Package com.illumon.iris.db.v2.sort.findruns

Class DoubleFindRunsKernel

java.lang.Object

com.illumon.iris.db.v2.sort.findruns.DoubleFindRunsKernel

public class DoubleFindRunsKernel
extends Object

Constructor Summary

Constructors

Constructor	Description
DoubleFindRunsKernel()

Method Summary

Modifier and Type	Method	Description
static FindRunsKernel	createContext()
static void	findRuns(DoubleChunk sortedValues, IntChunk<Attributes.ChunkPositions> offsetsIn, IntChunk<Attributes.ChunkLengths> lengthsIn, WritableIntChunk<Attributes.ChunkPositions> offsetsOut, WritableIntChunk<Attributes.ChunkLengths> lengthsOut)	Find runs of two or more identical values in a sorted chunk.
static void	findRunsSingles(DoubleChunk sortedValues, WritableIntChunk<Attributes.ChunkPositions> offsetsOut, WritableIntChunk<Attributes.ChunkLengths> lengthsOut)	Find runs of identical values in a sorted chunk.

Methods inherited from class java.lang.Object

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

Constructor Details

DoubleFindRunsKernel

public DoubleFindRunsKernel()

Method Details

findRunsSingles

public static void findRunsSingles(DoubleChunk sortedValues, WritableIntChunk<Attributes.ChunkPositions> offsetsOut, WritableIntChunk<Attributes.ChunkLengths> lengthsOut)

Find runs of identical values in a sorted chunk. If a single value exists, it is included as run of length 1.

Parameters:

sortedValues - a chunk of sorted values

offsetsOut - an output chunk, with offsets of starting locations that a run occurred

lengthsOut - an output chunk, parallel to offsetsOut, with the lengths of found runs

findRuns

public static void findRuns(DoubleChunk sortedValues, IntChunk<Attributes.ChunkPositions> offsetsIn, IntChunk<Attributes.ChunkLengths> lengthsIn, WritableIntChunk<Attributes.ChunkPositions> offsetsOut, WritableIntChunk<Attributes.ChunkLengths> lengthsOut)

Find runs of two or more identical values in a sorted chunk. This is used as part of an overall sort, after the timsort (or other sorting) kernel to identify the runs that must be sorted according to secondary keys.

Parameters:

sortedValues - a chunk of sorted values

offsetsIn - the offsets within the chunk to check for runs

lengthsIn - the lengths parallel to offsetsIn for run checking

offsetsOut - an output chunk, with offsets of starting locations that a run occurred

lengthsOut - an output chunk, parallel to offsetsOut, with the lengths of found runs Note, that lengthsIn must contain values greater than 1, and lengthsOut additionally only contain values greater than one

createContext

public static FindRunsKernel createContext()