[
https://issues.apache.org/jira/browse/LUCENE-1614?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12711303#action_12711303
]
Shai Erera commented on LUCENE-1614:
------------------------------------
Maybe there's a way out of this. In 2.9 we're already changing the contract of
DISI with the new nextDoc and advance. As opposed to before (boolean), we're
now stating what these methods should return when there are no more documents.
So what if we do this:
# In 2.9 we document that these methods should return MAX_VAL when there are no
more documents.
# Document doc() that in 3.0 it should return MAX_VAL when there are no more
docs. Also put that in CHANGES.
# I change the patch to return MAX_VAL for all current DISIs/Scorers, but don't
take advantage of that yet in BS and ConjunctionScorer.
# I open another issue for 3.0 which will take advantage of that, and ensure
all current DISIs/Scorers' doc() return MAX_VAL when there are no more docs?
I think that should work. We'll need to delay with that optimization beyond
2.9, but I don't think there's a nice and clean way around it otherwise. If 3.0
should quickly follow 2.9, we won't wait for long :).
> Add next() and skipTo() variants to DocIdSetIterator that return the current
> doc, instead of boolean
> ----------------------------------------------------------------------------------------------------
>
> Key: LUCENE-1614
> URL: https://issues.apache.org/jira/browse/LUCENE-1614
> Project: Lucene - Java
> Issue Type: Improvement
> Components: Search
> Reporter: Shai Erera
> Fix For: 2.9
>
> Attachments: LUCENE-1614.patch
>
>
> See
> http://www.nabble.com/Another-possible-optimization---now-in-DocIdSetIterator-p23223319.html
> for the full discussion. The basic idea is to add variants to those two
> methods that return the current doc they are at, to save successive calls to
> doc(). If there are no more docs, return -1. A summary of what was discussed
> so far:
> # Deprecate those two methods.
> # Add nextDoc() and skipToDoc(int) that return doc, with default impl in DISI
> (calls next() and skipTo() respectively, and will be changed to abstract in
> 3.0).
> #* I actually would like to propose an alternative to the names: advance()
> and advance(int) - the first advances by one, the second advances to target.
> # Wherever these are used, do something like '(doc = advance()) >= 0' instead
> of comparing to -1 for improved performance.
> I will post a patch shortly
--
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-unsubscr...@lucene.apache.org
For additional commands, e-mail: java-dev-h...@lucene.apache.org
