Package org.apache.cassandra.index.sai.plan

Class StorageAttachedIndexSearcher.ScoreOrderedResultRetriever

java.lang.Object

org.apache.cassandra.utils.AbstractIterator<UnfilteredRowIterator>

org.apache.cassandra.index.sai.plan.StorageAttachedIndexSearcher.ScoreOrderedResultRetriever

All Implemented Interfaces:

com.google.common.collect.PeekingIterator<UnfilteredRowIterator>, AutoCloseable, Iterator<UnfilteredRowIterator>, BasePartitionIterator<UnfilteredRowIterator>, UnfilteredPartitionIterator, CloseableIterator<UnfilteredRowIterator>

Enclosing class:

StorageAttachedIndexSearcher

public class StorageAttachedIndexSearcher.ScoreOrderedResultRetriever
extends AbstractIterator<UnfilteredRowIterator>

A result retriever that consumes an iterator primary keys sorted by some score, materializes the row for each primary key (currently, each primary key is required to be fully qualified and should only point to one row), apply the filter tree to the row to test that the real row satisfies the WHERE clause, and finally tests that the row is valid for the ORDER BY clause. The class performs some optimizations to avoid materializing rows unnecessarily. See the class for more details.

The resulting UnfilteredRowIterator objects are not guaranteed to be in any particular order. It is the responsibility of the caller to sort the results if necessary.

Method Summary

Modifier and Type	Method	Description
void	close()
UnfilteredRowIterator	computeNext()
TableMetadata	metadata()
UnfilteredRowIterator	readAndValidatePartition(PrimaryKey pk, List<PrimaryKeyWithScore> sourceKeys)	Reads and validates a partition for a given primary key against its sources.

Methods inherited from class org.apache.cassandra.utils.AbstractIterator

endOfData, hasNext, next, peek, remove

Methods inherited from class java.lang.Object

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

Methods inherited from interface java.util.Iterator

forEachRemaining, hasNext

Method Details

computeNext

public UnfilteredRowIterator computeNext()

Specified by:

computeNext in class AbstractIterator<UnfilteredRowIterator>

readAndValidatePartition

public UnfilteredRowIterator readAndValidatePartition(PrimaryKey pk,
 List<PrimaryKeyWithScore> sourceKeys)

Reads and validates a partition for a given primary key against its sources.

Parameters:

pk - The primary key of the partition to read and validate

sourceKeys - A list of PrimaryKeyWithScore objects associated with the primary key. Multiple sort keys can exist for the same primary key when data comes from different sstables or memtables.

Returns:

An UnfilteredRowIterator containing the validated partition data, or null if: - The key has already been processed - The partition does not pass index filters - The partition contains no valid rows - The row data does not match the index metadata for any of the provided primary keys

close

public void close()

Specified by:

close in interface AutoCloseable

Specified by:

close in interface BasePartitionIterator<UnfilteredRowIterator>

Specified by:

close in interface CloseableIterator<UnfilteredRowIterator>

Overrides:

close in class AbstractIterator<UnfilteredRowIterator>

metadata

public TableMetadata metadata()

Specified by:

metadata in interface UnfilteredPartitionIterator