On 2010 July 01, Thursday 14:57:53 you wrote:
Hi,
> > I have attached a patch that addresses the issues I described. I
> > introduced a
>
> thanks a lot for the patch. I have applied it (slightly modified) to
> HEAD. btw, you can add some test cases to tests/test_headers if you like.
>
> > boolean parameter to getHeader() which switches between the old and the
> > "correct" behaviour (defaulted to the correct) and patched all places to
> > use the old one so I hope nothing broke. Also had to add a new parameter
> > to
>
> I am wondering: in most cases, the new ("correct") behaviour would be
> better, right?
Sure, but I had no energy to understand every use in every application, so I
played safe.
> > findHeader. Modified removeHeader() also to remove all instances of the
> > header specified.
> >
> > Regarding the original problem in 0000024, I had trouble finding out the
> > exact format of the parameter from the RFC. Is it legal to the key be
> > quoted? Can it contain escaped characters? Which ones? Same with the
> > value. Is it legal to have whitespace around the "=" ? Etc. Any
> > pointers?
>
> there is actually no RFC for that - the P-App-Param header is SEMS'
> creation. So we are free to put the syntax the way it fits best. In
But the get_header_keyvalue() function is used for lots of other headers too,
not only P-App-Param.
br
Szo
> fact I think that
> - it would make sense to support quoted strings (with " as \")
> - it would make sense to support whitespace around "="
_______________________________________________
Semsdev mailing list
[email protected]
http://lists.iptel.org/mailman/listinfo/semsdev
