[TIP] Capture HTTP traffic

[C. Titus Brown]

On Wed, Nov 03, 2010 at 11:10:21AM +0100, Alex wrote:
> I am developing an automated test in Python/Windows and I have to
> ensure certain HTTP requests are made and verify the GET parameters
> and responses. I need a tool to help me easily to capture and filter
> the HTTP requests.
> 
> A typical test case is like:
> 
>     def setUp()
>         self.capture = HttpCapture() # hypothetic
>     def testFoo(self):
>         self.capture.start("http://update.foobar.com")
>         subprocess.Popen(...)
>         self.capture.end()
>         # suppose self.capture.result() returns a list of strings,
> containing the URLs
>         # have been visited betwen begin() and end()
>         self.failUnless("http://update.foobar.com/" in
> self.capture.result() )
> 
>  It seems to me that it's not trivial to integrate 'pycap' into this
> picture. I wouldn't mind that the hypothetic HttpCatpure() launchs an
> external application. In fact, I did try to automate Wireshark by
> myself but it didn't seem to fit well into this test case design
> neither.

Hi Alex,

you could write a simple HTTP server of your own, and run it in another
thread.  That doesn't help if you need to call remote URLs, though.

You could also develop a simple proxy server and funnel things through
that, recording the data that's passed through.  The 'scotch' toolkit is a stab
at doing that,

	http://darcs.idyll.org/~t/projects/scotch/doc/

and it used to work OK but I haven't played with it in years.

cheers,
--titus
-- 
C. Titus Brown, ctb at msu.edu