org.eclipse.persistence.config.QueryHints Maven / Gradle / Ivy
Show all versions of eclipselink Show documentation
/*
* Copyright (c) 1998, 2022 Oracle and/or its affiliates. All rights reserved.
* Copyright (c) 1998, 2021 IBM Corporation. All rights reserved.
*
* This program and the accompanying materials are made available under the
* terms of the Eclipse Public License v. 2.0 which is available at
* http://www.eclipse.org/legal/epl-2.0,
* or the Eclipse Distribution License v. 1.0 which is available at
* http://www.eclipse.org/org/documents/edl-v10.php.
*
* SPDX-License-Identifier: EPL-2.0 OR BSD-3-Clause
*/
// Contributors:
// Oracle - initial API and implementation from Oracle TopLink
// 07/16/2009-2.0 Guy Pelletier
// - 277039: JPA 2.0 Cache Usage Settings
// 10/29/2010-2.2 Michael O'Brien
// - 325167: Make reserved # bind parameter char generic to enable native SQL pass through
// 06/30/2011-2.3.1 Guy Pelletier
// - 341940: Add disable/enable allowing native queries
// 09/03/2015 - Will Dazey
// - 456067 : Added support for defining query timeout units
package org.eclipse.persistence.config;
import java.util.concurrent.TimeUnit;
import org.eclipse.persistence.annotations.BatchFetch;
import org.eclipse.persistence.annotations.BatchFetchType;
import org.eclipse.persistence.sessions.DataRecord;
/**
* The class defines EclipseLink query hints.
* These query hints allow a JPA Query to be customized or optimized beyond
* what is available in the JPA specification.
*
* JPA Query Hint Usage:
*
*
query.setHint(QueryHints.CACHE_USAGE, CacheUsage.CheckCacheOnly);
*
or
*
@QueryHint(name=QueryHints.CACHE_USAGE, value=CacheUsage.CheckCacheOnly)
*
*
Hint values are case-insensitive; "" could be used instead of default value.
*
* @see HintValues
* @see CacheUsage
* @see PessimisticLock
* @see QueryType
*/
public class QueryHints {
/**
* "eclipselink.jdbc.bind-parameters"
*
Configures parameter binding to be disabled or enabled just for this query (overrides persistent unit setting, which default to true).
* Valid values are: HintValues.PERSISTENCE_UNIT_DEFAULT, HintValues.TRUE, HintValues.FALSE,
* "" could be used instead of default value HintValues.PERSISTENCE_UNIT_DEFAULT
* @see PersistenceUnitProperties#JDBC_BIND_PARAMETERS
* @see org.eclipse.persistence.queries.DatabaseQuery#setShouldBindAllParameters(boolean)
*/
public static final String BIND_PARAMETERS = "eclipselink.jdbc.bind-parameters";
/**
* "eclipselink.jdbc.parameter-delimiter"
*
Configures parameter binding char to a user defined character (overrides default hash symbol).
* Valid values are: ParameterDelimiterType.DEFAULT, other valid single character,
* "" can not be used and will instead default to ParameterDelimiterType.DEFAULT
* @see PersistenceUnitProperties#JDBC_BIND_PARAMETERS
* @see org.eclipse.persistence.queries.DatabaseQuery#setShouldBindAllParameters(boolean)
*/
public static final String PARAMETER_DELIMITER = "eclipselink.jdbc.parameter-delimiter";
/**
* "jakarta.persistence.cache.retrieveMode"
*
Configures the behavior when data is retrieved by the find methods and
* by the execution of queries. The cache retrieve mode is ignored for the
* refresh method, which always causes data to be retrieved from the
* database and not the cache.
* @see org.eclipse.persistence.queries.DatabaseQuery#dontMaintainCache()
* @see org.eclipse.persistence.queries.ObjectLevelReadQuery#shouldCheckDescriptorForCacheUsage()
*/
public static final String CACHE_RETRIEVE_MODE = "jakarta.persistence.cache.retrieveMode";
/**
* "jakarta.persistence.cache.storeMode"
* Configures the behavior when data is read from the database and when
* data is committed into the database.
* @see org.eclipse.persistence.queries.DatabaseQuery#dontMaintainCache()
* @see org.eclipse.persistence.queries.DatabaseQuery#maintainCache()
* @see org.eclipse.persistence.queries.ObjectLevelReadQuery#refreshIdentityMapResult()
*/
public static final String CACHE_STORE_MODE = "jakarta.persistence.cache.storeMode";
/**
* "eclipselink.cache-usage"
* Configures the query to utilize the EclipseLink cache, by default the cache is not checked on queries before accessing the database.
* Valid values are all declared in CacheUsage class.
* For primary key cache hits the QUERY_TYPE hint must also be set to QueryType.ReadObject.
* @see CacheUsage
* @see #QUERY_TYPE
* @see org.eclipse.persistence.queries.ReadObjectQuery
* @see org.eclipse.persistence.queries.ObjectLevelReadQuery#setCacheUsage(int)
*/
public static final String CACHE_USAGE = "eclipselink.cache-usage";
/**
* "eclipselink.query-results-cache"
*
Configures the query to use a results cache.
* By default the query will cache 100 query results, such that the same named query with the same arguments
* is re-executed it will skip the database and just return the cached results.
* Valid values are: HintValues.PERSISTENCE_UNIT_DEFAULT, HintValues.TRUE, HintValues.FALSE,
* "" could be used instead of default value HintValues.PERSISTENCE_UNIT_DEFAULT.
* By default query results are not cached.
* This is not, and is independent from the object cache.
* @see org.eclipse.persistence.queries.QueryResultsCachePolicy
* @see org.eclipse.persistence.queries.ReadQuery#setQueryResultsCachePolicy(org.eclipse.persistence.queries.QueryResultsCachePolicy)
*/
public static final String QUERY_RESULTS_CACHE = "eclipselink.query-results-cache";
/**
* "eclipselink.query-results-cache.size"
*
Configures the size of the query's results cache.
