[ 
https://issues.apache.org/jira/browse/LUCENE-3280?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13060619#comment-13060619
 ] 

Yonik Seeley commented on LUCENE-3280:
--------------------------------------

I think FastBitSet should still have 
{code}
  /** Expert: returns the long[] storing the bits */
  public long[] getBits() { return bits; }
{code}

The whole reason I had to create OpenBitSet in the first place was that you 
couldn't do anything custom fast (on a word-for-word basis) because the bits 
were locked away from you.

> Add new bit set impl for caching filters
> ----------------------------------------
>
>                 Key: LUCENE-3280
>                 URL: https://issues.apache.org/jira/browse/LUCENE-3280
>             Project: Lucene - Java
>          Issue Type: Improvement
>            Reporter: Michael McCandless
>            Assignee: Michael McCandless
>             Fix For: 3.4, 4.0
>
>         Attachments: LUCENE-3280.patch
>
>
> I think OpenBitSet is trying to satisfy too many audiences, and it's
> confusing/error-proned as a result.  It has int/long variants of many
> methods.  Some methods require in-bound access, others don't; of those
> others, some methods auto-grow the bits, some don't.  OpenBitSet
> doesn't always know its numBits.
> I'd like to factor out a more "focused" bit set impl whose primary
> target usage is a cached Lucene Filter, ie a bit set indexed by docID
> (int, not long) whose size is known and fixed up front (backed by
> final long[]) and is always accessed in-bounds.

--
This message is automatically generated by JIRA.
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]

Reply via email to