com.avaje.ebeaninternal.api.SpiTransaction Maven / Gradle / Ivy
package com.avaje.ebeaninternal.api;
import java.sql.Connection;
import java.util.List;
import com.avaje.ebean.Transaction;
import com.avaje.ebean.annotation.DocStoreMode;
import com.avaje.ebean.bean.PersistenceContext;
import com.avaje.ebean.event.changelog.BeanChange;
import com.avaje.ebean.event.changelog.ChangeSet;
import com.avaje.ebeaninternal.server.core.PersistRequest;
import com.avaje.ebeaninternal.server.core.PersistRequestBean;
import com.avaje.ebeaninternal.server.persist.BatchControl;
/**
* Extends Transaction with additional API required on server.
*
* Provides support for batching and TransactionContext.
*
*/
public interface SpiTransaction extends Transaction {
/**
* Return the string prefix with the transaction id and label used in logging.
*/
String getLogPrefix();
/**
* Return true if generated SQL and Bind values should be logged to the
* transaction log.
*/
boolean isLogSql();
/**
* Return true if summary level events should be logged to the transaction
* log.
*/
boolean isLogSummary();
/**
* Log a message to the SQL logger.
*/
void logSql(String msg);
/**
* Log a message to the SUMMARY logger.
*/
void logSummary(String msg);
/**
* Register a "Derived Relationship" (that requires an additional update).
*/
void registerDerivedRelationship(DerivedRelationshipData assocBean);
/**
* Return the list of "Derived Relationships" that must be maintained after
* insert.
*/
List getDerivedRelationship(Object bean);
/**
* Add a deleting bean to the registered list.
*
* This is to handle bi-directional relationships where both sides Cascade.
*
*/
void registerDeleteBean(Integer hash);
/**
* Unregister the hash of the bean.
*/
void unregisterDeleteBean(Integer hash);
/**
* Return true if this is a bean that has already been saved/deleted.
*/
boolean isRegisteredDeleteBean(Integer hash);
/**
* Unregister the persisted bean.
*/
void unregisterBean(Object bean);
/**
* Return true if this is a bean that has already been persisted in the
* current recursive save request. The goal is to stop recursively saving
* the bean when cascade persist is on both sides of a relationship).
*
* This will register the bean if it is not already.
*
*/
boolean isRegisteredBean(Object bean);
/**
* Returns a String used to identify the transaction. This id is used for
* Transaction logging.
*/
String getId();
/**
* Return true if this transaction has updateAllLoadedProperties set.
* If null is returned the server default is used (set on ServerConfig).
*/
Boolean isUpdateAllLoadedProperties();
/**
* Return the batchSize specifically set for this transaction or 0.
*
* Returning 0 implies to use the system wide default batch size.
*
*/
DocStoreMode getDocStoreMode();
/**
* Return the batch size to us for ElasticSearch Bulk API calls
* as a result of this transaction.
*/
int getDocStoreBatchSize();
/**
* Return the batchSize specifically set for this transaction or 0.
*
* Returning 0 implies to use the system wide default batch size.
*
*/
int getBatchSize();
/**
* Modify and return the current 'depth' of the transaction.
*
* As we cascade save or delete we traverse the object graph tree. Going up
* to Assoc Ones the depth decreases and going down to Assoc Manys the depth
* increases.
*
*
* The depth is used for ordering batching statements. The lowest depth get
* executed first during save.
*
*/
void depth(int diff);
/**
* Return the current depth.
*/
int depth();
/**
* Return true if this transaction was created explicitly via
* Ebean.beginTransaction()
.
*/
boolean isExplicit();
/**
* Get the object that holds the event details.
*
* This information is used maintain the table state, cache and text
* indexes. On commit the Table modifications this generates is broadcast
* around the cluster (if you have a cluster).
*
*/
TransactionEvent getEvent();
/**
* Whether persistCascade is on for save and delete.
*/
boolean isPersistCascade();
/**
* Return true if this request should be batched. Conversely returns false
* if this request should be executed immediately.
*/
boolean isBatchThisRequest(PersistRequest.Type type);
/**
* Return the BatchControl used to batch up persist requests.
*/
BatchControl getBatchControl();
/**
* Set the BatchControl used to batch up persist requests. There should only be one
* PersistQueue set per transaction.
*/
void setBatchControl(BatchControl control);
/**
* Return the persistence context associated with this transaction.
*
* You may wish to hold onto this and set it against another transaction
* later. This is along the lines of 'extended persistence context'
* behaviour.
*
*/
PersistenceContext getPersistenceContext();
/**
* Set the persistence context to this transaction.
*
* This could be considered similar to 'EJB3 Extended Persistence Context'.
* In that you can get the PersistenceContext from a transaction, hold onto
* it, and then set it back later to a second transaction. In general there
* is one PersistenceContext per Transaction. The getPersistenceContext()
* and setPersistenceContext() enable a developer to reuse a single
* PersistenceContext with multiple transactions.
*
*/
void setPersistenceContext(PersistenceContext context);
/**
* Return the underlying Connection for internal use.
*
* If the connection is made from Transaction and the user code calls
* that method we can no longer trust the query only status of a
* Transaction.
*
*/
Connection getInternalConnection();
/**
* Return true if the manyToMany intersection should be persisted for this particular relationship direction.
*/
boolean isSaveAssocManyIntersection(String intersectionTable, String beanName);
/**
* Return true if batch mode got escalated for this request (and associated cascades).
*/
boolean checkBatchEscalationOnCascade(PersistRequestBean> request);
/**
* If batch mode was turned on for the request then flush the batch.
*/
void flushBatchOnCascade();
/**
* Mark the transaction explicitly as not being query only.
*/
void markNotQueryOnly();
/**
* Potentially escalate batch mode on saving or deleting a collection.
*/
void checkBatchEscalationOnCollection();
/**
* Flush batch if we escalated batch mode on saving or deleting a collection.
*/
void flushBatchOnCollection();
/**
* Add a bean change to the change log.
*/
void addBeanChange(BeanChange beanChange);
/**
* Send the change set to be prepared and then logged.
*/
void sendChangeLog(ChangeSet changeSet);
}