* By default the query will cache 100 query results, such that the same named query with the same arguments
* is re-executed it will skip the database and just return the cached results.
* If the query has no arguments a size of 1 should be used (as there is only a single result).
* Valid values are Integer or Strings that can be parsed to int values.
* @see org.eclipse.persistence.queries.QueryResultsCachePolicy#setMaximumCachedResults(int)
*/
public static final String QUERY_RESULTS_CACHE_SIZE = "eclipselink.query-results-cache.size";
/**
* "eclipselink.query-results-cache.expiry"
*
Configures the time to live, or expiry time of the query's results cache.
* By default the query results cache will not expiry results.
* Valid values are number of milliseconds, Integer or Strings that can be parsed to int values.
* @see org.eclipse.persistence.descriptors.invalidation.CacheInvalidationPolicy
* @see org.eclipse.persistence.queries.QueryResultsCachePolicy#setCacheInvalidationPolicy(org.eclipse.persistence.descriptors.invalidation.CacheInvalidationPolicy)
*/
public static final String QUERY_RESULTS_CACHE_EXPIRY = "eclipselink.query-results-cache.expiry";
/**
* "eclipselink.query-results-cache.randomize-expiry"
*
Configures a randomization on the expiry invalidation time.
* This can be used to avoid bottlenecks from the cached values expiring at the same time.
* By default expiry is not randomized.
* Valid values are "true" and "false", false is the default.
* @see org.eclipse.persistence.descriptors.invalidation.TimeToLiveCacheInvalidationPolicy#setIsInvalidationRandomized(boolean)
*/
public static final String QUERY_RESULTS_CACHE_RANDOMIZE_EXPIRY = "eclipselink.query-results-cache.randomize-expiry";
/**
* "eclipselink.query-results-cache.ignore-null"
*
Configures null results to not be cached.
* This can be used to use the query cache as a secondary key index, and allow inserts of new objects.
* By default null results are cached.
* Valid values are "true" and "false", false is the default.
* @see org.eclipse.persistence.queries.QueryResultsCachePolicy#setIsNullIgnored(boolean)
*/
public static final String QUERY_RESULTS_CACHE_IGNORE_NULL = "eclipselink.query-results-cache.ignore-null";
/**
* "eclipselink.query-results-cache.invalidate-on-change"
*
Configures if the query cache results should be invalidated if any object of any class used in the query if modified.
*
* Valid values are "true" and "false", true is the default.
* @see org.eclipse.persistence.queries.QueryResultsCachePolicy#setInvalidateOnChange(boolean)
*/
public static final String QUERY_RESULTS_CACHE_INVALIDATE = "eclipselink.query-results-cache.invalidate-on-change";
/**
* "eclipselink.query-results-cache.type"
*
Configures the cache type of the query's results cache.
* By default the a fixed sized LRU cache is used (CACHE).
* Valid values are defined in the CacheType enum (as Strings).
* @see org.eclipse.persistence.annotations.CacheType
* @see org.eclipse.persistence.queries.QueryResultsCachePolicy#setCacheType(java.lang.Class)
*/
public static final String QUERY_RESULTS_CACHE_TYPE = "eclipselink.query-results-cache.type";
/**
* "eclipselink.query-results-cache.expiry-time-of-day"
*
Configures the time of day expiry time of the query's results cache.
* By default the query results cache will not expiry results.
* Valid values are String Time format, "HH:MM:SS".
* @see org.eclipse.persistence.descriptors.invalidation.CacheInvalidationPolicy
* @see org.eclipse.persistence.queries.QueryResultsCachePolicy#setCacheInvalidationPolicy(org.eclipse.persistence.descriptors.invalidation.CacheInvalidationPolicy)
*/
public static final String QUERY_RESULTS_CACHE_EXPIRY_TIME_OF_DAY = "eclipselink.query-results-cache.expiry-time-of-day";
/**
* "eclipselink.query.redirector"
*
Used to provide a QueryRedirector to the executing query.
* The redirector must implement the QueryRedirector interface.
* This can be used to perform advanced query operations in code,
* or dynamically customize the query in code before execution.
* The value should be the name of the QueryRedirector class.
*
i.e. "org.acme.persistence.MyQueryRedirector"
*
The value could also be a Class, or an instance that implements the QueryRedirector interface.
* @see org.eclipse.persistence.queries.QueryRedirector
* @see org.eclipse.persistence.queries.DatabaseQuery#setRedirector(org.eclipse.persistence.queries.QueryRedirector)
*/
public static final String QUERY_REDIRECTOR = "eclipselink.query.redirector";
/**
* "eclipselink.partitioning"
*
Used to provide a PartitioningPolicy instance or name to the query.
* This allows the query to execute on a specific, or on multiple connection pools.
* @see org.eclipse.persistence.annotations.Partitioning
* @see org.eclipse.persistence.descriptors.partitioning.PartitioningPolicy
*/
public static final String PARTITIONING = "eclipselink.partitioning";
/**
* "eclipselink.query-type"
*
Configures the EclipseLink query type to use for the query.
* By default EclipseLink ReportQuery or ReadAllQuery are used for most JPQL queries, this allows other query types to be used,
* such as ReadObjectQuery which can be used for queries that are know to return a single object, and has different caching semantics.
* Valid values are all declared in QueryType class.
* A fully qualified class name of a valid subclass of DatabaseQuery can also be used.
*
i.e. "org.acme.persistence.CustomQuery"
* @see QueryType
*/
public static final String QUERY_TYPE = "eclipselink.query-type";
/**
* "eclipselink.pessimistic-lock"
*
Configures the query to acquire a pessimistic lock (write-lock) on the resulting rows in the database.
* Valid values are all declared in PessimisticLock class.
* Pessimistic locking support and behavior may differ on difference database platforms.
* @see PessimisticLock
* @see org.eclipse.persistence.queries.ObjectLevelReadQuery#setLockMode(short)
*
* Note: As of JPA 2.0 there is a standard way to configure pessimistic locking.
