[Baypiggies] urllib2, basic HTTP authentication, and tr.im

[Daryl Spitzer]

Thanks Benjamin.  I should have wrote sooner to add that I've
confirmed that curl -u works.  So the problem is definitely with my
Python basic HTTP authentication code.

I also should have explained that I want to use urllib2 because (as
Michael Foord wrote in
http://www.voidspace.org.uk/python/articles/urllib2.shtml#proxies),
urllib2 will auto-detect proxy settings and use them; curl--as far as
I know--doesn't auto-detect proxy settings.  (So my code using query
string parameters works fine at work through a proxy server and at
home.)

Also, I'd just like to learn why my code doesn't work!

I guess I should take jj's advice and look at what is being sent over
the wire.  (And compare that to what's going over the wire with curl.)
 I've never done that before.  Can anyone recommend a tool for that
(preferably that works on Mac OS X)?

--
Daryl


On Wed, Mar 11, 2009 at 2:07 PM, Benjamin Sergeant <bsergean at gmail.com> wrote:
> Warning: quick and dirty solution with no error checking using curl and a
> unix pipe.
>
> - Benjamin
>
> [bsergean at localhost bin]$ cat trim.py
> #!/usr/bin/env python
>
> import os, sys
> from getpass import getpass
> import xml.dom.minidom
>
> if len(sys.argv) != 2:
>     print 'Usage: trim.py <url to shorten>'
>     sys.exit(0)
>
> user = raw_input('username: ')
> passwd = getpass('password: ')
>
> # See http://tr.im/api
> url = 'http://api.tr.im/api/trim_url.xml?url=%s&newtrim=yes' % sys.argv[1]
>
> # Warning: --silent don't show error messages
> cmd  = 'curl --silent --basic -u %s:%s ' % (user, passwd)
> cmd += url
>
> fd = os.popen(cmd, 'r')
> response = fd.read()
> print response
> fd.close()
>
> Doc = xml.dom.minidom.parseString(response)
> N = Doc.getElementsByTagName("url")
> shortened_url = str((N[0].childNodes)[0].data)
>
> print 'shortened url:', shortened_url
>
>
> [bsergean at localhost bin]$ python trim.py
> Usage: trim.py <url to shorten>
> [bsergean at localhost bin]$ python trim.py
> http://www.reddit.com/r/programming/
> username: bsergean
> password:
> <?xml version="1.0" encoding="UTF-8"?>
> <trim>
>   <status result="OK" code="200" message="tr.im URL Added."/>
>   <url>http://tr.im/hglC</url>
>   <reference>ZfCAq7qcyouYXwfR2WsuBR6v3V4q7L</reference>
>   <trimpath>hglC</trimpath>
> </trim>
>
> shortened url: http://tr.im/hglC
>
>
>
>
>
>
> On Tue, Mar 10, 2009 at 11:51 PM, Shannon -jj Behrens <jjinux at gmail.com>
> wrote:
>>
>> It may help to look at what is being sent over the wire.
>>
>> -jj
>>
>> On Tue, Mar 10, 2009 at 8:26 AM, Daryl Spitzer <daryl.spitzer at gmail.com>
>> wrote:
>> > I'm playing around, trying to write some code to use the http://tr.im
>> > APIs (http://tr.im/api/) to shorten a URL.
>> >
>> > After reading http://docs.python.org/library/urllib2.html, I tried:
>> >
>> >    TRIM_API_URL = 'http://api.tr.im/api'
>> >    auth_handler = urllib2.HTTPBasicAuthHandler()
>> >    auth_handler.add_password(realm='tr.im',
>> >                              uri=TRIM_API_URL,
>> >                              user=USERNAME,
>> >                              passwd=PASSWORD)
>> >    opener = urllib2.build_opener(auth_handler)
>> >    urllib2.install_opener(opener)
>> >    response = urllib2.urlopen('%s/trim_simple?url=%s'
>> >                               % (TRIM_API_URL, url_to_trim))
>> >    url = response.read().strip()
>> >
>> > response.code is 200 (I think it should be 202).  url is valid, but
>> > the basic HTTP authentication doesn't seem to have worked, because the
>> > shortened URL isn't in my list of URLs (at http://tr.im/?page=1).
>> >
>> > After reading
>> > http://www.voidspace.org.uk/python/articles/authentication.shtml#doing-it-properly
>> > I also tried:
>> >
>> >    TRIM_API_URL = 'api.tr.im/api'
>> >    password_mgr = urllib2.HTTPPasswordMgrWithDefaultRealm()
>> >    password_mgr.add_password(None, TRIM_API_URL, USERNAME, PASSWORD)
>> >    auth_handler = urllib2.HTTPBasicAuthHandler(password_mgr)
>> >    opener = urllib2.build_opener(auth_handler)
>> >    urllib2.install_opener(opener)
>> >    response = urllib2.urlopen('http://%s/trim_simple?url=%s'
>> >                               % (TRIM_API_URL, url_to_trim))
>> >    url = response.read().strip()
>> >
>> > But I get the same results.  (response.code is 200 and url is valid,
>> > but not recorded in my account at http://tr.im/.)
>> >
>> > If I use query string parameters instead of basic HTTP authentication,
>> > like this:
>> >
>> >    TRIM_API_URL = 'http://api.tr.im/api'
>> >    response =
>> > urllib2.urlopen('%s/trim_simple?url=%s&username=%s&password=%s'
>> >                               % (TRIM_API_URL,
>> >                                  url_to_trim,
>> >                                  USERNAME,
>> >                                  PASSWORD))
>> >    url = response.read().strip()
>> >
>> > ...then not only is url valid but it's recorded in my tr.im account.
>> > (Though response.code is still 200.)
>> >
>> > Either there must be something wrong in my authentication code, or in
>> > tr.im's.  (I've written api at tr.im to see if it's a known problem and
>> > haven't received a response.)
>> >
>> > Can anyone spot any problems?
>> >
>> > --
>> > Daryl
>> > _______________________________________________
>> > Baypiggies mailing list
>> > Baypiggies at python.org
>> > To change your subscription options or unsubscribe:
>> > http://mail.python.org/mailman/listinfo/baypiggies
>> >
>>
>>
>>
>> --
>> In this life we cannot do great things. We can only do small things
>> with great love. -- Mother Teresa
>> http://jjinux.blogspot.com/
>> _______________________________________________
>> Baypiggies mailing list
>> Baypiggies at python.org
>> To change your subscription options or unsubscribe:
>> http://mail.python.org/mailman/listinfo/baypiggies
>
>