[LUCENE-5344] Flexible StandardQueryParser behaves differently than ClassicQueryParser

• Type: Bug
• Status: Closed
• Priority: Major
• Resolution: Fixed
• Affects Version/s: 4.5
• Fix Version/s: 4.6.1, 4.7, 6.0
• Component/s: core/queryparser
• Assignee: Adriano Crestani
• Reporter: Krishna Keldec
• Created: 20/Nov/13 12:01
• Updated: 09/May/16 18:38
• Resolved: 15/Jan/14 06:39

Description

AnalyzerQueryNodeProcessor creates a BooleanQueryNode instead of a MultiPhraseQueryNode for some circumstances.

Classic query parser output: +content:a +content:320 (correct)

QueryParser classicQueryParser;
classicQueryParser = new QueryParser(Version.LUCENE_45, "content", anaylzer);
classicQueryParser.setDefaultOperator(Operator.AND);
classicQueryParser.parse("a320"));

Flexible query parser output: content:a content:320 (wrong)

StandardQueryParser flexibleQueryParser;
flexibleQueryParser = new StandardQueryParser(anaylzer);
flexibleQueryParser.setDefaultOperator(Operator.AND);
flexibleQueryParser.parse("a320", "content"));

The used analyzer:

Analyzer anaylzer = return new Analyzer() {
  protected TokenStreamComponents createComponents(String field, Reader in) {
  Tokenizer   src = new WhitespaceTokenizer(Version.LUCENE_45, in);
  TokenStream tok = new WordDelimiterFilter(src,
     WordDelimiterFilter.SPLIT_ON_NUMERICS |
     WordDelimiterFilter.GENERATE_WORD_PARTS |
     WordDelimiterFilter.GENERATE_NUMBER_PARTS,
     CharArraySet.EMPTY_SET); 
  return new TokenStreamComponents(src, tok);
};