* @see jakarta.persistence.LockModeType
* @see jakarta.persistence.EntityManager (find(), refresh(), lock())
* @see jakarta.persistence.Query (setLockMode())
*/
public static final String PESSIMISTIC_LOCK = "eclipselink.pessimistic-lock";
/**
* "jakarta.persistence.lock.timeout"
*
Configures the WAIT timeout used in pessimistic locking, if the database
* query exceeds the timeout the database will terminate the query and
* return an exception. Valid values are Integer or Strings that can be
* parsed to int values.
* Some database platforms may not support lock timeouts, you may consider
* setting a JDBC_TIMEOUT hint for these platforms.
* @see #JDBC_TIMEOUT
* @see org.eclipse.persistence.queries.ObjectLevelReadQuery#setWaitTimeout(Integer)
*/
public static final String PESSIMISTIC_LOCK_TIMEOUT = "jakarta.persistence.lock.timeout";
/**
* "eclipselink.pessimistic.lock.timeout.unit"
*
Configures the pessimistic lock timeout unit value. Allows users more refinement.
* Valid Values:
*
* - "
java.util.concurrent.TimeUnit.MILLISECONDS" (DEFAULT),
* - "
java.util.concurrent.TimeUnit.SECONDS",
* - "
java.util.concurrent.TimeUnit.MINUTES".
*
* @see org.eclipse.persistence.queries.ObjectLevelReadQuery#setWaitTimeoutUnit(TimeUnit)
*/
public static final String PESSIMISTIC_LOCK_TIMEOUT_UNIT = "eclipselink.pessimistic.lock.timeout.unit";
/**
* "eclipselink.refresh"
* Configures the query to refresh the resulting objects in the cache and persistent context with the current state of the database.
* This will also refresh the objects in the shared cache, unless a flush has occurred.
* Any unflushed changes made to the objects will be lost (unless this query triggers a flush before execution).
* The refresh will cascade relationships based on the REFRESH_CASCADE hint value.
* Valid values are: HintValues.FALSE, HintValues.TRUE,
* "" could be used instead of default value HintValues.FALSE
* @see #REFRESH_CASCADE
* @see org.eclipse.persistence.queries.ObjectLevelReadQuery#setShouldRefreshIdentityMapResult(boolean)
*/
public static final String REFRESH = "eclipselink.refresh";
/**
* "eclipselink.batch"
*
Configures the query to optimize the retrieval of the related objects,
* the related objects for all the resulting objects will be read in a single query (instead of n queries).
* Batch reading is normally more efficient than join fetch, especially for collection relationships.
* Valid values are strings that represent JPQL style navigations to a relationship.
*
e.g. "e.manager.phoneNumbers"
*
Note: Batch reading produces two lines of SQL to read the requested object graph and uses the
* 'where clause' of the first SQL as part of the 'where clause' of the second. When indirect attributes
* of the objects returned from the first query are referenced the second query is executed to return those attributes.
* If changes are made to data referenced in the 'where clause' after the first SQL results are returned but before the
* second query executes then these attributes may not be returned by the second query. To prevent this reference these
* attributes before changing data associated with the query."
* @see BatchFetch
* @see #BATCH_TYPE
* @see BatchFetchType
* @see org.eclipse.persistence.queries.ReadAllQuery#addBatchReadAttribute(String)
*/
public static final String BATCH = "eclipselink.batch";
/**
* "eclipselink.batch.type"
*
Configures the type of batch fetching to use for any batch fetched relationships on the query.
* Valid values are defined in BatchFetchType ("JOIN", "EXISTS", "IN").
* @see #BATCH
* @see BatchFetchType
* @see BatchFetch
* @see org.eclipse.persistence.queries.ObjectLevelReadQuery#setBatchFetchType(BatchFetchType)
*/
public static final String BATCH_TYPE = "eclipselink.batch.type";
/**
* "eclipselink.batch.size"
*
Configures the batch size for batch fetching using the IN batch type.
* Valid values are integer values {@literal >} 0.
* The default size is 256, or the cursor pageSize.
* @see #BATCH
* @see #BATCH_TYPE
* @see BatchFetchType
* @see BatchFetch
* @see org.eclipse.persistence.queries.ObjectLevelReadQuery#setBatchFetchSize(int)
*/
public static final String BATCH_SIZE = "eclipselink.batch.size";
/**
* "eclipselink.join-fetch"
*
Configures the query to optimize the retrieval of the related objects,
* the related objects will be joined into the query instead of being queried independently.
* This allow for nested join fetching which is not supported in JPQL.
* It also allows for join fetching with native queries.
* This uses an INNER join and will filter null or empty values, see LEFT_FETCH for outer joins.
* Valid values are strings that represent JPQL style navigations to a relationship.
*
e.g. e.manager.phoneNumbers
* @see #BATCH
* @see #LEFT_FETCH
* @see org.eclipse.persistence.queries.ObjectLevelReadQuery#addJoinedAttribute(String)
*/
public static final String FETCH = "eclipselink.join-fetch";
/**
* "eclipselink.left-join-fetch"
*
Configures the query to optimize the retrieval of the related objects,
* the related objects will be joined into the query instead of being queried independently.
* This allow for nested join fetching which is not supported in JPQL.
* It also allows for join fetching with native queries.
* This uses an OUTER join to allow null or empty values.
* Valid values are strings that represent JPQL style navigations to a relationship.
*
e.g. e.manager.phoneNumbers
* @see #BATCH
* @see #FETCH
* @see org.eclipse.persistence.queries.ObjectLevelReadQuery#addJoinedAttribute(String)
*/
public static final String LEFT_FETCH = "eclipselink.left-join-fetch";
/**
* "eclipselink.read-only"
*
Configures the query to return shared (read-only) objects from the cache,
* instead of objects registered with the persistence context.
* This improves performance by avoiding the persistence context registration and change tracking overhead to read-only objects.
