Package org.apache.druid.sql.calcite.schema

Interface DruidSchemaManager

All Known Implementing Classes:

NoopDruidSchemaManager

public interface DruidSchemaManager

This interface provides a map of datasource names to DruidTable objects, used by the DruidSchema class as the SQL planner's view of Druid datasource schemas. If a mapping is found for the datasource name in CatalogResolver, then it is preferred during resolution time, to the entry found in this manager. See DruidSchema.getTable(String). If a non-default implementation is provided, the segment metadata polling-based view of the Druid tables will not be built in DruidSchema.

Method Summary

Modifier and Type	Method	Description
default DruidTable	getTable(String name, BrokerSegmentMetadataCache segmentMetadataCache)
default Set<String>	getTableNames(BrokerSegmentMetadataCache segmentMetadataCache)
Map<String,DruidTable>	getTables()	Deprecated.

Method Details

getTables

@Deprecated
Map<String,DruidTable> getTables()

Deprecated.

Return all tables known to this schema manager. Deprecated because getting all tables is never actually needed in the current code. Calcite asks for the information for a single table (via getTable(String, BrokerSegmentMetadataCache), or the list of all table names (via getTableNames(BrokerSegmentMetadataCache). This method was originally used to allow Calcite's AbstractSchema class to do the lookup. The current code implements the operations directly. For backward compatibility, the default methods emulate the old behavior. Newer implementations should omit this method and implement the other two.

getTable

default DruidTable getTable(String name,
 BrokerSegmentMetadataCache segmentMetadataCache)

getTableNames

default Set<String> getTableNames(BrokerSegmentMetadataCache segmentMetadataCache)