I completly agree. And as you mentioned different frameworks might implement different behaviors. I don't recall exactly but I think I was under the impression that Wicket didn't make a difference. /a?.. seemed to return the same result as /a/?.. Whether is by intention or by accident is what I wasn't sure about.
-Juergen On Thu, Jun 11, 2009 at 11:19 AM, Korbinian Bachl - privat<[email protected]> wrote: > Hi, > > IMHO WICKET-1597 should be applied soon, however there seems 1 last question > from Juergen Donnerstag needed to be clearified (and maybe implemented): > > "Applied the change and modified the test cases. Since I wasn't sure if > /a/?param is the same as /a?param, I didn't commit it but attached the > patch." > > IMHO /a/?param is NOT the same as /a?param > > Reason: different paths! > > while the queries are exactly the same, the paths arent; > > Example: > /a -> goes to path-root "/" and requests file "a"; > /a/ -> goes not to path-root but instead to folder called "a" and requests > document "null" that gets converted to the default-request document for that > folder - often some index.htm or sth. like that; in case there is no default > target however, it may be that the webserver/ container getting the requests > decides to truncate the last "/" and reroutes the request to "/a" - however, > this is dependend on specific container-configuration and mustn't be seen as > valid default; > > see also RFC 3986 example under http://en.wikipedia.org/wiki/URI_scheme > where exactly is defined that the parts "/a" and "/a/" each belong to the so > called <hierarchical part> while the ? sperates it from the query part ( [ ? > <query> ]); > > Best, > > Korbinian >
