[
https://issues.apache.org/jira/browse/LUCENE-5629?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13979875#comment-13979875
]
Ahmet Arslan commented on LUCENE-5629:
--------------------------------------
bq. The same problem exists for the Analyzer that is used.
Can't we use different analyzers for indexing and searching? e.g.
WordDelimiterFilter, SynonymFilter, NGramFilter, etc.
> Comparing the Version of Lucene , the Analyzer and the similarity function
> that are being used for indexing and searching.
> --------------------------------------------------------------------------------------------------------------------------
>
> Key: LUCENE-5629
> URL: https://issues.apache.org/jira/browse/LUCENE-5629
> Project: Lucene - Core
> Issue Type: New Feature
> Components: core/index, core/queryparser, core/search
> Environment: Operating system : Windows 8.1
> Software platform : Eclipse Kepler 4.3.2
> Reporter: Isabel Mendonca
> Priority: Minor
> Labels: features, patch
> Fix For: 4.8, 4.9, 5.0
>
> Original Estimate: 672h
> Remaining Estimate: 672h
>
> We have observed that Lucene does not check if the same Similarity function
> is used during indexing and searching. The same problem exists for the
> Analyzer that is used. This may lead to poor or misleading results.
> So we decided to create an xml file during indexing that will store
> information such as the Analyzer and the Similarity function that were used
> as well as the version of Lucene that was used. This xml file will always be
> available to the users.
> At search time , we will retrieve this information using SAX parsing and
> check if the utils used for searching , match those used for indexing. If not
> , a warning message will be displayed to the user.
--
This message was sent by Atlassian JIRA
(v6.2#6252)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
