPA does not start an actual database transaction until you flush or the optimistic transaction commits. This allows for very long-lived transactions without consuming database resources. When using the pessimistic lock manager, however, OpenJPA must begin a database transaction whenever you decide to lock an object during an optimistic transaction. This is because the pessimistic lock manager uses database locks, and databases cannot lock rows without a transaction in progress. OpenJPA will log an INFO message to the openjpa.Runtime logging channel when it begins a datastore transaction just to lock an object.

  • In order to maintain reasonable performance levels when loading object state, OpenJPA can only guarantee that an object is locked at the proper lock level after the state has been retrieved from the database. This means that it is technically possible for another transaction to "sneak in" and modify the database record after OpenJPA retrieves the state, but before it locks the object. The only way to positively guarantee that the object is locked and has the most recent state to refresh the object after locking it.

    When using the pessimistic lock manager, the case above can only occur when OpenJPA cannot issue the state-loading SELECT as a locking statement due to database limitations. For example, some databases cannot lock SELECTs that use joins. The pessimistic lock manager will log an INFO message to the openjpa.Runtime logging channel whenever it cannot lock the initial SELECT due to database limitations. By paying attention to these log messages, you can see where you might consider using an object refresh to guarantee that you have the most recent state, or where you might rethink the way you load the state in question to circumvent the database limitations that prevent OpenJPA from issuing a locking SELECT in the first place.

4.  Savepoints

Savepoints allow for fine grained control over the transactional behavior of your application. OpenJPA's savepoint API allow you to set intermediate rollback points in your transaction. You can then choose to rollback changes made only after a specific savepoint, then commit or continue making new changes in the transaction. This feature is useful for multi-stage transactions, such as editing a set of objects over several web pages or user screens. Savepoints also provide more flexibilty to conditional transaction behavior, such as choosing to commit or rollback a portion of the transaction based on the results of the changes. This chapter describes how to use and configure OpenJPA savepoints.

4.1.  Using Savepoints

OpenJPA's OpenJPAEntityManager have the following methods to control savepoint behavior. Note that the savepoints work in tandem with the current transaction. This means that savepoints require an open transaction, and that a rollback of the transaction will rollback all of the changes in the transaction regardless of any savepoints set.

void setSavepoint(String name);
void releaseSavepoint(String name);
void rollbackToSavepoint(String name);

To set a savepoint, simply call setSavepoint, passing in a symbolic savepoint name. This savepoint will define a point at which you can preserve the state of transactional objects for the duration of the current transaction.

Having set a named savepoint, you can rollback changes made after that point by calling rollbackToSavepoint. This method will keep the current transaction active, while restoring all transactional instances back to their saved state. Instances that were deleted after the save point will no longer be marked for deletion. Similarly, transient instances that were made persistent after the savepoint will become transient again. Savepoints made after this savepoint will be released and no longer valid, although you can still set new savepoints. Savepoints will also be cleared after the current transaction is committed or rolled back.

If a savepoint is no longer needed, you can release any resources it is consuming resources by calling releaseSavepoint. This method should not be called for savepoints that have been released automatically through other means, such as commit of a transaction or rollback to a prior savepoint. While savepoints made after this savepoint will also be released, there are no other effects on the current transaction.

The following simple example illustrates setting, releasing, and rolling back to a savepoint.

Example 9.7.  Using Savepoints

import org.apache.openjpa.persistence.*;

...

OpenJPAEntityManager oem = OpenJPAPersistence.cast(em);
oem.getTransaction().begin();

Magazine mag = oem.find(Magazine.class, id);
mag.setPageCount(300);
oem.setSavepoint("pages");

mag.setPrice(mag.getPageCount() * pricePerPage);
// we decide to release "pages"...
oem.releaseSavepoint("pages");
// ... and set a new savepoint which includes all changes
oem.setSavepoint("price");

mag.setPrice(testPrice);
// we determine the test price is not good
oem.rollbackToSavepoint("price");
// had we chosen to not release "pages", we might have rolled back to
// "pages" instead

// the price is now restored to mag.getPageCount() * pricePerPage
oem.getTransaction().commit();

4.2.  Configuring Savepoints

OpenJPA uses the org.apache.openjpa.kernel.SavepointManager plugin to handle perserving the savepoint state. OpenJPA includes the following SavepointManager plugins:

  • in-mem: The default. This is an alias for the org.apache.openjpa.kernel.InMemorySavepointManager. This plugin stores all state, including field values, in memory. Due to this behavior, each set savepoint is designed for small to medium transactional object counts.

  • jdbc: This is an alias for the org.apache.openjpa.jdbc.kernel.JDBCSavepointManager. This plugin requires JDBC 3 and java.sql.Savepoint support to operate. Note that this plugin implements savepoints by issuing a flush to the database.

  • oracle: This is an alias for the org.apache.openjpa.jdbc.sql.OracleSavepointManager. This plugin operates similarly to the JDBC plugin; however, it uses Oracle-specific calls. This plugin requires using the Oracle JDBC driver and database, versions 9.2 or higher. Note that this plugin implements savepoints by issuing a flush to the database.

5.  MethodQL

If JPQL and SQL queries do not match your needs, OpenJPA also allows you to name a Java method to use to load a set of objects. In a MethodQL query, the query string names a static method to invoke to determine the matching objects:

import org.apache.openjpa.persistence.*;

...

// the method query language is 'openjpa.MethodQL'.
// set the query string to the method to execute, including full class name; if
// the class is in the candidate class' package or in the query imports, you
// can omit the package; if the method is in the candidate class, you can omit
// the class name and just specify the method name
OpenJPAEntityManager oem = OpenJPAPersistence.cast(emf);
OpenJPAQuery q = oem.createQuery("openjpa.MethodQL", "com.xyz.Finder.getByName");

// set the type of objects that the method returns
q.setResultClass(Person.class);

// parameters are passed the same way as in standard queries
q.setParameter("firstName", "Fred").setParameter("lastName", "Lucas");

// this executes your method to get the results
List results = q.getResultList();

For datastore queries, the method must have the following signature:

public static ResultObjectProvider xxx(StoreContext ctx, ClassMetaData meta, boolean subclasses, Map params, FetchConfiguration  fetch)

The returned result object provider should produce objects of the candidate class that match the method's search criteria. If the returned objects do not have all fields in the given fetch configuration loaded, OpenJPA will make additional trips to the datastore as necessary to fill in the data for the missing fields.

In-memory execution is slightly different, taking in one object at a time and returning a boolean on whether the object matches the query:

public static boolean xxx(StoreContext ctx, ClassMetaData meta, boolean subclasses, Object obj, Map params, FetchConfiguration fetch)

In both method versions, the given params map contains the names and values of all the parameters for the query.

6.  Generators

The JPA Overview's Chapter 12, Mapping Metadata details using generators to automatically populate identity fields in JPA.

OpenJPA represents all generators internally with the org.apache.openjpa.kernel.Seq interface. This interface supplies all the context you need to create your own custom generators, including the current persistence environment, the JDBC DataSource , and other essentials. The org.apache.openjpa.jdbc.kernel.AbstractJDBCSeq helps you create custom JDBC-based sequences. OpenJPA also supplies the following built-in Seqs:

  • table: This is OpenJPA's default implementation. It is an alias for the org.apache.openjpa.jdbc.kernel.TableJDBCSeq class. The TableJDBCSeq uses a special single-row table to store a global sequence number. If the table does not already exist, it is created the first time you run the mapping tool on a class that requires it. You can also use the class's main method to manipulate the table; see the TableJDBCSeq.main method Javadoc for usage details.

    This Seq has the following properties:

    • Table: The name of the sequence number table to use. Defaults to OPENJPA_SEQUENCE_TABLE.

    • PrimaryKeyColumn: The name of the primary key column for the sequence table. Defaults to ID.

    • SequenceColumn: The name of the column that will hold the current sequence value. Defaults to SEQUENCE_VALUE.

    • Allocate: The number of values to allocate on each database trip. Defaults to 50, meaning the class will set aside the next 50 numbers each time it accesses the sequence table, which in turn means it only has to make a database trip to get new sequence numbers once every 50 sequence number requests.

  • class-table: This is an alias for the org.apache.openjpa.jdbc.kernel.ClassTableJDBCSeq . This Seq is like the TableJDBCSeq above, but maintains a separate table row, and therefore a separate sequence number, for each base persistent class. It has all the properties of the TableJDBCSeq. Its table name defaults to OPENJPA_SEQUENCES_TABLE. It also adds the following properties:

    • IgnoreUnmapped: Whether to ignore unmapped base classes, and instead use one row per least-derived mapped class. Defaults to false.

    • UseAliases: Whether to use each class' entity name as the primary key value of each row, rather than the full class name. Defaults to false.

    As with the TableJDBCSeq, the ClassTableJDBCSeq creates its table automatically during mapping tool runs. However, you can manually manipulate the table through the class' main method. See the Javadoc for the ClassTableJDBCSeq.main method for usage details.

  • value-table: This is an alias for the org.apache.openjpa.jdbc.kernel.ValueTableJDBCSeq . This Seq is like the ClassTableJDBCSeq above, but has an arbitrary number of rows for sequence values, rather than a fixed pattern of one row per class. Its table defaults to OPENJPA_SEQUENCES_TABLE. It has all the properties of the TableJDBCSeq, plus:

    • PrimaryKeyValue: The primary key value used by this instance.

