ComplexPhraseQueryParser/Enumerating Wildcard Queries

[GoogleCodeExporter]

We are troubleshooting performance issues related to wildcard queries 
running against the ComplexPhraseQueryParser (introduced in Lucene 3.0). 
Some wildcard queries like "quality operations providing quality food 
services job requirements: click here to apply for this job*" require a 
maxClauseCount > 40K to match against the DBSight implementation that 
we're using (and run longer than 2 seconds too).

We'd like to understand what the 40K clauses are as we don't think we have 
nearly that many unique words that have job* as a prefix. 

Can Luke support the ComplexPhraseQueryParser or otherwise show the fully 
enumerated query for wildcard searches? 

Original issue reported on code.google.com by auspici...@gmail.com on 1 Feb 2010 at 8:20

[GoogleCodeExporter]

There is no specific support for this - however, if you suspect that the last 
term
with the widlcard is the source of the problem you could perhaps use just this 
term
as a simple wildcard query, click the "Update" button, and see the Rewritten 
list of
clauses.

Original comment by sig...@gmail.com on 2 Feb 2010 at 11:05

• Added labels: Type-Enhancement
• Removed labels: Type-Defect

[GoogleCodeExporter]

Thanks but the simple WildCard query "job*" is rewritten as 
ConstantScore(TEXT:job*) 
which doesn't reveal the list of clauses.

Original comment by auspici...@gmail.com on 2 Feb 2010 at 12:39

[GoogleCodeExporter]

Original comment by sig...@gmail.com on 31 Mar 2010 at 4:35

• Changed state: Later