org.sindice.siren.index.codecs.siren10

Class Siren10PostingsWriter

java.lang.Object

org.apache.lucene.codecs.PostingsConsumer

org.apache.lucene.codecs.PostingsWriterBase

org.sindice.siren.index.codecs.siren10.Siren10PostingsWriter

All Implemented Interfaces:

Closeable, AutoCloseable

public class Siren10PostingsWriter
extends org.apache.lucene.codecs.PostingsWriterBase

Writes the document identifiers, node frequencies, node labels, term frequencies, term positions and block skip data.

Field Summary

Fields

Modifier and Type	Field and Description
protected static org.slf4j.Logger	logger

Constructor Summary

Constructors

Constructor and Description
Siren10PostingsWriter(org.apache.lucene.index.SegmentWriteState state, int blockSkipInterval, Siren10BlockStreamFactory factory)
Siren10PostingsWriter(org.apache.lucene.index.SegmentWriteState state, Siren10BlockStreamFactory factory)

Method Summary

Methods

Modifier and Type	Method and Description
void	addPosition(int position, org.apache.lucene.util.BytesRef payload, int startOffset, int endOffset)
void	close()
void	finishDoc()
void	finishTerm(org.apache.lucene.codecs.TermStats stats) Called when we are done adding docs to this term
void	flushTermsBlock(int start, int count)
org.apache.lucene.codecs.TermStats	merge(org.apache.lucene.index.MergeState mergeState, org.apache.lucene.index.DocsEnum postings, org.apache.lucene.util.FixedBitSet visitedDocs) Default merge impl: append documents, nodes and positions, mapping around deletes.
void	setField(org.apache.lucene.index.FieldInfo fieldInfo)
void	start(org.apache.lucene.store.IndexOutput termsOut)
void	startDoc(int docID, int termDocFreq) Adds a new doc in this term.
void	startTerm()

Methods inherited from class java.lang.Object

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

Field Detail

logger

protected static final org.slf4j.Logger logger

Constructor Detail

Siren10PostingsWriter

public Siren10PostingsWriter(org.apache.lucene.index.SegmentWriteState state,
                     Siren10BlockStreamFactory factory)
                      throws IOException

Throws:

IOException

Siren10PostingsWriter

public Siren10PostingsWriter(org.apache.lucene.index.SegmentWriteState state,
                     int blockSkipInterval,
                     Siren10BlockStreamFactory factory)
                      throws IOException

Throws:

IOException

Method Detail

start

public void start(org.apache.lucene.store.IndexOutput termsOut)
           throws IOException

Specified by:

start in class org.apache.lucene.codecs.PostingsWriterBase

Throws:

IOException

startTerm

public void startTerm()
               throws IOException

Specified by:

startTerm in class org.apache.lucene.codecs.PostingsWriterBase

Throws:

IOException

setField

public void setField(org.apache.lucene.index.FieldInfo fieldInfo)

Specified by:

setField in class org.apache.lucene.codecs.PostingsWriterBase

startDoc

public void startDoc(int docID,
            int termDocFreq)
              throws IOException

Adds a new doc in this term. If this returns null then we just skip consuming positions.

termDocFreq parameter is ignored as term frequency in document is not used.

Specified by:

startDoc in class org.apache.lucene.codecs.PostingsConsumer

Throws:

IOException

addPosition

public void addPosition(int position,
               org.apache.lucene.util.BytesRef payload,
               int startOffset,
               int endOffset)
                 throws IOException

Specified by:

addPosition in class org.apache.lucene.codecs.PostingsConsumer

Throws:

IOException

finishDoc

public void finishDoc()

Specified by:

finishDoc in class org.apache.lucene.codecs.PostingsConsumer

finishTerm

public void finishTerm(org.apache.lucene.codecs.TermStats stats)
                throws IOException

Called when we are done adding docs to this term

Specified by:

finishTerm in class org.apache.lucene.codecs.PostingsWriterBase

Throws:

IOException

flushTermsBlock

public void flushTermsBlock(int start,
                   int count)
                     throws IOException

Specified by:

flushTermsBlock in class org.apache.lucene.codecs.PostingsWriterBase

Throws:

IOException

close

public void close()
           throws IOException

Specified by:

close in interface Closeable

Specified by:

close in interface AutoCloseable

Specified by:

close in class org.apache.lucene.codecs.PostingsWriterBase

Throws:

IOException

merge

public org.apache.lucene.codecs.TermStats merge(org.apache.lucene.index.MergeState mergeState,
                                       org.apache.lucene.index.DocsEnum postings,
                                       org.apache.lucene.util.FixedBitSet visitedDocs)
                                         throws IOException

Default merge impl: append documents, nodes and positions, mapping around deletes.

Bypass the Siren10PostingsWriter methods and work directly with the BlockWriters for maximum efficiency.

TODO - Optimisation: If document blocks match the block size, and no document deleted, then it would be possible to copy block directly as byte array, avoiding decoding and encoding.

Overrides:

merge in class org.apache.lucene.codecs.PostingsConsumer

Throws:

IOException