    As with the TableJDBCSeq, the ValueTableJDBCSeq creates its table automatically during mapping tool runs. However, you can manually manipulate the table through the class' main method. See the Javadoc for the ValueTableJDBCSeq.main method for usage details.

  • native: This is an alias for the org.apache.openjpa.jdbc.kernel.NativeJDBCSeq. Many databases have a concept of "native sequences" - a built-in mechanism for obtaining incrementing numbers. For example, in Oracle, you can create a database sequence with a statement like CREATE SEQUENCE MYSEQUENCE . Sequence values can then be atomically obtained and incremented with the statement SELECT MYSEQUENCE.NEXTVAL FROM DUAL. OpenJPA provides support for this common mechanism of sequence generation with the NativeJDBCSeq, which accepts the following properties:

    • Sequence: The name of the database sequence. Defaults to OPENJPA_SEQUENCE.

    • InitialValue: The initial sequence value. Defaults to 1.

    • Increment: The amount the sequence increments. Defaults to 1.

    • Allocate: Some database can allocate values in-memory to service subsequent sequence requests faster.

  • time: This is an alias for the org.apache.openjpa.kernel.TimeSeededSeq. This type uses an in-memory static counter, initialized to the current time in milliseconds and monotonically incremented for each value requested. It is only suitable for single-JVM environments.

You can use JPA SequenceGenerators to describe any built-in Seqs or your own Seq implementation. Set the sequenceName attribute to a plugin string describing your choice. See Section 5, “ Generators ” in the JPA Overview for details on defining SequenceGenerators.

See Section 4, “ Plugin Configuration ” for plugin string formatting.

Example 9.8.  Named Seq Sequence

@Entity
@Table(name="AUTO")
public class Author {

    @Id
    @GeneratedValue(strategy=GenerationType.SEQUENCE, generator="AuthorSeq")
    @SequenceGenerator(name="AuthorSeq" sequence="table(Table=AUTO_SEQ, Increment=100)")
    @Column(name="AID")
    private long id;
 
    ...
}

Note that if you want to use a plugin string without any arguments, you must still suffix the plugin type with () to differentiate it from a sequence name in the SequenceGenerator.sequence attribute:

@SequenceGenerator(name="AuthorSeq", sequence="table()")

OpenJPA maintains a system sequence to generate datastore identity values for classes that do not declare a specific datastore identity strategy. You can configure the system sequence through the openjpa.Sequence configuration property. This property accepts a plugin string describing a Seq instance.

Example 9.9.  System Sequence Configuration

<property name="openjpa.Sequence" value="table(Table=OPENJPASEQ, Increment=100)"/>

In JPA, set your GeneratedValue annotation's strategy attribute to AUTO to use the configured system sequence. Or, because AUTO is the default strategy, use the annotation without attributes:

@GeneratedValue
private long id;

6.1.  Runtime Access

OpenJPA allows you to access named generators at runtime through the OpenJPAEntityManager.getNamedGenerator method:

public Generator getNamedGenerator(String name);

The returned org.apache.openjpa.persistence.Generator is a facade over an internal OpenJPA Seq.

The OpenJPAEntityManager includes additional APIs to retrieve the identity generator of any class, or the generator of any field. With these APIs, you do not have to know the generator name. Additionally, they allow you to access the implicit generator used by default for datastore identity classes. See the Javadoc for the OpenJPAEntityManager.getIdentityGenerator and OpenJPAEntityManager.getFieldGenerator methods for API details.

7.  Transaction Events

The OpenJPA runtime supports broadcasting transaction-related events. By registering one or more org.apache.openjpa.event.TransactionListener s, you can receive notifications when transactions begin, flush, rollback, commit, and more. Where appropriate, event notifications include the set of persistence-capable objects participating in the transaction.

public void addTransactionListener(Object listener);
public void removeTransactionListener(Object listener);

These OpenJPAEntityManager methods allow you to add and remove listeners.

For details on the transaction framework, see the org.apache.openjpa.event package Javadoc. Also see Section 2, “ Remote Event Notification Framework ” for a description of OpenJPA's remote event support.

8.  Non-Relational Stores

It is possible to adapt OpenJPA to access a non-relational datastore by creating an implementation of the org.apache.openjpa.kernel.StoreManager interface. OpenJPA provides an abstract StoreManager implementation to facilitate this process. See the org.apache.openjpa.abstractstore package Javadoc for details.

Chapter 10.  Caching

OpenJPA utilizes several configurable caches to maximize performance. This chapter explores OpenJPA's data cache, query cache, and query compilation cache.

1.  Data Cache

The OpenJPA data cache is an optional cache of persistent object data that operates at the EntityManagerFactory level. This cache is designed to significantly increase performance while remaining in full compliance with the JPA standard. This means that turning on the caching option can transparently increase the performance of your application, with no changes to your code.

OpenJPA's data cache is not related to the EntityManager cache dictated by the JPA specification. The JPA specification mandates behavior for the EntityManager cache aimed at guaranteeing transaction isolation when operating on persistent objects.

OpenJPA's data cache is designed to provide significant performance increases over cacheless operation, while guaranteeing that behavior will be identical in both cache-enabled and cacheless operation.

There are five ways to access data via the OpenJPA APIs: standard relation traversal, large result set relation traversal, queries, looking up an object by id, and iteration over an Extent. OpenJPA's cache plugin accelerates three of these mechanisms. It does not provide any caching of large result set relations or Extent iterators. If you find yourself in need of higher-performance Extent iteration, see Example 10.13, “ Query Replaces Extent ”.

Table 10.1.  Data access methods

Access methodUses cache
Standard relation traversal Yes
Large result set relation traversal No
QueryYes
Lookups by object id Yes
Iteration over an Extent No

When enabled, the cache is checked before making a trip to the datastore. Data is stored in the cache when objects are committed and when persistent objects are loaded from the datastore.

OpenJPA's data cache can in both single-JVM and multi-JVM environments. Multi-JVM caching is achieved through the use of the distributed event notification framework described in Section 2, “ Remote Event Notification Framework ”, or through custom integrations with a third-party distributed cache.

The single JVM mode of operation maintains and shares a data cache across all EntityManager instances obtained from a particular EntityManagerFactory. This is not appropriate for use in a distributed environment, as caches in different JVMs or created from different EntityManagerFactory objects will not be synchronized.

1.1.  Data Cache Configuration

To enable the basic single-factory cache set the openjpa.DataCache property to true, and set the openjpa.RemoteCommitProvider property to sjvm :

Example 10.1.  Single-JVM Data Cache

<property name="openjpa.DataCache" value="true"/>
<property name="openjpa.RemoteCommitProvider" value="sjvm"/>

To configure the data cache to remain up-to-date in a distributed environment, set the openjpa.RemoteCommitProvider property appropriately, or integrate OpenJPA with a third-party caching solution. Remote commit providers are described in Section 2, “ Remote Event Notification Framework ”.

OpenJPA's default implementation maintains a map of object ids to cache data. By default, 1000 elements are kept in cache. When the cache overflows, random entries are evicted. The maximum cache size can be adjusted by setting the CacheSize property in your plugin string - see below for an example. Objects that are pinned into the cache are not counted when determining if the cache size exceeds its maximum size.

Expired objects are moved to a soft reference map, so they may stick around for a little while longer. You can control the number of soft references OpenJPA keeps with the SoftReferenceSize property. Soft references are unlimited by default. Set to 0 to disable soft references completely.

Example 10.2.  Data Cache Size

<property name="openjpa.DataCache" value="true(CacheSize=5000, SoftReferenceSize=0)"/>

You can specify a cache timeout value for a class by setting the timeout metadata extension to the amount of time in milliseconds a class's data is valid. Use a value of -1 for no expiration. This is the default value.

Example 10.3.  Data Cache Timeout

Timeout Employee objects after 10 seconds.

@Entity
@DataCache(timeout=10000)
public class Employee {
    ...
}

See the org.apache.openjpa.persistence.DataCache Javadoc for more information on the DataCache annotation.

A cache can specify that it should be cleared at certain times rather than using data timeouts. The EvictionSchedule property of OpenJPA's cache implementation accepts a cron style eviction schedule. The format of this property is a whitespace-separated list of five tokens, where the * symbol (asterisk), indicates match all. The tokens are, in order:

  • Minute

  • Hour of Day

  • Day of Month

  • Month

  • Day of Week

For example, the following openjpa.DataCache setting schedules the default cache to evict values from the cache at 15 and 45 minutes past 3 PM on Sunday.

true(EvictionSchedule='15,45 15 * * 1')

1.2.  Data Cache Usage

The org.apache.openjpa.datacache package defines OpenJPA's data caching framework. While you may use this framework directly (see its Javadoc for details), its APIs are meant primarily for service providers. In fact, Section 1.4, “ Cache Extension ” below has tips on how to use this package to extend OpenJPA's caching service yourself.

Rather than use the low-level org.apache.openjpa.datacache package APIs, JPA users should typically access the data cache through OpenJPA's high-level org.apache.openjpa.persistence.StoreCache facade. This facade has methods to pin and unpin records, evict data from the cache, and more.

public StoreCache getStoreCache();

You obtain the StoreCache through the OpenJPAEntityManagerFactory.getStoreCache method.

Example 10.4.  Accessing the StoreCache

import org.apache.openjpa.persistence.*;

...

