urllib2.urlopen(req) error........

[John F Dutcher]

I don't really think there's any Python bug...just thought my hoster's
implementation (it's new to them) might be flawed.
The dismay is that there are (9) scripts which dynamically create html
pages
which work flawlessly on Win XP where I develope them. As with another
similar site I did....every single script has to be tweaked before it
will run on Linux
servers with Linux Python. In the current case, as with 90% of the
others, it's always 'premature end of script headers' as the error.
While the 'printing' of
the 'Content type:....' statements are very much in place, with the
accompanying '\n\n' to generate a blank line....there is still an
unfortunate session of trial and error to get past this issue. The
offending script noted here fails as it does 'process_req(ls)' (ls
being a list of parameters for the script being loaded (read)).
I just thought something 'new' might be going on given the point of
failure.
  


jjl at pobox.com (John J. Lee) wrote in message news:<87zn7ivtn9.fsf at pobox.com>...
> John_Dutcher at urmc.rochester.edu (John F Dutcher) writes:
> 
> > Can anyone comment on why the code shown in the Python error is in
> > some way incorrect...or is there a problem with Python on my hoster's
> > site ??
> > The highlites don't seem to show here...but line #80 and line # 38 are
> > the first line offenders.
> 
> So (writing your post for you, here) you've written a CGI script in
> Python.  Your script uses urllib2 to POST some data to a server, and
> you're including the tracback from cgitb.
> 
> >    36         req = urllib2.Request(url='http://www.euromillions.us/scgi-bin/euro8.py',\
> >    37         data = strdata)
>  [...]
> > code = 500, msg = 'Internal Server Error', hdrs = <httplib.HTTPMessage
> > instance>
> [...]
> 
> euromillions.us is trying to tell you that euromillions.us is broken
> (see RFC 2616, section 10.5.1 '500 Internal Server Error').  Maybe you
> tickled the bug in the software running on euromillons.us by POSTing
> bad data?
> 
> The problem is highly unlikely to be the result of a bug in the Python
> library.
> 
> 
> John