Hi,

On 11.01.2010 21:53, Ray Davis wrote:
> It's been a while, so this might be out-of-date, or I might have
> misunderstood the intent. Here's what seemed to be happening:
> 
> Assuming that SlingHttpServletRequestWrapper was intended to be used
> like HttpServletRequestWrapper, I subclassed it to override some
> relevant "get" methods. But before those overrides could come into play,
> SlingMainServlet triggered a call to RequestData's "unwrap" method,
> which discarded all layers of wrapper so as to go straight through to an
> instance of SlingHttpServletRequestImpl.

I think the current state of the SlingRequestDispatcher in the Engine
bundle does not do that and correctly uses the provided request.

> 
> That seemed to block some of the intended functionality of
> SlingHttpServletRequestWrapper. But, as I say, I didn't pursue the
> question.

It may well be that some time ago, this was wrong.

Regards
Felix

> 
> Best,
> Ray
> 
> Felix Meschberger wrote:
>> Hi,
>>
>> On 11.01.2010 18:07, Ray Davis wrote:
>>> FWIW, as a Sling newcomer I got tripped up by the same code in Sling's
>>> RequestData class when I tried to mix in some request parameters in the
>>> usual way. This  seemed to put a sizable dent in the usefulness of
>>> SlingHttpServletRequestWrapper, but I didn't have time to pursue the
>>> matter then.
>>
>> What is the exact use case you didn't succeed in implementing ?
>>
>> This sounds very much like a bug.
>> Regards
>> Felix
> 

Reply via email to