OpenJPAEntityManagerFactory oemf = OpenJPAPersistence.cast(emf);
StoreCache cache = oemf.getStoreCache();
...
public void evict(Class cls, Object oid);
public void evictAll();
public void evictAll(Class cls, Object... oids);
public void evictAll(Class cls, Collection oids);

The evict methods tell the cache to release data. Each method takes an entity class and one or more identity values, and releases the cached data for the corresponding persistent instances. The evictAll method with no arguments clears the cache. Eviction is useful when the datastore is changed by a separate process outside OpenJPA's control. In this scenario, you typically have to manually evict the data from the datastore cache; otherwise the OpenJPA runtime, oblivious to the changes, will maintain its stale copy.

public void pin(Class cls, Object oid);
public void pinAll(Class cls, Object... oids);
public void pinAll(Class cls, Collection oids);
public void unpin(Class cls, Object oid);
public void unpinAll(Class cls, Object... oids);
public void unpinAll(Class cls, Collection oids);

Most caches are of limited size. Pinning an identity to the cache ensures that the cache will mill not kick the data for the corresponding instance out of the cache, unless you manually evict it. Note that even after manual eviction, the data will get pinned again the next time it is fetched from the store. You can only remove a pin and make the data once again available for normal cache overflow eviction through the unpin methods. Use pinning when you want a guarantee that a certain object will always be available from cache, rather than requiring a datastore trip.

Example 10.5.  StoreCache Usage

import org.apache.openjpa.persistence.*;

...

OpenJPAEntityManagerFactory oemf = OpenJPAPersistence.cast(emf);
StoreCache cache = oemf.getStoreCache();
cache.pin(Magazine.class, popularMag.getId());
cache.evict(Magazine.class, changedMag.getId());

See the StoreCache Javadoc for information on additional functionality it provides. Also, Chapter 9, Runtime Extensions discusses OpenJPA's other extensions to the standard set of JPA runtime interfaces.

The examples above include calls to evict to manually remove data from the data cache. Rather than evicting objects from the data cache directly, you can also configure OpenJPA to automatically evict objects from the data cache when you use the OpenJPAEntityManager's eviction APIs.

Example 10.6.  Automatic Data Cache Eviction

<property name="openjpa.BrokerImpl" value="EvictFromDataCache=true"/>
import org.apache.openjpa.persistence.*;

...

OpenJPAEntityManager oem = OpenJPAPersistence.cast(em);
oem.evict(changedMag);  // will evict from data cache also

1.3.  Query Cache

In addition to the data cache, the org.apache.openjpa.datacache package defines service provider interfaces for a query cache. The query cache is enabled by default when the data cache is enabled. The query cache stores the object ids returned by query executions. When you run a query, OpenJPA assembles a key based on the query properties and the parameters used at execution time, and checks for a cached query result. If one is found, the object ids in the cached result are looked up, and the resultant persistence-capable objects are returned. Otherwise, the query is executed against the database, and the object ids loaded by the query are put into the cache. The object id list is not cached until the list returned at query execution time is fully traversed.

OpenJPA exposes a high-level interface to the query cache through the org.apache.openjpa.persistence.QueryResultCache class. You can access this class through the OpenJPAEntityManagerFactory.

Example 10.7.  Accessing the QueryResultCache

import org.apache.openjpa.persistence.*;

...

OpenJPAEntityManagerFactory oemf = OpenJPAPersistence.cast(emf);
QueryResultCache qcache = oemf.getQueryResultCache();

The default query cache implementation caches 100 query executions in a least-recently-used cache. This can be changed by setting the cache size in the CacheSize plugin property. Like the data cache, the query cache also has a backing soft reference map. The SoftReferenceSize property controls the size of this map. It is disabled by default.

Example 10.8.  Query Cache Size

<property name="openjpa.QueryCache" value="CacheSize=1000, SoftReferenceSize=100"/>

To disable the query cache completely, set the openjpa.QueryCache property to false:

Example 10.9.  Disabling the Query Cache

<property name="openjpa.QueryCache" value="false"/>

There are certain situations in which the query cache is bypassed:

  • Caching is not used for in-memory queries (queries in which the candidates are a collection instead of a class or Extent).

  • Caching is not used in transactions that have IgnoreChanges set to false and in which modifications to classes in the query's access path have occurred. If none of the classes in the access path have been touched, then cached results are still valid and are used.

  • Caching is not used in pessimistic transactions, since OpenJPA must go to the database to lock the appropriate rows.

  • Caching is not used when the the data cache does not have any cached data for an id in a query result.

  • Queries that use persistence-capable objects as parameters are only cached if the parameter is directly compared to field, as in:

    select e from Employee e where e.company.address = :addr
    

    If you extract field values from the parameter in your query string, or if the parameter is used in collection element comparisons, the query is not cached.

  • Queries that result in projections of custom field types or BigDecimal or BigInteger fields are not cached.

Cache results are removed from the cache when instances of classes in a cached query's access path are touched. That is, if a query accesses data in class A, and instances of class A are modified, deleted, or inserted, then the cached query result is dropped from the cache.

It is possible to tell the query cache that a class has been altered. This is only necessary when the changes occur via direct modification of the database outside of OpenJPA's control. You can also evict individual queries, or clear the entire cache.

public void evict(Query q);
public void evictAll(Class cls);
public void evictAll();

For JPA queries with parameters, set the desired parameter values into the Query instance before calling the above methods.

Example 10.10.  Evicting Queries

import org.apache.openjpa.persistence.*;

...

OpenJPAEntityManagerFactory oemf = OpenJPAPersistence.cast(emf);
QueryResultCache qcache = oemf.getQueryResultCache();

// evict all queries that can be affected by changes to Magazines
qcache.evictAll(Magazine.class);

// evict an individual query with parameters
EntityManager em = emf.createEntityManager();
Query q = em.createQuery(...).
    setParameter(0, paramVal0).
    setParameter(1, paramVal1);
qcache.evict (q);

When using one of OpenJPA's distributed cache implementations, it is necessary to perform this in every JVM - the change notification is not propagated automatically. When using a third-party coherent caching solution, it is not necessary to do this in every JVM (although it won't hurt to do so), as the cache results are stored directly in the coherent cache.

Queries can also be pinned and unpinned through the QueryResultCache. The semantics of these operations are the same as pinning and unpinning data from the data cache.

public void pin(Query q);
public void unpin(Query q);

For JPA queries with parameters, set the desired parameter values into the Query instance before calling the above methods.

The following example shows these APIs in action.

Example 10.11.  Pinning, and Unpinning Query Results

import org.apache.openjpa.persistence.*;

...

OpenJPAEntityManagerFactory oemf = OpenJPAPersistence.cast(emf);
QueryResultCache qcache = oemf.getQueryResultCache();
EntityManager em = emf.createEntityManager();

Query pinQuery = em.createQuery(...).
    setParameter(0, paramVal0).
    setParameter(1, paramVal1);
qcache.pin(pinQuery);
Query unpinQuery = em.createQuery(...).
    setParameter(0, paramVal0).
    setParameter(1, paramVal1);
qcache.unpin(unpinQuery);

Pinning data into the cache instructs the cache to not expire the pinned results when cache flushing occurs. However, pinned results will be removed from the cache if an event occurs that invalidates the results.

You can disable caching on a per-EntityManager or per-Query basis:

Example 10.12.  Disabling and Enabling Query Caching

import org.apache.openjpa.persistence.*;

...

// temporarily disable query caching for all queries created from em
OpenJPAEntityManager oem = OpenJPAPersistence.cast(em);
oem.getFetchPlan ().setQueryResultCache(false);

// re-enable caching for a particular query
OpenJPAQuery oq = oem.createQuery(...);
oq.getFetchPlan().setQueryResultCache(true);

1.4.  Cache Extension

The provided data cache classes can be easily extended to add additional functionality. If you are adding new behavior, you should extend org.apache.openjpa.datacache.DataCacheImpl. To use your own storage mechanism, extend org.apache.openjpa.datacache.AbstractDataCache , or implement org.apache.openjpa.datacache.DataCache directly. If you want to implement a distributed cache that uses an unsupported method for communications, create an implementation of org.apache.openjpa.event.RemoteCommitProvider. This process is described in greater detail in Section 2.2, “ Customization ”.

The query cache is just as easy to extend. Add functionality by extending the default org.apache.openjpa.datacache.QueryCacheImpl. Implement your own storage mechanism for query results by extending org.apache.openjpa.datacache.AbstractQueryCache or implementing the org.apache.openjpa.datacache.QueryCache interface directly.

1.5.  Important Notes

  • The default cache implementations do not automatically refresh objects in other EntityManagers when the cache is updated or invalidated. This behavior would not be compliant with the JPA specification.

  • Invoking OpenJPAEntityManager.evictdoes not result in the corresponding data being dropped from the data cache, unless you have set the proper configuration options as explained above (see Example 10.6, “ Automatic Data Cache Eviction ”). Other methods related to the EntityManager cache also do not effect the data cache.

    The data cache assumes that it is up-to-date with respect to the datastore, so it is effectively an in-memory extension of the database. To manipulate the data cache, you should generally use the data cache facades presented in this chapter.

  • You must specify a org.apache.openjpa.event.RemoteCommitProvider (via the openjpa.RemoteCommitProvider property) in order to use the data cache, even when using the cache in a single-JVM mode. When using it in a single-JVM context, set this property to sjvm.

1.6.  Known Issues and Limitations

  • When using datastore (pessimistic) transactions in concert with the distributed caching implementations, it is possible to read stale data when reading data outside a transaction.