* Valid values are: HintValues.FALSE, HintValues.TRUE,
* "" could be used instead of default value HintValues.FALSE
* @see org.eclipse.persistence.queries.ObjectLevelReadQuery#setIsReadOnly(boolean)
*/
public static final String READ_ONLY = "eclipselink.read-only";
/**
* "eclipselink.jdbc.timeout"
*
Configures the JDBC timeout of the query execution, if the database query exceeds the timeout
* the database will terminate the query and return an exception.
* Valid values are Integer or Strings that can be parsed to int values.
* @see org.eclipse.persistence.queries.DatabaseQuery#setQueryTimeout(int)
*/
public static final String JDBC_TIMEOUT = "eclipselink.jdbc.timeout";
/**
* "jakarta.persistence.query.timeout"
*
Configures the default query timeout value per the JPA specification.
* Valid values are strings containing a zero or greater integer value
* Defaults to use seconds as the unit of time.
* @see org.eclipse.persistence.queries.DatabaseQuery#setQueryTimeout(int)
*/
public static final String QUERY_TIMEOUT = "jakarta.persistence.query.timeout";
/**
* "eclipselink.jdbc.timeout.unit"
*
Configures the query timeout unit value. Allows users more refinement.
* Valid Values:
*
* - "
java.util.concurrent.TimeUnit.MILLISECONDS" (DEFAULT),
* - "
java.util.concurrent.TimeUnit.SECONDS",
* - "
java.util.concurrent.TimeUnit.MINUTES".
*
* @see org.eclipse.persistence.queries.DatabaseQuery#setQueryTimeoutUnit(TimeUnit)
*/
public static final String QUERY_TIMEOUT_UNIT = "eclipselink.query.timeout.unit";
/**
* "eclipselink.jdbc.fetch-size"
* Configures the JDBC fetch-size for the queries result-set.
* This can improve the performance for queries that return large result-sets.
* Valid values are Integer or Strings that can be parsed to int values.
* @see org.eclipse.persistence.queries.ReadQuery#setFetchSize(int)
*/
public static final String JDBC_FETCH_SIZE = "eclipselink.jdbc.fetch-size";
/**
* "eclipselink.jdbc.max-rows"
*
Configures the JDBC max-rows, if the query returns more rows than the max-rows
* the trailing rows will not be returned by the database.
* This is the same as JPA Query setMaxResults(), but can be set in meta-data for NamedQuerys.
* Valid values are Integer or Strings that can be parsed to int values.
* @see jakarta.persistence.Query#setMaxResults(int)
* @see org.eclipse.persistence.queries.ReadQuery#setMaxRows(int)
*/
public static final String JDBC_MAX_ROWS = "eclipselink.jdbc.max-rows";
/**
* "eclipselink.jdbc.first-result"
*
Configures the query to skip the firstResult number of rows.
* This is the same as JPA Query setFirstResults(), but can be set in meta-data for NamedQuerys.
* Valid values are Integer or Strings that can be parsed to int values.
* @see jakarta.persistence.Query#setFirstResult(int)
* @see org.eclipse.persistence.queries.ReadQuery#setFirstResult(int)
*/
public static final String JDBC_FIRST_RESULT = "eclipselink.jdbc.first-result";
/**
* "eclipselink.result-collection-type"
*
Configures the collection class implementation for the queries result.
* The fully qualified class name must be used, without the .class.
* Valid values are a Class representing a collection type or a String representing the class' name
* of the collection type.
* If a Collection type that is not a List is used, getResultCollection() or getSingleResult() must be used instead of
* getResultList().
*
e.g. "java.util.ArrayList"
* @see "org.eclipse.persistence.jpa.JpaQuery#getResultCollection()"
* @see org.eclipse.persistence.queries.ReadAllQuery#useCollectionClass(Class)
*/
public static final String RESULT_COLLECTION_TYPE = "eclipselink.result-collection-type";
/**
* "eclipselink.refresh.cascade"
*
Defines if a refresh query should cascade the refresh to relationships.
* Valid values are all declared in CascadePolicy class.
* The REFRESH hint should also be set to cause a refresh.
* @see CascadePolicy
* @see #REFRESH
* @see org.eclipse.persistence.queries.DatabaseQuery#setCascadePolicy(int)
*/
public static final String REFRESH_CASCADE = "eclipselink.refresh.cascade";
/**
* "eclipselink.maintain-cache"
*
Configures the query to not use both the shared cache, and the transactional cache/persistence context.
* Resulting objects will be read and built directly from the database, and not registered in the persistence context.
* Changes made to the objects will not be updated unless merged, object identity will not be maintained.
* This can be used to read the current state of the database, without affecting the current persistence context.
* By default the cache is always maintained.
* Valid values are: HintValues.FALSE, HintValues.TRUE,
* "" could be used instead of default value HintValues.FALSE
* @see org.eclipse.persistence.queries.DatabaseQuery#setShouldMaintainCache(boolean)
*/
public static final String MAINTAIN_CACHE = "eclipselink.maintain-cache";
/**
* "eclipselink.prepare"
*
Configures the query to not prepare its SQL.
* By default queries generate their SQL the first time they are executed,
* and avoid the cost of generating the SQL on subsequent executions.
* This can be used to generate the SQL on every execution if the
* query requires usage of dynamic SQL, such as to handle null parameters (use IS NULL instead of = null).
* This only effects the SQL generation, not parameter binding or statement caching.
* Valid values are: HintValues.FALSE, HintValues.TRUE,
* "" could be used instead of default value HintValues.FALSE
* @see org.eclipse.persistence.queries.DatabaseQuery#setShouldPrepare(boolean)
*/
public static final String PREPARE = "eclipselink.prepare";
/**
* "eclipselink.jdbc.cache-statement"
*
Configures if the query will cache its JDBC statement.
* This allows queries to use parameterized SQL with statement caching.
* It also allows a specific query to not cache its statement,
* if statement caching is enable for the persistence unit.
* If statement caching is desired, it should normally be set for the entire persistence unit
* in the persistence.xml, not in each query.
* If a DataSource is used statement caching must be set in the DataSource configuration,
* not in EclipseLink.
