[
https://issues.apache.org/jira/browse/LUCENE-5078?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13693162#comment-13693162
]
Shai Erera commented on LUCENE-5078:
------------------------------------
That's a good point! What about if we make the whole change smaller by keeping
the byte-ness and cache in TfIdfSimilarity as a default impl (as it is now),
and only change the signature of encode/decode to long. Then, if all you care
about is encode the full norm, you can extend DefaultSimilarity and not
implement tf, idf and friends?
I think jdocs is enough. It's quite an expert API, and whoever overrides encode
should know better not to override decode ...
If however we keep them on Default, I think they should be final there.
> Allow TfIdfSimilarity implementations to encode norm values into more than a
> single byte
> ----------------------------------------------------------------------------------------
>
> Key: LUCENE-5078
> URL: https://issues.apache.org/jira/browse/LUCENE-5078
> Project: Lucene - Core
> Issue Type: Improvement
> Components: core/index
> Reporter: Shai Erera
> Assignee: Shai Erera
> Attachments: LUCENE-5078.patch
>
>
> Continuation from here: http://lucene.markmail.org/message/jtwit3pwu5oiqr2h.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