    For example, if you have two JVMs (JVM A and JVM B) both communicating with each other, and JVM A obtains a data store lock on a particular object's underlying data, it is possible for JVM B to load the data from the cache without going to the datastore, and therefore load data that should be locked. This will only happen if JVM B attempts to read data that is already in its cache during the period between when JVM A locked the data and JVM B received and processed the invalidation notification.

    This problem is impossible to solve without putting together a two-phase commit system for cache notifications, which would add significant overhead to the caching implementation. As a result, we recommend that people use optimistic locking when using data caching. If you do not, then understand that some of your non-transactional data may not be consistent with the datastore.

    Note that when loading objects in a transaction, the appropriate datastore transactions will be obtained. So, transactional code will maintain its integrity.

  • Extents are not cached. So, if you plan on iterating over a list of all the objects in an Extent on a regular basis, you will only benefit from caching if you do so with a Query instead:

    Example 10.13.  Query Replaces Extent

    import org.apache.openjpa.persistence.*;
    
    ...
    
    OpenJPAEntityManager oem = OpenJPAPersistence.cast(em);
    Extent extent = oem.getExtent(Magazine.class, false);
    
    // This iterator does not benefit from caching...
    Iterator uncachedIterator = extent.iterator();
    
    // ... but this one does.
    OpenJPAQuery extentQuery = oem.createQuery(...);
    extentQuery.setSubclasses(false);
    Iterator cachedIterator = extentQuery.getResultList().iterator();
    

2.  Query Compilation Cache

The query compilation cache is a Map used to cache parsed query strings. As a result, most queries are only parsed once in OpenJPA, and cached thereafter. You can control the compilation cache through the openjpa.QueryCompilationCache configuration property. This property accepts a plugin string (see Section 4, “ Plugin Configuration ”) describing the Map used to associate query strings and their parsed form. This property accepts the following aliases:

Table 10.2.  Pre-defined aliases

AliasValueNotes
true org.apache.openjpa.util.CacheMap The default option. Uses a CacheMap to store compilation data. CacheMap maintains a fixed number of cache entries, and an optional soft reference map for entries that are moved out of the LRU space. So, for applications that have a monotonically increasing number of distinct queries, this option can be used to ensure that a fixed amount of memory is used by the cache.
all org.apache.openjpa.lib.util.ConcurrentHashMap This is the fastest option, but compilation data is never dropped from the cache, so if you use a large number of dynamic queries, this option may result in ever-increasing memory usage. Note that if your queries only differ in the values of the parameters, this should not be an issue.
falsenone Disables the compilation cache.

Chapter 11.  Remote and Offline Operation

The standard JPA runtime environment is local and online. It is local in that components such as EntityManagers and queries connect directly to the datastore and execute their actions in the same JVM as the code using them. It is online in that all changes to managed objects must be made in the context of an active EntityManager. These two properties, combined with the fact that EntityManagers cannot be serialized for storage or network transfer, make the standard JPA runtime difficult to incorporate into some enterprise and client/server program designs.

OpenJPA extends the standard runtime to add remote and offline capabilities in the form of enhanced Detach and Attach APIs and Remote Commit Events. The following sections explain these capabilities in detail.

1.  Detach and Attach

The JPA Overview describes the specification's standard detach and attach APIs in Section 2, “ Entity Lifecycle Management ”. This section enumerates OpenJPA's enhancements to the standard behavior.

1.1.  Detach Behavior

In JPA, objects detach automatically when they are serialized or when a persistence context ends. The specification does not define any way to explicitly detach objects. The extended OpenJPAEntityManager, however, allows you to explicitly detach objects at any time.

public Object detach(Object pc):
public Object[] detachAll(Object... pcs):
public Collection detachAll(Collection pcs):

Each detach method returns detached copies of the given instances. The copy mechanism is similar to serialization, except that only certain fields are traversed. We will see how to control which fields are detached in a later section.

When detaching an instance that has been modified in the current transaction (and thus made dirty), the current transaction is flushed. This means that when subsequently re-attaching the detached instances, OpenJPA assumes that the transaction from which they were originally detached was committed; if it has been rolled back, then the re-attachment process will throw an optimistic concurrency exception.

You can stop OpenJPA from assuming the transaction will commit by invoking OpenJPAEntityManager.setRollbackOnly prior to detaching your objects. Setting the RollbackOnly flag prevents OpenJPA from flushing when detaching dirty objects; instead OpenJPA just runs its pre-flush actions (see the OpenJPAEntityManager.preFlush Javadoc for details).

This allows you to use the same instances in multiple attach/modify/detach/rollback cycles. Alternatively, you might also prevent a flush by making your modifications outside of a transaction (with NontransactionalWrite enabled) before detaching.

1.2.  Attach Behavior

When attaching, OpenJPA uses several strategies to determine the optimal way to merge changes made to the detached instance. As you will see, these strategies can even be used to attach changes made to a transient instance which was never detached in the first place.

  • If the instance was detached and detached state is enabled, OpenJPA will use the detached state to determine the object's version and primary key values. In addition, this state will tell OpenJPA which fields were loaded at the time of detach, and in turn where to expect changes. Loaded detached fields with null values will set the attached instance's corresponding fields to null.

  • If the instance has a Version field, OpenJPA will consider the object detached if the version field has a non-default value, and new otherwise.

    When attaching null fields in these cases, OpenJPA cannot distinguish between a field that was unloaded and one that was intentionally set to null. In this case, OpenJPA will use the current detach state setting to determine how to handle null fields: fields that would have been included in the detached state are treated as loaded, and will in turn set the corresponding attached field to null.

  • If neither of the above cases apply, OpenJPA will check to see if an instance with the same primary key values exists in the database. If so, the object is considered detached. Otherwise, it is considered new.

These strategies will be assigned on a per-instance basis, such that during the attachment of an object graph more than one of the above strategies may be used.

If you attempt to attach a versioned instance whose representation has changed in the datastore since detachment, OpenJPA will throw an optimistic concurrency exception upon commit or flush, just as if a normal optimistic conflict was detected. When attaching an instance whose database record has been deleted since detaching, or when attaching a detached instance into a manager that has a stale version of the object, OpenJPA will throw an optimistic concurrency exception from the attach method. In these cases, OpenJPA sets the RollbackOnly flag on the transaction.

1.3.  Defining the Detached Object Graph

When detached objects lose their association with the OpenJPA runtime, they also lose the ability to load additional state from the datastore. It is important, therefore, to populate objects with all the persistent state you will need before detaching them. While you are free to do this manually, OpenJPA includes facilities for automatically populating objects when they detach. The openjpa.DetachState configuration property determines which fields and relations are detached by default. All settings are recursive. They are:

  1. loaded: Detach all fields and relations that are already loaded, but don't include unloaded fields in the detached graph. This is the default.

  2. fgs: Detach all fields and relations in the current fetch configuration. For more information on custom fetch groups, see Section 6, “ Fetch Groups ”.

  3. all: Detach all fields and relations. Be very careful when using this mode; if you have a highly-connected domain model, you could end up bringing every object in the database into memory!

Any field that is not included in the set determined by the detach mode is set to its Java default value in the detached instance.

The openjpa.DetachState option is actually a plugin string (see Section 4, “ Plugin Configuration ”) that allows you to also configure the following options related to detached state:

  • DetachedStateField: As described in Section 1.2, “ Attach Behavior ” above, OpenJPA can take advantage of a detached state field to make the attach process more efficient. This field is added by the enhancer and is not visible to your application. Set this property to one of the following values:

    • transient: Use a transient detached state field. This gives the benefits of a detached state field to local objects that are never serialized, but retains serialization compatibility for client tiers without access to the enhanced versions of your classes. This is the default.

    • true: Use a non-transient detached state field so that objects crossing serialization barriers can still be attached efficiently. This requires, however, that your client tier have the enhanced versions of your classes and the OpenJPA libraries.

    • false: Do not use a detached state field.

    You can override the setting of this property or declare your own detached state field on individual classes using OpenJPA's metadata extensions. See Section 1.3.1, “ Detached State Field ” below.

  • DetachedStateManager: Whether to use a detached state manager. A detached state manager makes attachment much more efficient. Like a detached state field, however, it breaks serialization compatibility with the unenhanced class if it isn't transient.

    This setting piggybacks on the DetachedStateField setting above. If your detached state field is transient, the detached state manager will also be transient. If the detached state field is disabled, the detached state manager will also be disabled. This is typically what you'll want. By setting DetachedStateField to true (or transient) and setting this property to false, however, you can use a detached state field without using a detached state manager. This may be useful for debugging or for legacy OpenJPA users who find differences between OpenJPA's behavior with a detached state manager and OpenJPA's older behavior without one.

  • AccessUnloaded: Whether to allow access to unloaded fields of detached objects. Defaults to true. Set to false to throw an exception whenever an unloaded field is accessed. This option is only available when you use detached state managers, as determined by the settings above.

Example 11.1.  Configuring Detached State

<property name="openjpa.DetachState" value="fgs(DetachedStateField=true)"/>

You can also alter the set of fields that will be included in the detached graph at runtime. OpenJPAEntityManagers expose the following APIs for controlling detached state:

public static final int DETACH_LOADED;
public static final int DETACH_FGS;
public static final int DETACH_ALL;
public int getDetachState();
public void setDetachState(int mode);

1.3.1.  Detached State Field

When the detached state field is enabled, the OpenJPA enhancer adds an additional field to the enhanced version of your class. This field of type Object. OpenJPA uses this field for bookkeeping information, such as the versioning data needed to detect optimistic concurrency violations when the object is re-attached.

