Shibboleth Developers

["Howard Gilbert" <>]

> > If you want to know the URL
> > of the original resource, to do some form of RequestMap
> > processing, then the code that extracts and decodes the URL
> > from the target= parameter has to check for nested targets
> > and do a second parse-decode. 
> 
> How would you know to do this (unless it was doing this all the time)?
> What
> if it was just a regular resource that happened to have a query string
> parameter called "target"?

The middle URL can contain something specific. It has to map to two new
functions: a new Servlet function of CAS and a new operation of the Filter.
So the URL has to have something in it that matches the servlet-mapping and
filter-mapping of the two WEB-INF/web.xml files in the CAS and Resource
contexts. For example, right now I do a mapping on "*.shire" to the Filter
because that's the way the C++ Filter installation instructions tell me to
map the ISAPI Filter. In practice, the only actual SHIRE URL is typically
/contextname/Shibboleth.shire and, I suppose, the mapping is too general if
anyone ever writes an application with files that actually end in .shire.

I don't want to pin down the exact syntax now. However, if you get a target
that points to a URL with a file name of (example)
"ProcessNewShibSessionId.passiton" and it has a target= parameter, then you
can be pretty sure it's a nested target.