🚀 go-pugleaf

For a multithreading REXX HTTP daemon I wrote using Object REXX, I want 
to have full WPS integration. This means that the daemon uses some object,
say <WWW_ROOT>, (normally a folder) as root for HTTP requests.

Now, when the server has to resolve a request for /archives/rexx, I want
the server to resolve the path even if it contains SOM proxies, instead
of folders. For example, shadows should be followed. In this way it's easy
to link in CD-ROM files or folders in the WWW tree. Further more, this
allows program-objects to be used as HTML-scripts even if they are in the
same path as the other HTML files (opposed to the /cgi-bin/ path that 
other daemons require).

The problems I have are:

  * given some path, say c:\WWW\Some_shadow_folder\Some_object, how
    do I get a reference to Some_object, so I can access it's contents
    if it's a file, or execute it if it's a program object. 

  * If the standard folder-class is replaced by some other class like
    for example Lan Server does (which uses LS... classes), how do I
    know what objects to treat like a container of other objects and
    what objects I should consider a data file?

I've got the Developer's Connection but even after browsing for a few
hours I couldn't find the right information. It might be I overlooked
some help-file: in that case, could anyone give me a pointer to it?

The Object REXX information always assumes the .wpDesktop standard object
to be used as a starting point, while I'd like to be able to use any
path as a starting point, even if it isn't a default REXX object.

Greetings,
  Geert

BTW: the REXX HTTP Daemon already works for normal HTML files and pictures,
and is just 500 lines of Object REXX code, including limited HTTP client
functionality. Object REXX is *really* great and simple. Hopefully some
visual environment will be built around it.  If anyone wants to have the 
current (premature) code to have a look, I'd be glad to mail it.
-- 
E-Mail: geert@sun3.iaf.nl 
 Phone: +31-53-4303054