* Valid values are: HintValues.FALSE, HintValues.TRUE,
* "" could be used instead of default value HintValues.FALSE
* @see PersistenceUnitProperties#CACHE_STATEMENTS
* @see org.eclipse.persistence.queries.DatabaseQuery#setShouldCacheStatement(boolean)
*/
public static final String CACHE_STATMENT = "eclipselink.jdbc.cache-statement";
/**
* "eclipselink.flush"
*
Configures if the query should trigger a flush of the persistence context before execution.
* If the query may access objects that have been changed in the persistence context,
* trigger a flush is required for the query to see these changes.
* If the query does not require seeing the changes, then avoid the flush can improve performance.
* The default flush-mode can be set on the EntityManager or configured as a persistence unit property.
* By default the flush-mode is AUTO, which requires a flush before any query execution.
* Conforming can also be used to query changes without requiring a flush,
* refer to the CACHE_USAGE query hint for conforming.
* Valid values are: HintValues.FALSE, HintValues.TRUE,
* "" could be used instead of default value HintValues.FALSE
* @see jakarta.persistence.EntityManager#setFlushMode(jakarta.persistence.FlushModeType)
* @see PersistenceUnitProperties#PERSISTENCE_CONTEXT_FLUSH_MODE
* @see org.eclipse.persistence.queries.DatabaseQuery#setFlushOnExecute(Boolean)
*/
public static final String FLUSH = "eclipselink.flush";
/**
* "eclipselink.sql.hint"
*
* Sets a SQL hint string into the query that will be generated into the SQL statement after
* the SELECT/INSERT/UPDATE/DELETE command.
*
* A SQL hint can be used on certain database platforms to define how the query uses indexes
* and other such low level usages. This should be the full hint string including the comment \
* delimiters.
*
* @see org.eclipse.persistence.queries.DatabaseQuery#setHintString(String)
*/
public static final String HINT = "eclipselink.sql.hint";
/**
* "eclipselink.jdbc.native-connection"
*
Configures if the query requires a native JDBC connection.
* This may be required for some queries on some server platforms that
* have DataSource implementations that wrap the JDBC connection in their own proxy.
* If the query requires custom JDBC access, it may require a native connection.
* A ServerPlatform is required to be set as a persistence property to be able to use a native connection.
* For features that are known by EclipseLink to require native connections this
* will default to true, otherwise is false.
* Valid values are: HintValues.FALSE, HintValues.TRUE,
* "" could be used instead of default value HintValues.FALSE
* @see PersistenceUnitProperties#TARGET_SERVER
* @see org.eclipse.persistence.queries.DatabaseQuery#setIsNativeConnectionRequired(boolean)
*/
public static final String NATIVE_CONNECTION = "eclipselink.jdbc.native-connection";
/**
* "eclipselink.cursor"
*
Configures the query to return a CursoredStream.
* A cursor is a stream of the JDBC ResultSet.
* Cursor implements Enumeration, when the each next() will fetch the next from the JDBC ResultSet,
* and build the resulting Object or value.
* A Cursor requires and will keep a live JDBC connection, close() must be called
* to free the Cursor's resources.
* A Cursor can be accessed from a JPA Query through getSingleResult(), or from JpaQuery using getResultCursor().
* Cursors are useful for large results sets, and if only the first few results are desired.
* MAX_ROWS and FIRST_RESULT can also be used instead of cursors to obtain a page of results.
* Valid values are: HintValues.FALSE, HintValues.TRUE,
* "" could be used instead of default value HintValues.FALSE
* @see "org.eclipse.persistence.jpa.JpaQuery#getSingleResult()"
* @see "org.eclipse.persistence.jpa.JpaQuery#getResultCursor()"
* @see org.eclipse.persistence.queries.Cursor
* @see org.eclipse.persistence.queries.CursoredStream
* @see org.eclipse.persistence.queries.ReadAllQuery#useCursoredStream()
*/
public static final String CURSOR = "eclipselink.cursor";
/**
* "eclipselink.cursor.initial-size"
*
Configures the query to return a CursoredStream with the initial threshold size.
* The initial size is the initial number of objects that are prebuilt for the stream before a next() is called.
* Valid values are Integer or Strings that can be parsed to int values.
* @see org.eclipse.persistence.queries.ReadAllQuery#useCursoredStream(int, int)
*/
public static final String CURSOR_INITIAL_SIZE = "eclipselink.cursor.initial-size";
/**
* "eclipselink.cursor.page-size"
*
Configures the query to return a CursoredStream with the page size.
* The page size is the number of objects that are fetched from the stream on a next() called,
* if the buffer of objects is empty.
* Valid values are Integer or Strings that can be parsed to int values.
* @see org.eclipse.persistence.queries.ReadAllQuery#useCursoredStream(int, int)
*/
public static final String CURSOR_PAGE_SIZE = "eclipselink.cursor.page-size";
/**
* "eclipselink.cursor.size-sql"
*
Configures the SQL string for the size query of a Cursor query.
* This is only required for cursor queries that use native SQL or procedures.
* The size query is only used if the size() is called on the Cursor.
* The SQL should perform a COUNT of the rows returned by the original query.
* @see #CURSOR
* @see org.eclipse.persistence.queries.Cursor#size()
* @see org.eclipse.persistence.queries.ReadAllQuery#useCursoredStream(int, int, org.eclipse.persistence.queries.ValueReadQuery)
*/
public static final String CURSOR_SIZE = "eclipselink.cursor.size-sql";
/**
* "eclipselink.cursor.scrollable"
*
Configures the query to return a ScrollableCursor.
* A cursor is a stream of the JDBC ResultSet.
* ScrollableCursor implements ListIterator, when the each next() will fetch the next from the JDBC ResultSet,
* and build the resulting Object or value.
* ScrollableCursor can scroll forwards and backwards and position into the ResultSet.
* A Cursor requires and will keep a live JDBC connection, close() must be called
* to free the Cursor's resources.
* A Cursor can be accessed from a JPA Query through getSingleResult(), or from JpaQuery using getResultCursor().