It is possible to define this detached state field yourself. Declaring this field in your class metadata prevents the enhancer from adding any extra fields to the class, and keeps the enhanced class serialization-compatible with the unenhanced version. The detached state field must not be persistent. See Section 3.1.3, “ Detached State ” for details on how to declare a detached state field.

import org.apache.openjpa.persistence.*;

@Entity
public class Magazine
    implements Serializable {

    private String name;
    @DetachedState private Object state;
    ...
}

2.  Remote Event Notification Framework

The remote event notification framework allows a subset of the information available through OpenJPA's transaction events (see Section 7, “ Transaction Events ”) to be broadcast to remote listeners. OpenJPA's data cache, for example, uses remote events to remain synchronized when deployed in multiple JVMs.

To enable remote events, you must configure the EntityManagerFactory to use a RemoteCommitProvider (see below).

When a RemoteCommitProvider is properly configured, you can register RemoteCommitListeners that will be alerted with a list of modified object ids whenever a transaction on a remote machine successfully commits.

2.1.  Remote Commit Provider Configuration

OpenJPA includes built in remote commit providers for JMS and TCP communication.

2.1.1.  JMS

The JMS remote commit provider can be configured by setting the openjpa.RemoteCommitProvider property to contain the appropriate configuration properties. The JMS provider understands the following properties:

  • Topic: The topic that the remote commit provider should publish notifications to and subscribe to for notifications sent from other JVMs. Defaults to topic/OpenJPACommitProviderTopic

  • TopicConnectionFactory: The JNDI name of a javax.jms.TopicConnectionFactory factory to use for finding topics. Defaults to java:/ConnectionFactory. This setting may vary depending on the application server in use; consult the application server's documentation for details of the default JNDI name for the javax.jms.TopicConnectionFactory instance. For example, under Weblogic, the JNDI name for the TopicConnectionFactory is javax.jms.TopicConnectionFactory.

  • ExceptionReconnectAttempts: The number of times to attempt to reconnect if the JMS system notifies OpenJPA of a serious connection error. Defaults to 0, meaning OpenJPA will log the error but otherwise ignore it, hoping the connection is still valid.

  • *: All other configuration properties will be interpreted as settings to pass to the JNDI InitialContext on construction. For example, you might set the java.naming.provider.url property to the URL of the context provider.

To configure a factory to use the JMS provider, your properties might look like the following:

Note

Because of the nature of JMS, it is important that you invoke EntityManagerFactory.close when finished with a factory. If you do not do so, a daemon thread will stay up in the JVM, preventing the JVM from exiting.

2.1.2.  TCP

The TCP remote commit provider has several options that are defined as host specifications containing a host name or IP address and an optional port separated by a colon. For example, the host specification saturn.bea.com:1234 represents an InetAddress retrieved by invoking InetAddress.getByName("saturn.bea.com") and a port of 1234.

The TCP provider can be configured by setting the openjpa.RemoteCommitProvider plugin property to contain the appropriate configuration settings. The TCP provider understands the following properties:

  • Port: The TCP port that the provider should listen on for commit notifications. Defaults to 5636.

  • Addresses: A semicolon-separated list of IP addresses to which notifications should be sent. No default value.

  • NumBroadcastThreads: The number of threads to create for the purpose of transmitting events to peers. You sould increase this value as the number of concurrent transactions increases. The maximum number of concurrent transactions is a function of the size of the connection pool. See the the MaxActive property of openjpa.ConnectionFactoryProperties in Section 1, “ Using the OpenJPA DataSource ”. Setting a value of 0 will result in behavior where the thread invoking commit will perform the broadcast directly. Defaults to 2.

  • RecoveryTimeMillis: Amount of time to wait in milliseconds before attempting to reconnect to a peer of the cluster when connectivity to the peer is lost. Defaults to 15000.

  • MaxIdle: The number of TCP sockets (channels) to keep open to each peer in the cluster for the transmission of events. Defaults to 2.

  • MaxActive: The maximum allowed number of TCP sockets (channels) to open simultaneously between each peer in the cluster. Defaults to 2.

To configure a factory to use the TCP provider, your properties might look like the following:

Example 11.2.  TCP Remote Commit Provider Configuration

<property name="openjpa.RemoteCommitProvider"
    value="tcp(Addresses=10.0.1.10;10.0.1.11;10.0.1.12;10.0.1.13)"/>

2.1.3.  Common Properties

In addition to the provider-specific configuration options above, all providers accept the following plugin properties:

  • TransmitPersistedObjectIds: Whether remote commit events will include the object ids of instances persisted in the transaction. By default only the class names of types persisted in the transaction are sent. This results in smaller events and more efficient network utilization. If you have registered your own remote commit listeners, however, you may require the persisted object ids as well.

To transmit persisted object ids in our remote commit events using the JMS provider, we modify the previous example as follows:

2.2.  Customization

You can develop additional mechanisms for remote event notification be by creating an implementation of the RemoteCommitProvider interface, possibly by extending the AbstractRemoteCommitProvider abstract class..

Chapter 12.  Third Party Integration

OpenJPA provides a number of mechanisms for integrating with third-party tools. The following chapter will illustrate these integration features.

1.  Apache Ant

Ant is a very popular tool for building Java projects. It is similar to the make command, but is Java-centric and has more modern features. Ant is open source, and can be downloaded from Apache's Ant web page at http://jakarta.apache.org/ant/ . Ant has become the de-facto standard build tool for Java, and many commercial integrated development environments provide some support for using ant build files. The remainder of this section assumes familiarity with writing Ant build.xml files.

OpenJPA provides pre-built Ant task definitions for all bundled tools:

The source code for all the ant tasks is provided with the distribution under the src directory. This allows you to customize various aspects of the ant tasks in order to better integrate into your development environment.

1.1.  Common Ant Configuration Options

All OpenJPA tasks accept a nested config element, which defines the configuration environment in which the specified task will run. The attributes for the config tag are defined by the JDBCConfiguration bean methods. Note that excluding the config element will cause the Ant task to use the default system configuration mechanism, such as the configuration defined in the org.apache.openjpa.xml file.

Following is an example of how to use the nested config tag in a build.xml file:

Example 12.1.  Using the <config> Ant Tag

<mappingtool>
  <fileset dir="${basedir}">
    <include name="**/model/*.java" />
  </fileset>
  <config connectionUserName="scott" connectionPassword="tiger"
    connectionURL="jdbc:oracle:thin:@saturn:1521:solarsid"
    connectionDriverName="oracle.jdbc.driver.OracleDriver" />
</mappingtool>

It is also possible to specify a properties or propertiesFile attribute on the config tag, which will be used to locate a properties resource or file. The resource will be loaded relative to the current CLASSPATH.

Example 12.2.  Using the Properties Attribute of the <config> Tag

<mappingtool>
  <fileset dir="${basedir}">
    <include name="**/model/*.java"/>
  </fileset>
  <config properties="openjpa-dev.xml"/>
</mappingtool>

Example 12.3.  Using the PropertiesFile Attribute of the <config> Tag

<mappingtool>
  <fileset dir="${basedir}">
    <include name="**/model/*.java"/>
  </fileset>
  <config propertiesFile="../conf/openjpa-dev.xml"/>
</mappingtool>

Tasks also accept a nested classpath element, which you can use in place of the default classpath. The classpath argument behaves the same as it does for Ant's standard javac element. It is sometimes the case that projects are compiled to a separate directory than the source tree. If the target path for compiled classes is not included in the project's classpath, then a classpath element that includes the target class directory needs to be included so the enhancer and mapping tool can locate the relevant classes.

Following is an example of using a classpath tag:

Example 12.4.  Using the <classpath> Ant Tag

<openjpac>
  <fileset dir="${basedir}/source">
    <include name="**/model/*.java" />
  </fileset>
  <classpath>
    <pathelement location="${basedir}/classes"/>
    <pathelement location="${basedir}/source"/>
    <pathelement path="${java.class.path}"/>
  </classpath>
</openjpac>

Finally, tasks that invoke code-generation tools like the application identity tool and reverse mapping tool accept a nested codeformat element. See the code formatting documentation in Section 3.1, “ Code Formatting ” for a list of code formatting attributes.

Example 12.5.  Using the <codeformat> Ant Tag

<reversemappingtool package="com.xyz.jdo" directory="${basedir}/src">
  <codeformat tabSpaces="4" spaceBeforeParen="true" braceOnSameLine="false"/>
</reversemappingtool>

1.2.  Enhancer Ant Task

The enhancer task allows you to invoke the OpenJPA enhancer directly from within the Ant build process. The task's parameters correspond exactly to the long versions of the command-line arguments to the org.apache.openjpa.enhance.PCEnhancer.

The enhancer task accepts a nested fileset tag to specify the files that should be processed. You can specify .java or .class files. If you do not specify any files, the task will run on the classes listed in your persistence.xml or openjpa.MetaDataFactory property.

