Hi Andreas, Andreas Jung schrieb:
--On Montag, 15. Dezember 2003 14:36 Uhr +0100 Tino Wildenhain <tino@wildenhain.de> wrote:
Hi Andreas,
Andreas Jung schrieb:
This a single REQUEST is either a POST or GET request there is no need for this separation IMO. Everything's in the REQUEST.
That's not entirely true. Sadly there is a design flaw in cgi.py which ignores GET variables still present when doing POST.
The whole lib is a bit ugly so its not easy to fix it. And its even worser to fix if you want to achieve compatibility with current implementations.
Should variables be merget? Get variables before POST or vice versa?
in the HTTP specs we can find: """ The POST method is used to request that the origin server accept the entity enclosed in the request as a new subordinate of the resource identified by the Request-URI in the Request-Line"""
How shall we interpret that? Parameters passed as QUERYSTRING in a POST request do override parameters passed in the body?
Anyway I don't see why there is a need to mix both. For clearity one should either pass the parameters as POST or GET.
There are really a lot of use cases for that. You can only use <a hrefs> and redirect to do parameter passing via GET. If the target page uses POST to submit another information, the information on GET is lost for current Zope. Maybe next example makes it more transparent: you make a link published in a e-mail which provides information of an article in a shop referenced by QUERY-String, of course. In the resulting page there is the option to log in first - of course with POST. Currently you have to transport every paramater from query-String to a hidden form field to not loose the values. Thats a bit nasty. Regards Tino