[
https://issues.apache.org/jira/browse/OAK-2118?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14143189#comment-14143189
]
Davide Giannella edited comment on OAK-2118 at 9/22/14 1:26 PM:
----------------------------------------------------------------
[~alex.parvulescu] looks good to me. I remember having seen as well while
digging the code some operations to avoid dups during inserts but I can't
remember right now where exactly.
I'll resolve as 1.1 if don't mind. always in time to re-open in case.
was (Author: edivad):
[~alex.parvulescu] looks good to me. I remember having seen as well while
digging the code some operations to avoid dups during inserts but I can't
remember right now where exactly.
> Aggregation cursor introduces doubles in the result set
> -------------------------------------------------------
>
> Key: OAK-2118
> URL: https://issues.apache.org/jira/browse/OAK-2118
> Project: Jackrabbit Oak
> Issue Type: Bug
> Components: core
> Affects Versions: 1.0, 1.1
> Reporter: Alex Parvulescu
> Assignee: Alex Parvulescu
> Attachments: OAK-2118.patch
>
>
> In some cases (double aggregation) the aggregation cursor in some cases will
> introduce doubles in the result set by replacing some good results.
> This makes the result set look like (for the attached test case):
> {code}
> "/myFolder", "/myFolder", "/myFolder/myFile/jcr:content"
> {code}
> instead of
> {code}
> "/myFolder", "/myFolder/myFile", "/myFolder/myFile/jcr:content"
> {code}
> What actually happens is the cursor will replace the current row value with
> the new one ("/myFolder/myFile" will turn into "/myFolder") in the #next
> call. Not only this replaces the good value with the double, but it also
> skips the de-duplication bits, resulting in an incomplete result set with
> doubles sprinkled in.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
