On Wed, Jan 6, 2010 at 6:30 AM, Claes Jakobsson <claes_at_versed.se> wrote:
> On 6 jan 2010, at 15.00, Julien Chaffraix wrote:
>> On Sun, Jan 3, 2010 at 2:05 PM, Daniel Stenberg <daniel_at_haxx.se> wrote:
>>>
>>> Yes, I agree. 'path' in this context is only the path fragment of the
>>> URI/URL anyway so doing a strchr() for '#' should be a better idea I think.
>>>
>>
>> I also completely agree with the proposal. It would make us degrade
>> gracefully on such non valid URI.
>>
>> An updated patch is attached. Do you think we should add a test for such URI?
>
> Yes, such a test would be valuable as we then can verify that we remove from the left-most #.
>
> Care to fix it? and I'll apply the patch.

Sure, the patch with the new test is attached. I have also added a
comment in each new test to explain what they verify and updated the
comment in url.c to mirror your comments on this list.

Thanks a lot.

Regards,
Julien

-------------------------------------------------------------------
List admin: http://cool.haxx.se/list/listinfo/curl-library
Etiquette: http://curl.haxx.se/mail/etiquette.html