Timothy Potter created SOLR-16199:
-------------------------------------
Summary: Solr SQL using {!complexphrase} for wildcard filters
doesn't always work as expected
Key: SOLR-16199
URL: https://issues.apache.org/jira/browse/SOLR-16199
Project: Solr
Issue Type: Bug
Security Level: Public (Default Security Level. Issues are Public)
Components: Parallel SQL
Reporter: Timothy Potter
Assignee: Timothy Potter
If a field is text analyzed, then doing {{{!complexphrase}name:\"Test*\"}}
causes:
{code}
\"msg\":\"maxClauseCount is set to 1024\",\n
\"trace\":\"org.apache.lucene.search.BooleanQuery$TooManyClauses:
maxClauseCount is set to 1024\\n\\tat
{code}
Working around the issue with the parens hack like {{name LIKE '(Test*)'}}
makes the query work as expected. Perhaps we can just drop using the
{{complexphrase}} parser if the predicate has a trailing wildcard only (leading
too), that'll probably cover the bulk of the problems and use parens instead of
double quotes if it has multiple wild card in it?
Need to build some better test cases around wildcard matching for string and
text analyzed fields.
--
This message was sent by Atlassian Jira
(v8.20.7#820007)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
