[
https://issues.apache.org/jira/browse/OAK-2754?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14502529#comment-14502529
]
Chetan Mehrotra commented on OAK-2754:
--------------------------------------
bq. Maybe you are referring to the _ seenPaths_ set [0], just wanted to make
sure
Missed that. I was thinking that as of now it would not be possible to have
duplicated paths at all at Lucene query level itself. If we already keep a set
then its very much safe to not use UniquePathCursor. And then I would also
think about removing that set also (but later!)
> Use non unique PathCursor in LucenePropertyIndex
> ------------------------------------------------
>
> Key: OAK-2754
> URL: https://issues.apache.org/jira/browse/OAK-2754
> Project: Jackrabbit Oak
> Issue Type: Improvement
> Components: lucene
> Reporter: Chetan Mehrotra
> Assignee: Chetan Mehrotra
> Priority: Minor
> Fix For: 1.3.0
>
>
> {{LucenePropertyIndex}} currently uses unique PathCursor [1] due to which the
> cursor would maintain an in memory set of visited path. This might grow big
> if result size is big and cursor is traversed completely.
> As with current impl the path would not be duplicated we can avoid using
> unique cursor
> [1]
> https://github.com/apache/jackrabbit-oak/blob/trunk/oak-lucene/src/main/java/org/apache/jackrabbit/oak/plugins/index/lucene/LucenePropertyIndex.java#L1153-1154
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