* Cursors are useful for large results sets, and if only some of the results are desired.
* MAX_ROWS and FIRST_RESULT can also be used instead of cursors to obtain a page of results.
* Valid values are: HintValues.FALSE, HintValues.TRUE,
* "" could be used instead of default value HintValues.FALSE
* @see "org.eclipse.persistence.jpa.JpaQuery#getSingleResult()"
* @see "org.eclipse.persistence.jpa.JpaQuery#getResultCursor()"
* @see org.eclipse.persistence.queries.Cursor
* @see org.eclipse.persistence.queries.ScrollableCursor
* @see org.eclipse.persistence.queries.ReadAllQuery#useScrollableCursor()
*/
public static final String SCROLLABLE_CURSOR = "eclipselink.cursor.scrollable";
/**
* "eclipselink.cursor.scrollable.result-set-type"
*
This can be used on ScrollableCursor queries to set the JDBC ResultSet scroll type.
* @see ResultSetType
* @see #SCROLLABLE_CURSOR
* @see org.eclipse.persistence.queries.ScrollableCursorPolicy#setResultSetType(int)
*/
public static final String RESULT_SET_TYPE = "eclipselink.cursor.scrollable.result-set-type";
/**
* "eclipselink.cursor.scrollable.result-set-concurrency"
*
This can be used on ScrollableCursor queries to set the JDBC ResultSet concurrency.
* @see ResultSetConcurrency
* @see #SCROLLABLE_CURSOR
* @see org.eclipse.persistence.queries.ScrollableCursorPolicy#setResultSetConcurrency(int)
*/
public static final String RESULT_SET_CONCURRENCY = "eclipselink.cursor.scrollable.result-set-concurrency";
/**
* "eclipselink.fetch-group"
*
Configures the query to use the fetch group object.
* This is an instance of FetchGroup.
* The query will only fetch the attributes defined in the fetch group, if any other attribute is accessed
* it will cause the object to be refreshed.
* To load all FetchGroup's relationship attributes set the FetchGroup's boolean flag "shouldLoad" to true.
* FetchGroups are only supported for queries returning objects (only a single alias can be the select clause).
* Weaving is required to allow usage of fetch groups.
* @see #FETCH_GROUP_NAME
* @see #FETCH_GROUP_ATTRIBUTE
* @see #FETCH_GROUP_DEFAULT
* @see #LOAD_GROUP
* @see org.eclipse.persistence.queries.FetchGroup
* @see org.eclipse.persistence.queries.FetchGroup#setShouldLoad
* @see org.eclipse.persistence.descriptors.FetchGroupManager
* @see org.eclipse.persistence.queries.ObjectLevelReadQuery#setFetchGroup(org.eclipse.persistence.queries.FetchGroup)
*/
public static final String FETCH_GROUP = "eclipselink.fetch-group";
/**
* "eclipselink.fetch-group.name"
*
Configures the query to use a named fetch group defined for the result class.
* This is the name of the fetch group, as defined on the ClassDescriptor.
* Currently FetchGroups can only be defined on the ClassDescriptor using a DescriptorCustomizer.
* The query will only fetch the attributes defined in the fetch group, if any other attribute is accessed
* it will cause the object to be refreshed.
* To load all FetchGroup's relationship attributes set the FetchGroup's boolean flag "load" to true.
* FetchGroups are only supported for queries returning objects (only a single alias can be the select clause).
* Weaving is required to allow usage of fetch groups.
* @see #FETCH_GROUP_ATTRIBUTE
* @see #FETCH_GROUP_DEFAULT
* @see #LOAD_GROUP
* @see org.eclipse.persistence.descriptors.FetchGroupManager
* @see org.eclipse.persistence.queries.FetchGroup
* @see org.eclipse.persistence.queries.ObjectLevelReadQuery#setFetchGroupName(String)
*/
public static final String FETCH_GROUP_NAME = "eclipselink.fetch-group.name";
/**
* "eclipselink.fetch-group.attribute"
*
Configures the query to use a dynamic fetch group that includes a list of attributes.
* Each attribute must be defined using a separate hint.
* The primary key and version are always included.
* The query will only fetch the attributes defined in the fetch group, if any other attribute is accessed
* it will cause the object to be refreshed.
* To load all FetchGroup's relationship attributes set FETCH_GROUP_LOAD to "true".
* FetchGroups are only supported for queries returning objects (only a single alias can be the select clause).
* Weaving is required to allow usage of fetch groups.
* Both local and nested attributes are supported.
* @see #FETCH_GROUP_NAME
* @see #FETCH_GROUP_DEFAULT
* @see #FETCH_GROUP_LOAD
* @see #LOAD_GROUP
* @see org.eclipse.persistence.queries.FetchGroup
* @see org.eclipse.persistence.queries.ObjectLevelReadQuery#setFetchGroup(org.eclipse.persistence.queries.FetchGroup)
*/
public static final String FETCH_GROUP_ATTRIBUTE = "eclipselink.fetch-group.attribute";
/**
* "eclipselink.fetch-group.load"
*
Configures a dynamic fetch group to load (default) or not to load its attributes.
* Could be used with FETCH_ATTRIBUTES hint.
* To load all FetchGroup's relationship attributes set this hint to "true".
* Weaving is required to allow usage of fetch groups.
* Valid values are: HintValues.FALSE, HintValues.TRUE,
* "" could be used instead of default value HintValues.TRUE
* @see #FETCH_GROUP_ATTRIBUTE
* @see org.eclipse.persistence.queries.FetchGroup
* @see org.eclipse.persistence.queries.FetchGroup#setShouldLoad(boolean)
*/
public static final String FETCH_GROUP_LOAD = "eclipselink.fetch-group.load";
/**
* "jakarta.persistence.fetchgraph"
*
Configures a jpa entity graph to be used as a fetch graph template.
* Those attributes included in the entity graph will have their values fetched and populated.
* Attributes of the represented entity that are not present in the entity graph will be treated
* as FetchType.LAZY.
