Package org.apache.lucene.sandbox.search
Class TermAutomatonQuery
java.lang.Object
org.apache.lucene.search.Query
org.apache.lucene.sandbox.search.TermAutomatonQuery
- All Implemented Interfaces:
Accountable
A proximity query that lets you express an automaton, whose transitions are terms, to match
documents. This is a generalization of other proximity queries like
PhraseQuery, MultiPhraseQuery and SpanNearQuery. It is likely slow, since it visits any document
having any of the terms (i.e. it acts like a disjunction, not a conjunction like PhraseQuery), and then it must merge-sort all positions within each document to test whether/how
many times the automaton matches.
After creating the query, use createState(), setAccept(int, boolean), addTransition(int, int, java.lang.String)
and addAnyTransition(int, int) to build up the automaton. Once you are done, call finish()
and then execute the query.
This code is very new and likely has exciting bugs!
- WARNING: This API is experimental and might change in incompatible ways in the next release.
-
Field Summary
Fields inherited from interface org.apache.lucene.util.Accountable
NULL_ACCOUNTABLE -
Constructor Summary
Constructors -
Method Summary
Modifier and TypeMethodDescriptionvoidaddAnyTransition(int source, int dest) Adds a transition matching any term.voidaddTransition(int source, int dest, String term) Adds a transition to the automaton.voidaddTransition(int source, int dest, BytesRef term) Adds a transition to the automaton.intReturns a new state; state 0 is always the initial state.createWeight(IndexSearcher searcher, ScoreMode scoreMode, float boost) booleanReturns true iffois equal to this.voidfinish()Call this once you are done adding states/transitions.voidfinish(int determinizeWorkLimit) Call this once you are done adding states/transitions.inthashCode()longrewrite(IndexSearcher indexSearcher) voidsetAccept(int state, boolean accept) Marks the specified state as accept or not.toDot()Returns the dot (graphviz) representation of this automaton.voidvisit(QueryVisitor visitor) Methods inherited from class org.apache.lucene.search.Query
classHash, sameClassAs, toStringMethods inherited from class java.lang.Object
clone, finalize, getClass, notify, notifyAll, wait, wait, waitMethods inherited from interface org.apache.lucene.util.Accountable
getChildResources
-
Constructor Details
-
TermAutomatonQuery
-
-
Method Details
-
createState
public int createState()Returns a new state; state 0 is always the initial state. -
setAccept
public void setAccept(int state, boolean accept) Marks the specified state as accept or not. -
addTransition
Adds a transition to the automaton. -
addTransition
Adds a transition to the automaton. -
addAnyTransition
public void addAnyTransition(int source, int dest) Adds a transition matching any term. -
finish
public void finish()Call this once you are done adding states/transitions. -
finish
public void finish(int determinizeWorkLimit) Call this once you are done adding states/transitions.- Parameters:
determinizeWorkLimit- Maximum effort to spend determinizing the automaton. Higher numbers allow this operation to consume more memory but allow more complex automatons. UseOperations.DEFAULT_DETERMINIZE_WORK_LIMITas a decent default if you don't otherwise know what to specify.
-
createWeight
public Weight createWeight(IndexSearcher searcher, ScoreMode scoreMode, float boost) throws IOException - Overrides:
createWeightin classQuery- Throws:
IOException
-
toString
-
equals
Returns true iffois equal to this. -
hashCode
public int hashCode() -
ramBytesUsed
public long ramBytesUsed()- Specified by:
ramBytesUsedin interfaceAccountable
-
toDot
Returns the dot (graphviz) representation of this automaton. This is extremely useful for visualizing the automaton. -
rewrite
- Overrides:
rewritein classQuery- Throws:
IOException
-
visit
-