[
https://issues.apache.org/jira/browse/OAK-2943?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14593414#comment-14593414
]
Thomas Mueller commented on OAK-2943:
-------------------------------------
Some remarks:
It would be great to also have a unit test case in oak-core, in addition to, or
instead of the test in oak-lucene.
With the patch, there is only one entry in the result if both subqueries have
the same selectors. OK, that's a possible solution. Another solution would be
to have separate entries (with slightly different key names, for example
selector + "/1" or "/2". One could argue that separate entries can be made by
running each query separately. To me, it does not really matter all that much.
Some small formatting changes:
Add a space after the comma for:
Map<String,Long> getSelectorScanCount();
so it becomes
Map<String, Long> getSelectorScanCount();
This:
static abstract class MeasuringIterator extends
AbstractIterator<ResultRowImpl> {
should be:
abstract static ...
No need for another review, you can commit once you think it's OK.
> Support measure for union queries
> ---------------------------------
>
> Key: OAK-2943
> URL: https://issues.apache.org/jira/browse/OAK-2943
> Project: Jackrabbit Oak
> Issue Type: Sub-task
> Components: query
> Reporter: Amit Jain
> Assignee: Amit Jain
> Fix For: 1.3.1
>
> Attachments: OAK-2943.patch
>
>
> Currently, the {{measure}} for union queries does not take into consideration
> the optimizations done by the query engine and returns the scan count for
> each individual query.
> It would be better if the optimizations and the actual iterations on the
> underlying left/right queries are taken into account.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
