Package com.atlassian.jira.issue.statistics.util

Class AbstractOneDimensionalHitCollector

java.lang.Object

org.apache.lucene.search.SimpleCollector

com.atlassian.jira.issue.statistics.util.AbstractOneDimensionalHitCollector

All Implemented Interfaces:

org.apache.lucene.search.Collector, org.apache.lucene.search.LeafCollector

Direct Known Subclasses:

OneDimensionalDocIssueHitCollector, OneDimensionalTermHitCollector

@Internal
@NotThreadSafe
public abstract class AbstractOneDimensionalHitCollector
extends org.apache.lucene.search.SimpleCollector

Used to traverse a query and collect if the specified one dimension is relevant, this will keep track of the irrelevant count.

This class implements Collector which is a callback mechanism for use by IndexSearcher, but it is widely abused and called directly from within JIRA.

This class assumes the collected values will be sorted. This is not possible in the current API (SearchProvider). Users of this class should - Retrieve the sorted document ids from Lucene - Use doSetNextReader(LeafReaderContext) or doSetNextIndexReader(IndexReaderContext) to set context - Feed sorted ids into collect(int) method This class is not thread-safe.

Since:

v4.0

Constructor Summary

Constructors

Constructor	Description
AbstractOneDimensionalHitCollector(String documentFieldId, FieldVisibilityManager fieldVisibilityManager, ReaderCache readerCache, FieldManager fieldManager)

Method Summary

Modifier and Type	Method	Description
void	collect(int i)	Collect hits.
protected abstract void	collectIrrelevant(int docId)	Implement this if you would like to do something when the hit collector has encountered a docId that contains an irrelevant data match
protected abstract void	collectWithTerms(int docId, JiraBytesRef[] terms)	Implement this if you would like to do something with the hit we have found.
void	doSetNextIndexReader(org.apache.lucene.index.IndexReaderContext context)
void	doSetNextReader(org.apache.lucene.index.LeafReaderContext context)
long	getHitCount()	Returns the number of times the collect(int) method was called.
long	getIrrelevantCount()
boolean	needsScores()

Methods inherited from class org.apache.lucene.search.SimpleCollector

getLeafCollector, setScorer

Methods inherited from class java.lang.Object

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

Constructor Details

AbstractOneDimensionalHitCollector

public AbstractOneDimensionalHitCollector(String documentFieldId,
 FieldVisibilityManager fieldVisibilityManager,
 ReaderCache readerCache,
 FieldManager fieldManager)

Since:

v6.4

Method Details

doSetNextReader

public void doSetNextReader(org.apache.lucene.index.LeafReaderContext context)

Overrides:

doSetNextReader in class org.apache.lucene.search.SimpleCollector

doSetNextIndexReader

public void doSetNextIndexReader(org.apache.lucene.index.IndexReaderContext context)

collect

public void collect(int i)

Collect hits. When accessing our value caches we do not need to increment i by the docBase as the value caches are built against the current (segment) reader. When calling the abstract collectXXX() methods these expect the index of the whole reader, so we need to use (i + docBase).

Specified by:

collect in interface org.apache.lucene.search.LeafCollector

Specified by:

collect in class org.apache.lucene.search.SimpleCollector

Parameters:

i - Index to the current reader

getIrrelevantCount

public long getIrrelevantCount()

getHitCount

public long getHitCount()

Returns the number of times the collect(int) method was called. This should return the number of unique issues that was matched during a search.

Returns:

number of times the collect(int) method was called.

collectIrrelevant

protected abstract void collectIrrelevant(int docId)

Implement this if you would like to do something when the hit collector has encountered a docId that contains an irrelevant data match

Parameters:

docId - the match we have found. This is an index into the overall directory reader.

collectWithTerms

protected abstract void collectWithTerms(int docId,
 JiraBytesRef[] terms)

Implement this if you would like to do something with the hit we have found.

Parameters:

docId - the doc id of the hit. This is an index into the overall directory reader.

terms - the terms for the fieldId for this document, pre-calculated so you may not need to call getDocument

needsScores

public boolean needsScores()