de.ipcon.schema.generators

Class PostgreSQLIndexSynchronizer

java.lang.Object

de.ipcon.schema.generators.PostgreSQLIndexSynchronizer

public class PostgreSQLIndexSynchronizer
extends java.lang.Object

Handler for PostgreSQL Indexes (Generation, Parsing, Cleanup)

Constructor Summary

Constructors

Constructor and Description
PostgreSQLIndexSynchronizer(PostgreSQL _pg, PostgreSQLMigrationContext _ctx)

Method Summary

Modifier and Type	Method and Description
void	determineHighestIndexNumber() Fetches all existing tables and coordinates the underlying PostgreSQLIndex to adjust its internal tracking of the highest used custom index number.
void	dropAllIndexesForTable(java.lang.String tableName) Drops all existing indices for a specific table.
void	generateIndexes(SchemaI schema, boolean nonDestructive) Coordinates the generation, verification, and cleanup of indices across all persistent entities in the schema.
void	generateIndexesPost(boolean nonDestructive) Executes pending index additions and removals sequentially.
java.util.Map<java.lang.String,PostgreSQLIndex>	generatePostgreSQLIndexesForEntity(CBOEntity entity) Generates the expected PostgreSQL indices for an entity based on its attributes and cross-checks them against the currently existing indices in the database.
java.util.Map<java.lang.String,PostgreSQLIndex>	getIndexesForTableJDBC(java.lang.String tableName) Retrieves all existing indices for a given table from the pre-loaded database catalog map (cached via initAllExistingIndexesMapForAllEntities).
java.lang.String	getKey(CBOAttribute att) Constructs a unique string key for an index to map definitions to physical tables.
void	markSuperfluousIndexesForRemoval(java.util.Collection indexes) Marks indices that exist in the database but are no longer defined in the schema for deletion (unless strict mode is deactivated).

Methods inherited from class java.lang.Object

equals, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

PostgreSQLIndexSynchronizer

public PostgreSQLIndexSynchronizer(PostgreSQL _pg,
                                   PostgreSQLMigrationContext _ctx)

Method Detail

determineHighestIndexNumber

public void determineHighestIndexNumber()

Fetches all existing tables and coordinates the underlying PostgreSQLIndex to adjust its internal tracking of the highest used custom index number.

generatePostgreSQLIndexesForEntity

public java.util.Map<java.lang.String,PostgreSQLIndex> generatePostgreSQLIndexesForEntity(CBOEntity entity)

Generates the expected PostgreSQL indices for an entity based on its attributes and cross-checks them against the currently existing indices in the database.

Uses a 3-pass system (Determine definitions, Flag deletions, Flag creations) to avoid duplicate processing.

Parameters:

entity - The entity whose indices should be verified/generated.

Returns:

A map containing the existing indexes in the database for the given table.

getKey

public java.lang.String getKey(CBOAttribute att)

Constructs a unique string key for an index to map definitions to physical tables. Must be kept strictly in sync with PostgreSQLIndex#getKey.

Parameters:

att - The attribute representing the indexed column.

Returns:

The canonical string key for the index.

markSuperfluousIndexesForRemoval

public void markSuperfluousIndexesForRemoval(java.util.Collection indexes)

Marks indices that exist in the database but are no longer defined in the schema for deletion (unless strict mode is deactivated).

Parameters:

indexes - The collection of obsolete indexes discovered during analysis.

generateIndexes

public void generateIndexes(SchemaI schema,
                            boolean nonDestructive)

Coordinates the generation, verification, and cleanup of indices across all persistent entities in the schema. Populates the "addidx" and "delidx" queues.

Parameters:

schema - The schema containing the definitions.

nonDestructive - If true, destructive actions (dropping indexes) are bypassed.

generateIndexesPost

public void generateIndexesPost(boolean nonDestructive)

Executes pending index additions and removals sequentially.

CRITICAL CONCURRENCY WARNING:
job.setSequential(1) MUST remain set to 1! The getConn() method returns the defaultConnection, which is NOT thread-safe for concurrent DDL changes. If parallel index updates are ever enabled, each thread MUST obtain an isolated, fresh connection from the pool.

Parameters:

nonDestructive - If true, no index modifications are executed.

dropAllIndexesForTable

public void dropAllIndexesForTable(java.lang.String tableName)

Drops all existing indices for a specific table. Primarily used during a forced database-wide re-indexing pass.

Parameters:

tableName - The table whose indexes should be purged.

getIndexesForTableJDBC

public java.util.Map<java.lang.String,PostgreSQLIndex> getIndexesForTableJDBC(java.lang.String tableName)

Retrieves all existing indices for a given table from the pre-loaded database catalog map (cached via initAllExistingIndexesMapForAllEntities).

Parameters:

tableName - The table name.

Returns:

A map of canonical index keys to their PostgreSQLIndex objects.