[
https://issues.apache.org/jira/browse/LUCENE-3417?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13099000#comment-13099000
]
Robert Muir commented on LUCENE-3417:
-------------------------------------
thanks for the patch: all tests pass here with it, and I think the added tests
are clear.
> DictionaryCompoundWordTokenFilter does not properly add tokens from the end
> compound word.
> ------------------------------------------------------------------------------------------
>
> Key: LUCENE-3417
> URL: https://issues.apache.org/jira/browse/LUCENE-3417
> Project: Lucene - Java
> Issue Type: Bug
> Components: modules/analysis
> Affects Versions: 3.3, 4.0
> Reporter: Njal Karevoll
> Attachments: LUCENE-3417.patch
>
> Original Estimate: 5m
> Remaining Estimate: 5m
>
> Due to an off-by-one error, a subword placed at the end of a compound word
> will not get a token added to the token stream.
> For example (from the unit test in the attached patch):
> Dictionary: {"ab", "cd", "ef"}
> Input: "abcdef"
> Created tokens: {"abcdef", "ab", "cd"}
> Expected tokens: {"abcdef", "ab", "cd", "ef"}
> Additionally, it could produce tokens that were shorter than the
> minSubwordSize due to another off-by-one error. For example (again, from the
> attached patch):
> Dictionary: {"abc", "d", "efg"}
> Minimum subword length: 2
> Input: "abcdefg"
> Created tokens: {"abcdef", "abc", "d", "efg"}
> Expected tokens: {"abcdef", "abc", "efg"}
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org
For additional commands, e-mail: dev-h...@lucene.apache.org