Following is an example of using the enhancer task in a build.xml file:

Example 12.6.  Invoking the Enhancer from Ant

<target name="enhance">
  <!-- define the openjpac task; this can be done at the top of the    -->
  <!-- build.xml file, so it will be available for all targets      -->
  <taskdef name="openjpac" classname="org.apache.openjpa.ant.PCEnhancerTask"/>

  <!-- invoke enhancer on all .java files below the model directory -->
  <openjpac>
    <fileset dir=".">
      <include name="**/model/*.java" />
    </fileset>
  </openjpac>
</target>

1.3.  Application Identity Tool Ant Task

The application identity tool task allows you to invoke the application identity tool directly from within the Ant build process. The task's parameters correspond exactly to the long versions of the command-line arguments to the org.apache.openjpa.enhance.ApplicationIdTool.

The application identity tool task accepts a nested fileset tag to specify the files that should be processed. You can specify .java or .class files. If you do not specify any files, the task will run on the classes listed in your persistence.xml file or openjpa.MetaDataFactory property.

Following is an example of using the application identity tool task in a build.xml file:

Example 12.7.  Invoking the Application Identity Tool from Ant

<target name="appids">
  <!-- define the appidtool task; this can be done at the top of     -->
  <!-- the build.xml file, so it will be available for all targets   -->
  <taskdef name="appidtool" classname="org.apache.openjpa.ant.ApplicationIdToolTask"/>

  <!-- invoke tool on all .jdo files below the current directory     -->
  <appidtool>
    <fileset dir=".">
      <include name="**/model/*.java" />
    </fileset>
    <codeformat spaceBeforeParen="true" braceOnSameLine="false"/>
  </appidtool>
</target>

1.4.  Mapping Tool Ant Task

The mapping tool task allows you to directly invoke the mapping tool from within the Ant build process. It is useful for making sure that the database schema and object-relational mapping data is always synchronized with your persistent class definitions, without needing to remember to invoke the mapping tool manually. The task's parameters correspond exactly to the long versions of the command-line arguments to the org.apache.openjpa.jdbc.meta.MappingTool.

The mapping tool task accepts a nested fileset tag to specify the files that should be processed. You can specify .java or .class files. If you do not specify any files, the task will run on the classes listed in your persistence.xml file or openjpa.MetaDataFactory property.

Following is an example of a build.xml target that invokes the mapping tool:

Example 12.8.  Invoking the Mapping Tool from Ant

<target name="refresh">
  <!-- define the mappingtool task; this can be done at the top of -->
  <!-- the build.xml file, so it will be available for all targets -->
  <taskdef name="mappingtool" classname="org.apache.openjpa.jdbc.ant.MappingToolTask"/>

  <!-- add the schema components for all .jdo files below the      -->
  <!-- current directory                                           -->
  <mappingtool action="buildSchema">
    <fileset dir=".">
      <include name="**/*.jdo" />
    </fileset>
  </mappingtool>
</target>

1.5.  Reverse Mapping Tool Ant Task

The reverse mapping tool task allows you to directly invoke the reverse mapping tool from within Ant. While many users will only run the reverse mapping process once, others will make it part of their build process. The task's parameters correspond exactly to the long versions of the command-line arguments to the org.apache.openjpa.jdbc.meta.ReverseMappingTool.

Following is an example of a build.xml target that invokes the reverse mapping tool:

Example 12.9.  Invoking the Reverse Mapping Tool from Ant

<target name="reversemap">
  <!-- define the reversemappingtool task; this can be done at the top of -->
  <!-- the build.xml file, so it will be available for all targets        -->
  <taskdef name="reversemappingtool" 
    classname="org.apache.openjpa.jdbc.ant.ReverseMappingToolTask"/>

  <!-- reverse map the entire database -->
  <reversemappingtool package="com.xyz.model" directory="${basedir}/src"
    customizerProperties="${basedir}/conf/reverse.properties">
    <codeformat tabSpaces="4" spaceBeforeParen="true" braceOnSameLine="false"/>
  </reversemappingtool>
</target>

1.6.  Schema Tool Ant Task

The schema tool task allows you to directly invoke the schema tool from within the Ant build process. The task's parameters correspond exactly to the long versions of the command-line arguments to the org.apache.openjpa.jdbc.schema.SchemaTool.

Following is an example of a build.xml target that invokes the schema tool:

Example 12.10.  Invoking the Schema Tool from Ant

<target name="schema">
  <!-- define the schematool task; this can be done at the top of  -->
  <!-- the build.xml file, so it will be available for all targets -->
  <taskdef name="schematool" classname="org.apache.openjpa.jdbc.ant.SchemaToolTask"/>

  <!-- add the schema components for all .schema files below the   -->
  <!-- current directory                                           -->
  <schematool action="add">
    <fileset dir=".">
      <include name="**/*.schema" />
    </fileset>
  </schematool>
</target>

Chapter 13.  Optimization Guidelines

There are numerous techniques you can use in order to ensure that OpenJPA operates in the fastest and most efficient manner. Following are some guidelines. Each describes what impact it will have on performance and scalability. Note that general guidelines regarding performance or scalability issues are just that - guidelines. Depending on the particular characteristics of your application, the optimal settings may be considerably different than what is outlined below.

In the following table, each row is labeled with a list of italicized keywords. These keywords identify what characteristics the row in question may improve upon. Many of the rows are marked with one or both of the performance and scalability labels. It is important to bear in mind the differences between performance and scalability (for the most part, we are referring to system-wide scalability, and not necessarily only scalability within a single JVM). The performance-related hints will probably improve the performance of your application for a given user load, whereas the scalability-related hints will probably increase the total number of users that your application can service. Sometimes, increasing performance will decrease scalability, and vice versa. Typically, options that reduce the amount of work done on the database server will improve scalability, whereas those that push more work onto the server will have a negative impact on scalability.

Table 13.1.  Optimization Guidelines

Plugin in a Connection Pool

performance, scalability

OpenJPA's built-in datasource does not perform connection pooling or prepared statement caching. Plugging in a third-party pooling datasource may drastically improve performance.
Optimize database indexes

performance, scalability

The default set of indexes created by OpenJPA's mapping tool may not always be the most appropriate for your application. Manually setting indexes in your mapping metadata or manually manipulating database indexes to include frequently-queried fields (as well as dropping indexes on rarely-queried fields) can yield significant performance benefits.

A database must do extra work on insert, update, and delete to maintain an index. This extra work will benefit selects with WHERE clauses, which will execute much faster when the terms in the WHERE clause are appropriately indexed. So, for a read-mostly application, appropriate indexing will slow down updates (which are rare) but greatly accelerate reads. This means that the system as a whole will be faster, and also that the database will experience less load, meaning that the system will be more scalable.

Bear in mind that over-indexing is a bad thing, both for scalability and performance, especially for applications that perform lots of inserts, updates, or deletes.

JVM optimizations

performance, reliability

Manipulating various parameters of the Java Virtual Machine (such as hotspot compilation modes and the maximum memory) can result in performance improvements. For more details about optimizing the JVM execution environment, please see http://java.sun.com/docs/hotspot/PerformanceFAQ.html.
Use the data cache

performance, scalability

Using OpenJPA's data and query caching features can often result in a dramatic improvement in performance. Additionally, these caches can significantly reduce the amount of load on the database, increasing the scalability characteristics of your application.
Set LargeTransaction to true, or set PopulateDataCache to false

performance vs. scalability

When using OpenJPA's data caching features in a transaction that will delete, modify, or create a very large number of objects you can set LargeTransaction to true and perform periodic flushes during your transaction to reduce its memory requirements. See the Javadoc: OpenJPAEntityManager.setLargeTransaction. Note that transactions in large mode have to more aggressively flush items from the data cache.

If your transaction will visit objects that you know are very unlikely to be accessed by other transactions, for example an exhaustive report run only once a month, you can turn off population of the data cache so that the transaction doesn't fill the entire data cache with objects that won't be accessed again. Again, see the Javadoc: OpenJPAEntityManager.setPopulateDataCache

Disable logging, performance tracking

performance

Developer options such as verbose logging and the JDBC performance tracker can result in serious performance hits for your application. Before evaluating OpenJPA's performance, these options should all be disabled.
Set IgnoreChanges to true, or set FlushBeforeQueries to true

performance vs. scalability

When both the openjpa.IgnoreChanges and openjpa.FlushBeforeQueries properties are set to false, OpenJPA needs to consider in-memory dirty instances during queries. This can sometimes result in OpenJPA needing to evaluate the entire extent objects in order to return the correct query results, which can have drastic performance consequences. If it is appropriate for your application, configuring FlushBeforeQueries to automatically flush before queries involving dirty objects will ensure that this never happens. Setting IgnoreChanges to false will result in a small performance hit even if FlushBeforeQueries is true, as incremental flushing is not as efficient overall as delaying all flushing to a single operation during commit.

Setting IgnoreChanges to true will help performance, since dirty objects can be ignored for queries, meaning that incremental flushing or client-side processing is not necessary. It will also improve scalability, since overall database server usage is diminished. On the other hand, setting IgnoreChanges to false will have a negative impact on scalability, even when using automatic flushing before queries, since more operations will be performed on the database server.

Configure openjpa.ConnectionRetainMode appropriately

performance vs. scalability

The ConnectionRetainMode configuration option controls when OpenJPA will obtain a connection, and how long it will hold that connection. The optimal settings for this option will vary considerably depending on the particular behavior of your application. You may even benefit from using different retain modes for different parts of your application.

The default setting of on-demand minimizes the amount of time that OpenJPA holds onto a datastore connection. This is generally the best option from a scalability standpoind, as database resources are held for a minimal amount of time. However, if you are not using connection pooling, or if your DataSource is not efficient at managing its pool, then this default value could cause undesirable pool contention.

