Re: [Dclib-devel] Updated server

[Steven V. I. <st...@va...>]

That is maybe better, because than you could read the "raw post" as
sometimes it is needed as well.


Grtz,
Steven

2009/3/18 Davis King <dav...@gm...>

> Hmm.  Well, what if the body variable just always contains the body (so it
> has Content-Length bytes always)?  That is the part that I checked into
> subversion yesterday.  Is that enough to do what you want to be able to
> do?   Seems like it should be since you can look at the body to see any POST
> variables and any other post content.
>
> -Davis
>
>
> On Wed, Mar 18, 2009 at 5:41 AM, Steven Van Ingelgem <
> st...@va...> wrote:
>
>> Not exactly.
>>
>> The body contains the body (without the POST if there is a post).
>> The post contains ONLY the post variables
>> The get contains ONLY the get variables
>> The queries contains both the post and the get variables (with the get
>> variables overwriting the post if they have the exact same name).
>>
>>
>> 2009/3/18 Davis King <dav...@gm...>
>>
>> Cool.  It's definitely good to have access to the body of the POST data.
>>> But why did you add the GET and POST maps?  They seem to just contain the
>>> same thing as the queries map?
>>>
>>> -Davis
>>>
>>>
>>> On Tue, Mar 17, 2009 at 3:46 PM, Steven Van Ingelgem <
>>> st...@va...> wrote:
>>>
>>>> After recompiling ;-)
>>>>
>>>> 2009/3/17 Steven Van Ingelgem <st...@va...>
>>>>
>>>> Hi Davis,
>>>>>
>>>>>
>>>>> I added parse_url into the server, this way I can get GET, POST or both
>>>>> (little like REQUEST in PHP).
>>>>> I also added a parameter "body", which is needed for example in case of
>>>>> webdav @ "PROPFIND".
>>>>>
>>>>>
>>>>> Grtz,
>>>>> Steven
>>>>>
>>>>> 2009/3/15 Davis King <dav...@gm...>
>>>>>
>>>>> Nice.  I just finished checking all this into subversion.  I changed it
>>>>>> around a little.   The biggest change was I put the two structs you made
>>>>>> into the on_request() function's arguments because it occurred to me that
>>>>>> otherwise we needed to add thread synchronization to the http_memory map.
>>>>>> And that would have been a pain.  This is simpler anyway.  I'm not sure what
>>>>>> I was thinking when I suggested adding those other extra helper function and
>>>>>> the map :)
>>>>>>
>>>>>> So it looks like this now:
>>>>>> http://davis.sytes.net/~davis/dlib/server_http_ex.cpp.html<http://davis.sytes.net/%7Edavis/dlib/server_http_ex.cpp.html>
>>>>>> And I updated the specification file too:
>>>>>> http://davis.sytes.net/~davis/dlib/dlib/server/server_http_abstract.h.html<http://davis.sytes.net/%7Edavis/dlib/dlib/server/server_http_abstract.h.html>
>>>>>>
>>>>>> Cheers,
>>>>>> Davis
>>>>>>
>>>>>>
>>>>>> On Sat, Mar 14, 2009 at 7:34 PM, Steven Van Ingelgem <
>>>>>> st...@va...> wrote:
>>>>>>
>>>>>>> Corrected version (compiles on my linux and windows). Bugfixed all
>>>>>>> around via the previous example.cpp.
>>>>>>>
>>>>>>> Seems to work ok.
>>>>>>>
>>>>>>> 2009/3/14 Steven Van Ingelgem <st...@va...>
>>>>>>>
>>>>>>> Please find attached the updated http server for dlib.
>>>>>>>>
>>>>>>>> Now you can easily extend it to have additional features without
>>>>>>>> letting the on_request() string grow.
>>>>>>>>
>>>>>>>>
>>>>>>>> Grtz,
>>>>>>>> Steven
>>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>
>>>>>
>>>>
>>>
>>
>