[
https://issues.apache.org/jira/browse/OLINGO-1009?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15476905#comment-15476905
]
Steven Hawkins commented on OLINGO-1009:
----------------------------------------
> this does not handle the case where explicit $levels is defined but cycle is
> detected.
I don't follow you, can you elaborate?
> Minor concern is not thread safe.
I just didn't see any instance where it needed to be nor was it documented as
needing to be. That of course can be changed.
> Also couple unit tests were failing.
Yes, those should fail as the cycle is resolved with an entity reference rather
than the entity itself. Sorry I wan't clear that the patch was just a proposal
of the approach and not fully baked. This is a behavioral change which could
be limited to only the max case, but it makes just as much sense to resolve all
cycles regardless of when they appear.
> I reworked and came up with
No, that is not correct. Run the unit test with my patch to see the expected
output - the cycle should be resolved with an entity reference, not by omitting
the entity altogether.
> Provide serilization support for $levels with $expand
> -----------------------------------------------------
>
> Key: OLINGO-1009
> URL: https://issues.apache.org/jira/browse/OLINGO-1009
> Project: Olingo
> Issue Type: New Feature
> Components: odata4-server
> Affects Versions: (Java) V4 4.0.0-beta-01
> Reporter: Ramesh Reddy
> Assignee: Ramesh Reddy
> Fix For: (Java) V4 4.3.0
>
>
> Currently Olingo parses the query with $expand with $levels construct,
> however if the payload contains the nesting of the entities, it does not
> serialize to the defined $levels
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
