Re: Before I file a bug report...
On Sun, 30 Sep 2001, Jon Stevens wrote: > > Mmmm, I'm not too sure about all this. > > The servlet spec seems to imply that the extension mapping is on the end of > > the path. > > > > Section 11.1, item #3 : If the last segment in the URL path contains an > > extension (e.g. .jsp), the servlet container will try to match a servlet > > that handles requests for the extension. > > An extension is defined as the part of the last segment after the last '.' > > character. > > > > Remy > > I think the Servlet Spec is wrong then because HTTP says that anything after > the extension mapping is PATH_INFO. It Is also inconsistent with the way > that servlet path mapping works as well. Well, I agree with Jon, but we had this whole discussion almost a year ago, and the conclusion was that the HTTP spec is wrong and we should follow the Servlet spec. Older versions of tomcat supported path info on extension mapped servlets, that was removed since it was considered a spec compliance bug. Workaround - declare each page with exact mappings in web.xml. Costin
Re: Before I file a bug report...
> on 9/30/01 3:15 PM, "Remy Maucherat" <[EMAIL PROTECTED]> wrote: > > > Mmmm, I'm not too sure about all this. > > The servlet spec seems to imply that the extension mapping is on the end of > > the path. > > > > Section 11.1, item #3 : If the last segment in the URL path contains an > > extension (e.g. .jsp), the servlet container will try to match a servlet > > that handles requests for the extension. > > An extension is defined as the part of the last segment after the last '.' > > character. > > > > Remy > > I think the Servlet Spec is wrong then because HTTP says that anything after > the extension mapping is PATH_INFO. It Is also inconsistent with the way > that servlet path mapping works as well. We're supposed to implement path mapping the way it is specified in the servlet API. So if I read the spec correctly, it says that /foo.jsp goes to the *.jsp mapping, and /foo.jsp/bar goes to the default servlet, since here /bar would be the last segment of the URI. Remy
Re: Before I file a bug report...
on 9/30/01 3:15 PM, "Remy Maucherat" <[EMAIL PROTECTED]> wrote: > Mmmm, I'm not too sure about all this. > The servlet spec seems to imply that the extension mapping is on the end of > the path. > > Section 11.1, item #3 : If the last segment in the URL path contains an > extension (e.g. .jsp), the servlet container will try to match a servlet > that handles requests for the extension. > An extension is defined as the part of the last segment after the last '.' > character. > > Remy I think the Servlet Spec is wrong then because HTTP says that anything after the extension mapping is PATH_INFO. It Is also inconsistent with the way that servlet path mapping works as well. -jon
Re: Before I file a bug report...
> Ok, I have found the problem. It is a result of an incorrectly written piece > of code in StandardContextMapper.map() under // Rule 3 > > I'm working on a rewrite that will parse things the way that they should be > parsed. :-) The issue is that the relativeURI variable can be pretty much > anything so it is difficult to figure out what the actual file suffix is out > of the path... > > Some initial testing shows that the relativeURI variable could be any number > of bits of information... > > uri: /Links.vm > uri: /Links.vm/foo/bar > uri: /Links.vm;foo/bar > uri: /Links.vm:foo/bar > uri: /Links.vm.foo/bar > > So, what I'm going to do is start at the first "." and then attempt to > resolve each additional character in the URI after the period against the > context.findServletMapping() until a match is found or the end of the URI is > reached. Mmmm, I'm not too sure about all this. The servlet spec seems to imply that the extension mapping is on the end of the path. Section 11.1, item #3 : If the last segment in the URL path contains an extension (e.g. .jsp), the servlet container will try to match a servlet that handles requests for the extension. An extension is defined as the part of the last segment after the last '.' character. Remy
Re: Before I file a bug report...
Ok, I have found the problem. It is a result of an incorrectly written piece
of code in StandardContextMapper.map() under // Rule 3
I'm working on a rewrite that will parse things the way that they should be
parsed. :-) The issue is that the relativeURI variable can be pretty much
anything so it is difficult to figure out what the actual file suffix is out
of the path...
Some initial testing shows that the relativeURI variable could be any number
of bits of information...
uri: /Links.vm
uri: /Links.vm/foo/bar
uri: /Links.vm;foo/bar
uri: /Links.vm:foo/bar
uri: /Links.vm.foo/bar
So, what I'm going to do is start at the first "." and then attempt to
resolve each additional character in the URI after the period against the
context.findServletMapping() until a match is found or the end of the URI is
reached.
Something like (untested):
String dotToEnd = relativeURI.substring(relativeURI.indexOf('.'))
StringBuffer sb = new StringBuffer();
for (int i=0; i wrote:
> I want to get confirmation that this is a valid url:
>
> http://localhost:8080/examples/jsp/snp/snoop.jsp/foo
>
> Where:
>
> PATH_INFO=/foo
>
> The error that I get right now is:
>
> HTTP Status 404 - /jsp/snp/snoop.jsp/foo
> The requested resource (/jsp/snp/snoop.jsp/foo) is not available.
>
> Please note that:
>
> http://localhost:8080/context/servlet/Bar/foo
>
> Works just fine. I don't see why it shouldn't also work with file extension
> mapping.
>
> thanks,
>
> -jon
--
If you come from a Perl or PHP background, JSP is a way to take
your pain to new levels. --Anonymous
Before I file a bug report...
I want to get confirmation that this is a valid url: http://localhost:8080/examples/jsp/snp/snoop.jsp/foo Where: PATH_INFO=/foo The error that I get right now is: HTTP Status 404 - /jsp/snp/snoop.jsp/foo The requested resource (/jsp/snp/snoop.jsp/foo) is not available. Please note that: http://localhost:8080/context/servlet/Bar/foo Works just fine. I don't see why it shouldn't also work with file extension mapping. thanks, -jon