Use flat inheritance

performance, scalability vs. disk space

Mapping inheritance hierarchies to a single database table is faster for most operations than other strategies employing multiple tables. If it is appropriate for your application, you should use this strategy whenever possible.

However, this strategy will require more disk space on the database side. Disk space is relatively inexpensive, but if your object model is particularly large, it can become a factor.

High sequence increment

performance, scalability

For applications that perform large bulk inserts, the retrieval of sequence numbers can be a bottleneck. Increasing sequence increments and using table-based rather than native database sequences can reduce or eliminate this bottleneck. In some cases, implementing your own sequence factory can further optimize sequence number retrieval.
Use optimistic transactions

performance, scalability

Using datastore transactions translates into pessimistic database row locking, which can be a performance hit (depending on the database). If appropriate for your application, optimistic transactions are typically faster than datastore transactions.

Optimistic transactions provide the same transactional guarantees as datastore transactions, except that you must handle a potential optimistic verification exception at the end of a transaction instead of assuming that a transaction will successfully complete. In many applications, it is unlikely that different concurrent transactions will operate on the same set of data at the same time, so optimistic verification increases the concurrency, and therefore both the performance and scalability characteristics, of the application. A common approach to handling optimistic verification exceptions is to simply present the end user with the fact that concurrent modifications happened, and require that the user redo any work.

Use query aggregates and projections

performance, scalability

Using aggregates to compute reporting data on the database server can drastically speed up queries. Similarly, using projections when you are interested in specific object fields or relations rather than the entire object state can reduce the amount of data OpenJPA must transfer from the database to your application.
Always close resources

scalability

Under certain settings, EntityManager s, OpenJPA Extent iterators, and Query results may be backed by resources in the database.

For example, if you have configured OpenJPA to use scrollable cursors and lazy object instantiation by default, each query result will hold open a ResultSet object, which, in turn, will hold open a Statement object (preventing it from being re-used). Garbage collection will clean up these resources, so it is never necessary to explicitly close them, but it is always faster if it is done at the application level.

Use detached state managers

performance

Attaching and even persisting instances can be more efficient when your detached objects use detached state managers. By default, OpenJPA does not use detached state managers when serializing an instance across tiers. See Section 1.3, “ Defining the Detached Object Graph ” for how to force OpenJPA to use detached state managers across tiers, and for other options for more efficient attachment.

The downside of using a detached state manager across tiers is that your enhanced persistent classes and the OpenJPA libraries must be available on the client tier.

Utilize the EntityManager cache

performance, scalability

When possible and appropriate, re-using EntityManagers and setting the RetainState configuration option to true may result in significant performance gains, since the EntityManager's built-in object cache will be used.
Enable multithreaded operation only when necessary

performance

OpenJPA respects the openjpa.Multithreaded option in that it does not impose as much synchronization overhead for applications that do not set this value to true. If your application is guaranteed to only use single-threaded access to OpenJPA resources and persistent objects, leaving this option as false will reduce synchronization overhead, and may result in a modest performance increase.
Enable large data set handling

performance, scalability

If you execute queries that return large numbers of objects or have relations (collections or maps) that are large, and if you often only access parts of these data sets, enabling large result set handling where appropriate can dramatically speed up your application, since OpenJPA will bring the data sets into memory from the database only as necessary.
Disable large data set handling

performance, scalability

If you have enabled scrollable result sets and on-demand loading but do you not require it, consider disabling it again. Some JDBC drivers and databases (SQLServer for example) are much slower when used with scrolling result sets.
Use the DynamicSchemaFactory

performance, validation

If you are using a openjpa.jdbc.SchemaFactory setting of something other than the default of dynamic, consider switching back. While other factories can ensure that object-relational mapping information is valid when a persistent class is first used, this can be a slow process. Though the validation is only performed once for each class, switching back to the DynamicSchemaFactory can reduce the warm-up time for your application.
Do not use XA transactions

performance, scalability

XA transactions can be orders of magnitude slower than standard transactions. Unless distributed transaction functionality is required by your application, use standard transactions.

Recall that XA transactions are distinct from managed transactions - managed transaction services such as that provided by EJB declarative transactions can be used both with XA and non-XA transactions. XA transactions should only be used when a given business transaction involves multiple different transactional resources (an Oracle database and an IBM transactional message queue, for example).

Use Sets instead of List/Collections

performance, scalability

There is a small amount of extra overhead for OpenJPA to maintain collections where each element is not guaranteed to be unique. If your application does not require duplicates for a collection, you should always declare your fields to be of type Set, SortedSet, HashSet, or TreeSet.
Use query parameters instead of encoding search data in filter strings

performance

If your queries depend on parameter data only known at runtime, you should use query parameters rather than dynamically building different query strings. OpenJPA performs aggressive caching of query compilation data, and the effectiveness of this cache is diminished if multiple query filters are used where a single one could have sufficed.
Tune your fetch groups appropriately

performance, scalability

The fetch groups used when loading an object control how much data is eagerly loaded, and by extension, which fields must be lazily loaded at a future time. The ideal fetch group configuration loads all the data that is needed in one fetch, and no extra fields - this minimizes both the amount of data transferred from the database, and the number of trips to the database.

If extra fields are specified in the fetch groups (in particular, large fields such as binary data, or relations to other persistence-capable objects), then network overhead (for the extra data) and database processing (for any necessary additional joins) will hurt your application's performance. If too few fields are specified in the fetch groups, then OpenJPA will have to make additional trips to the database to load additional fields as necessary.

Use eager fetching

performance, scalability

Using eager fetching when loading subclass data or traversing relations for each instance in a large collection of results can speed up data loading by orders of magnitude.
Disable BrokerImpl finalization

performance, scalability

Outside of a Java EE 5 application server or other JPA persistence container, OpenJPA's EntityManagers use finalizers to ensure that resources get cleaned up. If you are properly managing your resources, this finalization is not necessary, and will introduce unneeded synchronization, leading to scalability problems. You can disable this protective behavior by setting the openjpa.BrokerImpl property to non-finalizing. See ??? for details.

Appendix A.  JPA Resources

Appendix B.  Supported Databases

Following is a table of the database and JDBC driver versions that are supported by OpenJPA.

Table B.1.  Supported Databases and JDBC Drivers

Database Name Database Version JDBC Driver Name JDBC Driver Version
Apache Derby 10.1.2.1 Apache Derby Embedded JDBC Driver 10.1.2.1
Borland Interbase 7.1.0.202 Interclient 4.5.1
Borland JDataStore 6.0 Borland JDataStore 6.0
DB2 8.1 IBM DB2 JDBC Universal Driver 1.0.581
Empress 8.62 Empress Category 2 JDBC Driver 8.62
Firebird 1.5 JayBird JCA/JDBC driver 1.0.1
H2 Database Engine 1.0 H2 1.0
Hypersonic Database Engine 1.8.0 Hypersonic 1.8.0
Informix Dynamic Server 9.30.UC10 Informix JDBC driver 2.21.JC2
InterSystems Cache 5.0 Cache JDBC Driver 5.0
Microsoft Access 9.0 (a.k.a. "2000") DataDirect SequeLink 5.4.0038
Microsoft SQL Server 9.00.1399 (SQL Server 2005) SQLServer 1.0.809.102
Microsoft Visual FoxPro 7.0 DataDirect SequeLink 5.4.0038
MySQL 3.23.43-log MySQL Driver 3.0.14
MySQL 5.0.26 MySQL Driver 3.0.14
Oracle 8.1,9.2,10.1 Oracle JDBC driver 10.2.0.1.0
Pointbase 4.4 Pointbase JDBC driver 4.4 (4.4)
PostgreSQL 7.2.1 PostgreSQL Native Driver 8.1
PostgreSQL 8.1.5 PostgreSQL Native Driver 8.1
Sybase Adaptive Server Enterprise 12.5 jConnect 5.5 (5.5)

1.  Apache Derby

Example B.1.  Example properties for Derby

openjpa.ConnectionDriverName: org.apache.derby.jdbc.EmbeddedDriver
openjpa.ConnectionURL: jdbc:derby:DB_NAME;create=true

2.  Borland Interbase

Example B.2.  Example properties for Interbase

openjpa.ConnectionDriverName: interbase.interclient.Driver
openjpa.ConnectionURL: jdbc:interbase://SERVER_NAME:SERVER_PORT/DB_PATH

2.1.  Known issues with Interbase

  • Interbase does not support record locking, so datastore transactions cannot use the pessimistic lock manager.

  • Interbase does not support the LOWER, SUBSTRING , or INSTR SQL functions>

3.  JDataStore

Example B.3.  Example properties for JDataStore

openjpa.ConnectionDriverName: com.borland.datastore.jdbc.DataStoreDriver
openjpa.ConnectionURL: jdbc:borland:dslocal:db-jdatastore.jds;create=true

4.  IBM DB2

Example B.4.  Example properties for IBM DB2

openjpa.ConnectionDriverName: com.ibm.db2.jcc.DB2Driver
openjpa.ConnectionURL: jdbc:db2://SERVER_NAME:SERVER_PORT/DB_NAME

4.1.  Known issues with DB2

  • Floats and doubles may lose precision when stored.

  • Empty char values are stored as NULL.

  • Fields of type BLOB and CLOB are limited to 1M. This number can be increased by extending DB2Dictionary.

