[
https://issues.apache.org/jira/browse/LUCENE-8288?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16461058#comment-16461058
]
Jim Ferenczi commented on LUCENE-8288:
--------------------------------------
Since it is possible to index suggestions with or without separators
(preservePositionIncrements) the context query adds an optional separator after
the context automaton. This character is optional so the regex "." can match
the context plus the separator label but nothing from the real suggestions.
Completion queries should always match a prefix from the suggestions (hence the
assertion) but it doesn't handle regex that starts with ".". I've attached a
patch to fix the issue that adds a parameter in the ContextQuery constructor to
indicate if suggestions are indexed with position increments or not.This is a
breaking change since it requires to match the value used for indexing but I
don't see how to do it differently if we want to match regex that starts with
any character accurately (e.g.: ".[s|t]").
> ContextQuery "." for RegexCompletionQuery produces an assertion failure
> -----------------------------------------------------------------------
>
> Key: LUCENE-8288
> URL: https://issues.apache.org/jira/browse/LUCENE-8288
> Project: Lucene - Core
> Issue Type: Bug
> Reporter: Julie Tibshirani
> Priority: Major
> Attachments: LUCENE-8288-repro.patch, LUCENE-8288.patch
>
>
> When a RegexCompletionQuery of "." is provided to ContextQuery, the following
> assertion failure occurs:
> {code:java}
> java.lang.AssertionError: input should not end with a context separator
> followed by SEP_LABEL
> at
> org.apache.lucene.search.suggest.document.ContextQuery$ContextCompletionWeight.setInnerWeight(ContextQuery.java:299)
> at
> org.apache.lucene.search.suggest.document.ContextQuery$ContextCompletionWeight.setNextMatch(ContextQuery.java:275)
> at
> org.apache.lucene.search.suggest.document.NRTSuggester.lookup(NRTSuggester.java:221)
> at
> org.apache.lucene.search.suggest.document.CompletionScorer.score(CompletionScorer.java:70)
> at org.apache.lucene.search.BulkScorer.score(BulkScorer.java:39)
> at
> org.apache.lucene.search.suggest.document.SuggestIndexSearcher.suggest(SuggestIndexSearcher.java:78)
> at
> org.apache.lucene.search.suggest.document.SuggestIndexSearcher.suggest(SuggestIndexSearcher.java:58)
> at
> org.apache.lucene.search.suggest.document.TestContextQuery.testDotRegexQuery(TestContextQuery.java:188)
> {code}
> Note that this is a related, but distinct issue from
> https://issues.apache.org/jira/browse/LUCENE-8287, where the
> RegexCompletionQuery is empty.
> The attached patch provides a reproduction of the issue, as the test case
> TestContextQuery#testRegexDotQuery. To reproduce, Java assertions must be
> enabled (as in the default configuration for tests).
> The patch also provides a test case for the normal behavior of an empty
> RegexCompletionQuery, when it is not wrapped in ContextQuery
> (TestRegexCompletionQuery#testRegexDotQuery). In this case, there is no
> error, and all suggestions are returned.
> From a quick look, it seems as though "." doesn't capture any characters past
> CompletionAnalyzer.SEP_LABEL, so the matching prefix in
> ContextCompletionWeight#setInnerWeight is unexpectedly empty.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org
For additional commands, e-mail: dev-h...@lucene.apache.org