*/
public static final String JPA_FETCH_GRAPH = "jakarta.persistence.fetchgraph";
/**
* "eclipselink.fetch-group.default"
*
Configures the query not to use the default fetch group.
* The default fetch group is defined by all non-lazy Basic mappings.
* If set to FALSE all attributes will be fetched, including lazy Basics,
* this still excludes lazy relationships, they will fetch their foreign keys, but not their values.
* FetchGroups are only supported for queries returning objects (only a single alias can be the select clause).
* Weaving is required to allow usage of fetch groups.
* Valid values are: HintValues.FALSE, HintValues.TRUE,
* "" could be used instead of default value HintValues.TRUE
* @see #FETCH_GROUP_NAME
* @see #FETCH_GROUP_ATTRIBUTE
* @see org.eclipse.persistence.queries.FetchGroup
* @see org.eclipse.persistence.queries.ObjectLevelReadQuery#setShouldUseDefaultFetchGroup(boolean)
*/
public static final String FETCH_GROUP_DEFAULT = "eclipselink.fetch-group.default";
/**
* "eclipselink.load-group"
*
Configures the query to use the load group object.
* This is an instance of LoadGroup.
* Ensures that all relational attributes an the group are loaded.
* LoadGroups are only supported for queries returning objects (only a single alias can be the select clause).
* @see #LOAD_GROUP_ATTRIBUTE
* @see org.eclipse.persistence.queries.LoadGroup
* @see org.eclipse.persistence.queries.ObjectLevelReadQuery#setLoadGroup(org.eclipse.persistence.queries.LoadGroup)
*/
public static final String LOAD_GROUP = "eclipselink.load-group";
/**
* "jakarta.persistence.loadgraph"
*
Configures a jpa entity graph to be used as a load graph template.
* Those attributes included in the entity graph will have their values fetched and populated.
* Attributes of the represented entity that are not present in the entity graph will be loaded based
* on their mapping based FetchType settings.
*/
public static final String JPA_LOAD_GRAPH = "jakarta.persistence.loadgraph";
/**
* "eclipselink.load-group.attribute"
*
Configures the query to use load group that includes a list of attributes.
* Each attribute must be defined using a separate hint.
* The query will load all relational attributes defined in the load group.
* LoadGroups are only supported for queries returning objects (only a single alias can be the select clause).
* Both local and nested attributes are supported.
* @see #LOAD_GROUP
* @see org.eclipse.persistence.queries.LoadGroup
* @see org.eclipse.persistence.queries.ObjectLevelReadQuery#setLoadGroup
*/
public static final String LOAD_GROUP_ATTRIBUTE = "eclipselink.load-group.attribute";
/**
* "eclipselink.exclusive-connection"
*
Configures the query to use the exclusive (transactional/write) connection.
* This is only relevant if a EXCLUSIVE_CONNECTION_MODE property has been set for the persistence unit (such as VPD).
* If an EXCLUSIVE_CONNECTION_MODE has been configured,
* this will ensure that the query is executed through the exclusive connection.
* This may be required in certain cases. An example being
* where database security will prevent a query joining to a secure table
* from returning the correct results when executed through the shared
* connection.
* Valid values are: HintValues.FALSE, HintValues.TRUE.
* @see PersistenceUnitProperties#EXCLUSIVE_CONNECTION_MODE
* @see org.eclipse.persistence.sessions.server.ConnectionPolicy#setExclusiveMode(org.eclipse.persistence.sessions.server.ConnectionPolicy.ExclusiveMode)
* @see org.eclipse.persistence.queries.ObjectLevelReadQuery#setShouldUseExclusiveConnection(boolean)
*/
public static final String EXCLUSIVE_CONNECTION = "eclipselink.exclusive-connection";
/**
* "eclipselink.inheritance.outer-join"
*
Configures the query to outer-join all subclasses.
* This is only relevant for queries to root or branch inherited classes.
* By default a separate query is executed for each subclass.
* This can also be configured for the class using a DescriptorCustomizer.
* This is required for correct ordering, firstResult, maxResult, and cursors.
* Valid values are: HintValues.FALSE, HintValues.TRUE.
* @see org.eclipse.persistence.descriptors.InheritancePolicy#setShouldOuterJoinSubclasses(boolean)
* @see org.eclipse.persistence.queries.ObjectLevelReadQuery#setShouldOuterJoinSubclasses(boolean)
*/
public static final String INHERITANCE_OUTER_JOIN = "eclipselink.inheritance.outer-join";
/**
* The "eclipselink.jdbc.allow-native-sql-query" property
* specifies whether a single native SQL query should override a persistence
* unit level setting (eclipselink.jdbc.allow-native-sql-queries). The PU
* level flag is of particular importance within a multitenant to minimize
* the potential impact of revealing multitenant. However in some cases the
* application may need to allow certain native SQL queries through.
*
* Allowed Values (String):
*
* - "true" - allow native SQL (and override the persistence unit flag)
*
- "false" - (DEFAULT) do not allow native SQL (and respect the persistence unit flag if set.)
*
*
* @see org.eclipse.persistence.queries.DatabaseQuery#setAllowNativeSQLQuery(Boolean) {
*/
public static final String ALLOW_NATIVE_SQL_QUERY = "eclipselink.jdbc.allow-native-sql-query";
/**
* "eclipselink.history.as-of"
* Configures the query to query the state of the object as-of a point in time.
* This can only be used if the class has been configured with historical support,
* or if Oracle Flashback is used.
* Valid values are timestamps in the form "YYYY/MM/DD HH:MM:SS.n".
* @see org.eclipse.persistence.history.HistoryPolicy
* @see org.eclipse.persistence.queries.ObjectLevelReadQuery#setAsOfClause(org.eclipse.persistence.history.AsOfClause)
*/
public static final String AS_OF = "eclipselink.history.as-of";
/**
* "eclipselink.history.as-of.scn"
*
Configures the query to query the state of the object as-of a database SCN (System Change Number).
* This can only be used with Oracle Flashback support.
