[LUCENE-1140] NPE in StopFilter caused by StandardAnalyzer(boolean replaceInvalidAcronym) constructor

• Type: Bug
• Status: Closed
• Priority: Blocker
• Resolution: Fixed
• Affects Version/s: 2.3
• Fix Version/s: 2.3
• Component/s: modules/analysis
• Assignee: Grant Ingersoll
• Reporter: Alexei Dets
• Created: 18/Jan/08 20:36
• Updated: 28/Aug/22 11:45
• Resolved: 18/Jan/08 21:28

Description

I think that I found a problem with the new code (https://issues.apache.org/jira/browse/LUCENE-1068).
Usage of the new constructor StandardAnalyzer(boolean replaceInvalidAcronym) causes NPE in
StopFilter:

java.lang.NullPointerException
at org.apache.lucene.analysis.StopFilter.<init>(StopFilter.java:74)
at org.apache.lucene.analysis.StopFilter.<init>(StopFilter.java:86)
at
org.apache.lucene.analysis.standard.StandardAnalyzer.tokenStream(StandardAnalyzer.java:151)
at
org.apache.lucene.queryParser.QueryParser.getFieldQuery(QueryParser.java:452)
at
org.apache.lucene.queryParser.QueryParser.Term(QueryParser.java:1133)
at
org.apache.lucene.queryParser.QueryParser.Clause(QueryParser.java:1020)
at
org.apache.lucene.queryParser.QueryParser.Query(QueryParser.java:948)
at
org.apache.lucene.queryParser.QueryParser.Clause(QueryParser.java:1024)
at
org.apache.lucene.queryParser.QueryParser.Query(QueryParser.java:948)
at
org.apache.lucene.queryParser.QueryParser.TopLevelQuery(QueryParser.java:937)
at
org.apache.lucene.queryParser.QueryParser.parse(QueryParser.java:147)

The reason is that new constructor forgets to initialize the stopSet field:
public StandardAnalyzer(boolean replaceInvalidAcronym)

{ this.replaceInvalidAcronym = replaceInvalidAcronym; }

I guess this should be changed to something like this:
public StandardAnalyzer(boolean replaceInvalidAcronym)

{ this(STOP_WORDS); this.replaceInvalidAcronym = replaceInvalidAcronym; }

The bug is present in RC3. Fix is one line, it'll be great to have it in 2.3
release.

Issue Links

• relates to: LUCENE-1068 Invalid behavior of StandardTokenizerImpl