If you are running behind, for example, apache, and the content type headers are
correct, apache might expand the gzip stuff for you. I wouldn't be surprised to
find that the python/django http request stuff doesn't handle gzip,
but it might.
It might be possible to return an error response saying coding not accepted,
prompting gnip to retry without gzip. I've certainly seen spotty adherence to
standards for applications to do posts for non-browserish things. The "None"
as request.encoding is suspicious, if that is spelled correctly.
As diagnostic stuff you might poke in the request object (using pdb on the
development server) to see if they are correctly setting the content type for
the mime part. And you could catch the post earlier than where it
gets converted
to a querydict (possibly using python urllib(2), et al rather than
django) to save
the post section to a file and see if it gunzips. If the data is good
gzipped stuff,
and the headers are correct, then you know to look on your (python,
django, front
end server) side. If the headers are bogus, you should push back to get gnip to
fix them.
Good luck.
On Thu, Feb 4, 2010 at 9:15 AM, Kenneth Loafman
<kenneth.loaf...@gmail.com> wrote:
> Guys,
>
> I hope you can help with a major Django problem I'm having. I have a
> feed from Gnip 2.0 using POST and its coming in as pure garbage in
> request.POST, like below:
>
> None
> twitter keyword
> [04/Feb/2010 13:55:15] "POST /api/gnip/collect/twitter/keyword/
> HTTP/1.0" 200 7
> POST = '<QueryDict:
> {u'm...@\u0480\ufffd\ufffd\ufffd\ufffda\ufffd\ufffdtp\ufffd\\\ufffd\ufffd...
> META = '{'PYTHONIOENCODING': 'UTF-8', 'wsgi.multiprocess': False,...
> len raw = 693
>
> The structure of the QueryDict, when printed fully, is a dictionary,
> complete with garbage keys and values.
>
> The code for this is quite simple:
> def api_gnip_collect( request, publisher=None, rule_type=None ):
> print request.encoding
> request.encoding = "utf-8" # I've tried utf-8,16,32,zlib,...
> print publisher, rule_type
> print "POST = '%s'" % request.POST
> print "META = '%s'" % request.META
> print "len raw =", len( request.raw_post_data )
>
> So, Gnip's other customers, not using Django, are not having this
> problem. Gnip says that the POST is gzip encoded. Is it possible that
> Django is not handling gzip'ed input?
>
> I've tried this on all versions 0.96 through 1.2alpha. No luck.
>
> ...Thanks,
> ...Ken
>
> --
> You received this message because you are subscribed to the Google Groups
> "Django users" group.
> To post to this group, send email to django-us...@googlegroups.com.
> To unsubscribe from this group, send email to
> django-users+unsubscr...@googlegroups.com.
> For more options, visit this group at
> http://groups.google.com/group/django-users?hl=en.
>
>
--
You received this message because you are subscribed to the Google Groups
"Django users" group.
To post to this group, send email to django-us...@googlegroups.com.
To unsubscribe from this group, send email to
django-users+unsubscr...@googlegroups.com.
For more options, visit this group at
http://groups.google.com/group/django-users?hl=en.