5.  Empress

Example B.5.  Example properties for Empress

openjpa.ConnectionDriverName: empress.jdbc.empressDriver
openjpa.ConnectionURL: jdbc:empress://SERVER=yourserver;PORT=6322;DATABASE=yourdb

5.1.  Known issues with Empress

  • Empress enforces pessimistic semantics (lock on read) when not using AllowConcurrentRead property (which bypasses row locking) for EmpressDictionary.

  • Only the category 2 non-local driver is supported.

6.  H2 Database Engine

Example B.6.  Example properties for H2 Database Engine

openjpa.ConnectionDriverName: org.h2.Driver
openjpa.ConnectionURL: jdbc:h2:DB_NAME

6.1.  Known issues with H2 Database Engine

  • H2 does not support cross joins

7.  Hypersonic

Example B.7.  Example properties for Hypersonic

openjpa.ConnectionDriverName: org.hsqldb.jdbcDriver
openjpa.ConnectionURL: jdbc:hsqldb:DB_NAME

7.1.  Known issues with Hypersonic

  • Hypersonic does not support pessimistic locking, so non-optimistic transactions will fail unless the SimulateLocking property is set for the openjpa.jdbc.DBDictionary

8.  Firebird

Example B.8.  Example properties for Firebird

openjpa.ConnectionDriverName: org.firebirdsql.jdbc.FBDriver
openjpa.ConnectionURL: jdbc:firebirdsql://SERVER_NAME:SERVER_PORT/DB_PATH

8.1.  Known issues with Firebird

  • Firebird does not support auto-increment columns.

  • Firebird does not support the LOWER, SUBSTRING , or INSTR SQL functions.

9.  Informix

Example B.9.  Example properties for Informix Dynamic Server

openjpa.ConnectionDriverName: com.informix.jdbc.IfxDriver
openjpa.ConnectionURL: \
  jdbc:informix-sqli://SERVER_NAME:SERVER_PORT/DB_NAME:INFORMIXSERVER=SERVER_ID

9.1.  Known issues with Informix

  • None

10.  InterSystems Cache

Example B.10.  Example properties for InterSystems Cache

openjpa.ConnectionDriverName: com.intersys.jdbc.CacheDriver
openjpa.ConnectionURL: jdbc:Cache://SERVER_NAME:SERVER_PORT/DB_NAME

10.1.  Known issues with InterSystems Cache

  • Support for Cache is done via SQL access over JDBC, not through their object database APIs.

11.  Microsoft Access

Example B.11.  Example properties for Microsoft Access

openjpa.ConnectionDriverName: com.ddtek.jdbc.sequelink.SequeLinkDriver
openjpa.ConnectionURL: jdbc:sequelink://SERVER_NAME:SERVER_PORT

11.1.  Known issues with Microsoft Access

  • Using the Sun JDBC-ODBC bridge to connect is not supported.

12.  Microsoft SQL Server

Example B.12.  Example properties for Microsoft SQLServer

openjpa.ConnectionDriverName: com.microsoft.sqlserver.jdbc.SQLServerDriver
openjpa.ConnectionURL: \
    jdbc:sqlserver://SERVER_NAME:1433;DatabaseName=DB_NAME;selectMethod=cursor;sendStringParametersAsUnicode=false

12.1.  Known issues with SQL Server

  • SQL Server date fields are accurate only to the nearest 3 milliseconds, possibly resulting in precision loss in stored dates.

  • The ConnectionURL must always contain the " selectMethod=cursor " string.

  • Adding sendStringParametersAsUnicode=false to the ConnectionURL may significantly increase performance.

  • The Microsoft SQL Server driver only emulates batch updates. The DataDirect JDBC driver has true support for batch updates, and may result in a significant performance gain.

  • Floats and doubles may lose precision when stored.

  • TEXT columns cannot be used in queries.

  • When using a SQL Server instance that has been configured to be case-sensitive in schema names, you need to set the "schemaCase=preserve" parameter in the openjpa.jdbc.DBDictionary property.

13.  Microsoft FoxPro

Example B.13.  Example properties for Microsoft FoxPro

openjpa.ConnectionDriverName: com.ddtek.jdbc.sequelink.SequeLinkDriver
openjpa.ConnectionURL: jdbc:sequelink://SERVER_NAME:SERVER_PORT

13.1.  Known issues with Microsoft FoxPro

  • Using the Sun JDBC-ODBC bridge to connect is not supported.

14.  MySQL

Example B.14.  Example properties for MySQL

openjpa.ConnectionDriverName: com.mysql.jdbc.Driver
openjpa.ConnectionURL: jdbc:mysql://SERVER_NAME/DB_NAME

14.1.  Known issues with MySQL

  • The default table types that MySQL uses do not support transactions, which will prevent OpenJPA from being able to roll back transactions. Use the InnoDB table type for any tables that OpenJPA will access.

  • MySQL does not support sub-selects in versions prior to 4.1, and are disabled by default. Some operations (such as the isEmpty() method in a query) will fail due to this. If you are using MySQL 4.1 or later, you can lift this restriction by setting the SupportsSubselect=true parameter of the openjpa.jdbc.DBDictionary property.

  • Rollback due to database error or optimistic lock violation is not supported unless the table type is one of the MySQL transactional types. Explicit calls to rollback() before a transaction has been committed, however, are always supported.

  • Floats and doubles may lose precision when stored in some datastores.

  • When storing a field of type java.math.BigDecimal, some datastores will add extraneous trailing 0 characters, causing an equality mismatch between the field that is stored and the field that is retrieved.

  • Some version of the MySQL JDBC driver have a bug that prevents OpenJPA from being able to interrogate the database for foreign keys. Version 3.0.14 (or higher) of the MySQL driver is required in order to get around this bug.

15.  Oracle

Example B.15.  Example properties for Oracle

openjpa.ConnectionDriverName: oracle.jdbc.driver.OracleDriver
openjpa.ConnectionURL: jdbc:oracle:thin:@SERVER_NAME:1521:DB_NAME

15.1.  Using Query Hints with Oracle

Oracle has support for "query hints", which are formatted comments embedded in SQL that provide some hint for how the query should be executed. These hints are usually designed to provide suggestions to the Oracle query optimizer for how to efficiently perform a certainly query, and aren't typically needed for any but the most intensive queries.

Example B.16.  Using Oracle Hints

Query query = em.createQuery(...);
query.setHint("openjpa.hint.OracleSelectHint", "/*+ first_rows(100) */");
List results = query.getResultList();

15.2.  Known issues with Oracle

  • The Oracle JDBC driver has significant differences between different versions. It is important to use the officially supported version of the driver (10.2.0.1.0), which is backward compatible with previous versions of the Oracle server. It can be downloaded from http://www.oracle.com/technology/software/tech/java/sqlj_jdbc/htdocs/jdbc101040.html.

  • For VARCHAR fields, null and a blank string are equivalent. This means that an object that stores a null string field will have it get read back as a blank string.

  • Oracle corp's JDBC driver for Oracle has only limited support for batch updates. The result for OpenJPA is that in some cases, the exact object that failed an optimistic lock check cannot be determined, and OpenJPA will throw an OptimisticVerificationException with more failed objects than actually failed.

  • Oracle cannot store numbers with more than 38 digits in numeric columns.

  • Floats and doubles may lose precision when stored.

  • CLOB columns cannot be used in queries.

16.  Pointbase

Example B.17.  Example properties for Pointbase

openjpa.ConnectionDriverName: com.pointbase.jdbc.jdbcUniversalDriver
openjpa.ConnectionURL: \
    jdbc:pointbase:DB_NAME,database.home=pointbasedb,create=true,cache.size=10000,database.pagesize=30720

16.1.  Known issues with Pointbase

  • Fields of type BLOB and CLOB are limited to 1M. Set the BlobTypeName and/or ClobTypeName properties of the openjpa.jdbc.DBDictionary setting to override.

17.  PostgreSQL

Example B.18.  Example properties for PostgreSQL

openjpa.ConnectionDriverName: org.postgresql.Driver
openjpa.ConnectionURL: jdbc:postgresql://SERVER_NAME:5432/DB_NAME

17.1.  Known issues with PostgreSQL

  • Floats and doubles may lose precision when stored.

  • PostgreSQL cannot store very low and very high dates.

  • Empty string/char values are stored as NULL.

18.  Sybase Adaptive Server

Example B.19.  Example properties for Sybase

openjpa.ConnectionDriverName: com.sybase.jdbc2.jdbc.SybDriver
openjpa.ConnectionURL: \
    jdbc:sybase:Tds:SERVER_NAME:4100/DB_NAME?ServiceName=DB_NAME&BE_AS_JDBC_COMPLIANT_AS_POSSIBLE=true

18.1.  Known issues with Sybase

  • The "DYNAMIC_PREPARE" parameter of the Sybase JDBC driver cannot be used with OpenJPA.

  • Datastore locking cannot be used when manipulating many-to-many relations using the default OpenJPA schema created by the schematool, unless an auto-increment primary key field is manually added to the table.

  • Persisting a zero-length string results in a string with a single space characted being returned from Sybase, Inc.'s JDBC driver.

  • The BE_AS_JDBC_COMPLIANT_AS_POSSIBLE is required in order to use datastore (pessimistic) locking. Failure to set this property may lead to obscure errors like " FOR UPDATE can not be used in a SELECT which is not part of the declaration of a cursor or which is not inside a stored procedure. ".