* Valid values are Integer SCN values.
* @see org.eclipse.persistence.queries.ObjectLevelReadQuery#setAsOfClause(org.eclipse.persistence.history.AsOfClause)
*/
public static final String AS_OF_SCN = "eclipselink.history.as-of.scn";
/**
* "eclipselink.result-type"
*
By default in JPA for non-single select queries an Array of values is returned.
* If getSingleResult() is called the first array is returned, for getResultList() a List of arrays is returned.
*
i.e. "Select e.firstName, e.lastName from Employee e" returns {@literal List}
* or the native query, "SELECT * FROM EMPLOYEE" returns {@literal List}
* The ResultType can be used to instead return a Map of values (DatabaseRecord, ReportQueryResult).
*
It can also be used to return a single column, or single value.
* Valid values are defined in ResultType.
* @see ResultType
* @see DataRecord
* @see org.eclipse.persistence.sessions.DatabaseRecord
* @see org.eclipse.persistence.queries.ReportQueryResult
* @see org.eclipse.persistence.queries.ReportQuery#setReturnType(int)
* @see org.eclipse.persistence.queries.DataReadQuery#setResultType
*/
public static final String RESULT_TYPE = "eclipselink.result-type";
/**
* "eclipselink.cache-usage.indirection-policy"
*
This can be used on a query with a CACHE_USAGE hint to configure the behavior of in-memory
* querying and conforming's treatment of uninstantiated indirection/lazy relationships.
* This is only relevant when the query traverses a join across a lazy relationship.
* Valid values are defined in CacheUsageIndirectionPolicy.
* @see CacheUsageIndirectionPolicy
* @see #CACHE_USAGE
* @see org.eclipse.persistence.queries.ObjectLevelReadQuery#setInMemoryQueryIndirectionPolicyState(int)
*/
public static final String INDIRECTION_POLICY = "eclipselink.cache-usage.indirection-policy";
/**
* "jakarta.persistence.lock.scope"
*
By default pessimistic lock applied to only the tables mapped to the object being locked.
*
It could be extended to apply also to relation (join) tables (ManyToMany and OneToOne case),
* and CollectionTables (ElementCollection case).
* Valid values are defined in PessimisticLockScope.
* @see jakarta.persistence.PessimisticLockScope
* @see jakarta.persistence.LockModeType
*/
public static final String PESSIMISTIC_LOCK_SCOPE = "jakarta.persistence.lock.scope";
/**
* "eclipselink.composite-unit.member"
*
This must be used on a native query executed on composite persistence unit.
* It specifies the name of composite member persistence unit to execute the query on.
* @see org.eclipse.persistence.queries.DatabaseQuery#setSessionName(String)
*/
public static final String COMPOSITE_UNIT_MEMBER = "eclipselink.composite-unit.member";
/**
* "eclipselink.jdbc.batch-writing"
*
Configures if this modify query can be batched through batch writing.
* Some types of queries cannot be batched, such as DDL on some databases.
* Disabling batch writing will also allow the row count to be returned.
* Valid values are: HintValues.PERSISTENCE_UNIT_DEFAULT, HintValues.TRUE, HintValues.FALSE,
* "" could be used instead of default value HintValues.PERSISTENCE_UNIT_DEFAULT
* @see org.eclipse.persistence.queries.ModifyQuery#setIsBatchExecutionSupported(boolean)
*/
public static final String BATCH_WRITING = "eclipselink.jdbc.batch-writing";
/**
* "eclipselink.result-set-access"
*
Configures if this read object(s) query should keep result set until all result object(s) are built.
* That allows to avoid getting potentially huge objects (LOBs) from ResultSet in case they would not be used
* because the object is found in the cache.
* Valid values are: HintValues.PERSISTENCE_UNIT_DEFAULT, HintValues.TRUE, HintValues.FALSE,
* "" could be used instead of default value HintValues.PERSISTENCE_UNIT_DEFAULT
* @see org.eclipse.persistence.queries.ObjectLevelReadQuery#setIsResultSetAccessOptimizedQuery(boolean)
* @see org.eclipse.persistence.queries.ObjectLevelReadQuery#clearIsResultSetOptimizedQuery
*/
public static final String RESULT_SET_ACCESS = "eclipselink.result-set-access";
/**
* "eclipselink.serialized-object"
*
Configures if this read object(s) query use SerializedObjectPolicy if it's set on the descriptor.
* Valid values are: HintValues.TRUE, HintValues.FALSE,
* "" could be used instead of default value HintValues.FALSE
* @see org.eclipse.persistence.queries.ObjectLevelReadQuery#setShouldUseSerializedObjectPolicy(boolean)
*/
public static final String SERIALIZED_OBJECT = "eclipselink.serialized-object";
/**
* "eclipselink.query-return-name-value-pairs"
*
Configures a ResultSetMappingQuery to return populated DatabaseRecord(s)
* instead of raw data. This is particularly useful when the structure of the
* returned data is not known.
* @see org.eclipse.persistence.queries.ResultSetMappingQuery#setShouldReturnNameValuePairs(boolean)
* @see org.eclipse.persistence.queries.ResultSetMappingQuery#shouldReturnNameValuePairs()
*/
public static final String RETURN_NAME_VALUE_PAIRS = "eclipselink.query-return-name-value-pairs";
/**
* "eclipselink.inner-join-in-where-clause"
*
Changes the way that inner joins are printed in generated SQL for the database.
* With a value of true, inner joins are printed in the WHERE clause, if false, inner joins are printed in the FROM clause.
* This query hint should override global/session switch {@link org.eclipse.persistence.internal.databaseaccess.DatabasePlatform#setPrintInnerJoinInWhereClause(boolean)}
* @see org.eclipse.persistence.internal.databaseaccess.DatabasePlatform#setPrintInnerJoinInWhereClause(boolean)
*/
public static final String INNER_JOIN_IN_WHERE_CLAUSE = "eclipselink.inner-join-in-where-clause";
private QueryHints() {
// no instance please
}
}