[
https://issues.apache.org/jira/browse/LUCENE-8192?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16385242#comment-16385242
]
Robert Muir commented on LUCENE-8192:
-------------------------------------
Second, less aggressive patch: it *changes the boolean* from
{{offsetsAreCorrect}} to {{graphOffsetsAreCorrect}} and always enables the
checks consistent with what indexwriter will do (e.g. offsets don't go
backwards).
I ran tests a couple times and nothing failed... probably deserves some
beasting but I think this is a good step? It removes some useless leniency.
> Remove offsetsAreCorrect from BaseTokenStreamTestCase
> -----------------------------------------------------
>
> Key: LUCENE-8192
> URL: https://issues.apache.org/jira/browse/LUCENE-8192
> Project: Lucene - Core
> Issue Type: Bug
> Reporter: Robert Muir
> Priority: Major
> Attachments: LUCENE-8192_prototype.patch, LUCENE-8192_take_two.patch
>
>
> Similar to LUCENE-8191, now that indexwriter checks the offsets, this boolean
> is useless: if offsets are broken it will still fail.
> We should just remove the boolean.
--
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
