[
https://issues.apache.org/jira/browse/OLINGO-735?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14636488#comment-14636488
]
Michael Bolz commented on OLINGO-735:
-------------------------------------
Hi [~gmurray],
After a more detailed look I have to say that the replacement of
{{'*Builder.reflection*'}} is more time consuming.
Because e.g. in the case of the {{ClientItem}} it can not just replaced with an
implementation at this class, it has also to be replaced in each sub-class.
For the classes were an implementation was possible (within the core library) I
have replaced it.
See [this
commit|https://git1-us-west.apache.org/repos/asf?p=olingo-odata4.git;a=commit;h=1ebbbc3698de644c097efe44472130ef248baa92].
Best regards,
Michael
> reflectionEquals in ClientItem and FullQualifiedName make for unacceptable
> parsing performance.
> -----------------------------------------------------------------------------------------------
>
> Key: OLINGO-735
> URL: https://issues.apache.org/jira/browse/OLINGO-735
> Project: Olingo
> Issue Type: Bug
> Components: odata4-client, odata4-commons
> Affects Versions: (Java) V4 4.0.0-beta-03
> Reporter: Graham Murray
> Assignee: Michael Bolz
> Fix For: (Java) V4 4.0.0
>
>
> Reviewing a CPU profiling result of a simple OData parse points out any use
> of reflectionEquals as demolishing the parse performance. It doesn't even
> seem like reflectionEquals is needed for ClientItem or FullQualifiedName,
> shouldn't they just be implemented properly?
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
