[
https://issues.apache.org/jira/browse/SLING-1596?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12888390#action_12888390
]
Bertrand Delacretaz commented on SLING-1596:
--------------------------------------------
> I think we should not do this...
How do you suggest fixing this then?
Currently Sling cannot process a request that is not a
SlingHttpServletRequestImpl, even if its is a SlingHttpServletRequest. That's
much uglier than exposing an implementation object as an attribute, IMHO.
> Reduce coupling between RequestData and SlingHttpServletRequestImpl
> -------------------------------------------------------------------
>
> Key: SLING-1596
> URL: https://issues.apache.org/jira/browse/SLING-1596
> Project: Sling
> Issue Type: Improvement
> Components: Engine
> Affects Versions: Engine 2.0.6
> Reporter: Bertrand Delacretaz
> Assignee: Bertrand Delacretaz
> Priority: Minor
> Attachments: SLING-1596.patch
>
>
> As discussed in http://markmail.org/thread/ldayz27ehldyvzr4 the tight
> coupling between RequestData and SlingHttpServletRequestImpl makes it
> impossible to use request classes that just implement SlingHttpServletRequest.
> I need this for example for SLING-550, where servlets run outside of the
> container's request/response cycle.
> I'll attach a patch that reduces coupling by grabbing the RequestData from a
> request attribute instead of relying on the SlingHttpServletRequestImpl class
> to provide it. All tests including integration pass with this patch.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
