[racket-dev] comment boxes

Either way, what I need is some way to pull out the bindings (with the
attachment) from a cgi script in some form.  Once I can do that, I'll
be able to add the attachement field.


On Jun 29, Jay McCarthy wrote:
> The Web server implements the multipart/form-data reading as well.
> 
> Jay
> 
> On Tue, Jun 29, 2010 at 3:07 PM, Neil Van Dyke <neil at neilvandyke.org> wrote:
> > Eli Barzilay wrote at 06/29/2010 04:46 PM:
> >>
> >> 2. Gnats *does* know how to deal with attachements.  The problem is
> >>   that our cgi library cannot handle a multipart/form-data encoding.
> >>   (If you convince the author to add support for that, I'll do the
> >>   rest and add a file attachement option to the web form, and
> >>   changing drracket to send a file is probably easy too.)
> >
> > I implemented "multipart/form-data" reading a couple years ago as part of
> > robust CGI support for a proprietary application.
> >
> > As I recall, the HTTP "post" MIME-reading code is only a little bit
> > application-specific.  (It writes file attachments directly to disk, and
> > adds metadata about the attachments to the form field alist in a certain
> > application-specific (constraining) way.  It might also do some stuff with
> > character encodings that are not strictly backwards-compatible with Racket's
> > "cgi" library.)
> >
> > If anyone has time and interest in adapting this HTTP "post" MIME-reading
> > code to improve core Racket, I suspect that the copyright owner would be
> > willing and able to contribute it.

-- 
          ((lambda (x) (x x)) (lambda (x) (x x)))          Eli Barzilay:
                    http://barzilay.org/                   Maze is Life!