[jira] Updated: (LUCENE-1410) PFOR implementation

["Michael McCandless (JIRA)" <ji...@apache.org>]

     [ https://issues.apache.org/jira/browse/LUCENE-1410?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Michael McCandless updated LUCENE-1410:
---------------------------------------

    Attachment: LUCENE-1410-codecs.tar.bz2

Attaching sep, intblock and pfordelta codecs, spun out of the last patch on LUCENE-1458.

Once LUCENE-1458 is in, we should finish the pfordelta codec to make it a real choice.

I actually think some combination of pulsing, standard, pfordelta and simple bit packing (in order by increasing term's docFreq), within a single codec, may be best.

Ie, rare terms (only in a doc or two) could be inlined into the the terms dict.  Slightly more common terms can use the more CPU intensive standard codec.  Common terms can use cpu-friendly-yet-still-decent-compression pfordelta.  Obsenely common terms can use bit packing for the fastest decode.

> PFOR implementation
> -------------------
>
>                 Key: LUCENE-1410
>                 URL: https://issues.apache.org/jira/browse/LUCENE-1410
>             Project: Lucene - Java
>          Issue Type: New Feature
>          Components: Other
>            Reporter: Paul Elschot
>            Priority: Minor
>         Attachments: autogen.tgz, LUCENE-1410-codecs.tar.bz2, LUCENE-1410b.patch, LUCENE-1410c.patch, LUCENE-1410d.patch, LUCENE-1410e.patch, TermQueryTests.tgz, TestPFor2.java, TestPFor2.java, TestPFor2.java
>
>   Original Estimate: 21840h
>  Remaining Estimate: 21840h
>
> Implementation of Patched Frame of Reference.

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.


---------------------------------------------------------------------
To unsubscribe, e-mail: java-dev-unsubscribe@lucene.apache.org
For additional commands, e-mail: java-dev-help@lucene.apache.org