Merge lp:~barry/wadllib/lp1132125 into lp:wadllib

[Barry Warsaw]

Fixes the Python 3.3 hash randomization failures by imposing a sort order in a couple of places. This may or may not be semantically correct. The alternative would be to add a urlparsing method to the README and parse the query string into a dict, then compare the dict, but it would make the README uglier.

Please let me know if this solution seems okay. If so, I'll happily merge to trunk, and upload to PyPI and Raring. If not, I'll adjust the tests.

[Gary Poster (gary)]

Hi Barry. I've not looked at this code in a loooong time, but your changes look eminently reasonable to me. Thanks.

Gary

review: Approve

[Barry Warsaw (barry)]

Thanks! Merging and uploading...

=== modified file 'src/wadllib/NEWS.txt'
--- src/wadllib/NEWS.txt	2012-03-22 17:01:20 +0000
+++ src/wadllib/NEWS.txt	2013-02-25 21:26:22 +0000
@@ -2,6 +2,15 @@
 NEWS for wadllib
 ================
 
+1.3.2 (2013-02-25)
+==================
+
+- Impose sort order to avoid test failures due to hash randomization.
+  LP: #1132125
+- Be sure to close streams opened by pkg_resources.resource_stream() to avoid
+  test suite complaints.
+
+
 1.3.1 (2012-03-22)
 ==================
 

=== modified file 'src/wadllib/README.txt'
--- src/wadllib/README.txt	2012-01-27 18:12:53 +0000
+++ src/wadllib/README.txt	2013-02-25 21:26:22 +0000
@@ -1,5 +1,5 @@
 ..
-   Copyright (C) 2008-2009 Canonical Ltd.
+   Copyright (C) 2008-2013 Canonical Ltd.
 
    This file is part of wadllib.
 
@@ -35,9 +35,11 @@
 
 Or the second argument may be an open filehandle containing the markup.
 
+   >>> cleanups = []
    >>> def application_for(filename, url="http://www.example.com/"):
    ...    wadl_stream = pkg_resources.resource_stream(
    ...    'wadllib.tests.data', filename)
+   ...    cleanups.append(wadl_stream)
    ...    return Application(url, wadl_stream)
    >>> wadl = application_for("launchpad-wadl.xml",
    ...                        "http://api.launchpad.dev/beta/")
@@ -535,7 +537,7 @@
 
    >>> create_team_method.build_representation(
    ...     display_name='Joe Bloggs', name='joebloggs')
-   ('application/x-www-form-urlencoded', 'display_name=Joe+Bloggs&ws.op=newTeam&name=joebloggs')
+   ('application/x-www-form-urlencoded', 'display_name=Joe+Bloggs&name=joebloggs&ws.op=newTeam')
 
 The return value of build_representation is a 2-tuple containing the
 media type of the built representation, and the string representation
@@ -561,6 +563,7 @@
 
    >>> binary_stream = pkg_resources.resource_stream(
    ...     'wadllib.tests.data', 'multipart-binary-wadl.xml')
+   >>> cleanups.append(binary_stream)
    >>> binary_wadl = Application(
    ...     "http://www.example.com/", binary_stream)
    >>> service_root = binary_wadl.get_resource_by_path('')
@@ -647,6 +650,11 @@
    ...     'post', representation_params={ 'foo' : 'bar' }))
    None
 
+.. cleanup
+   >>> for stream in cleanups:
+   ...    stream.close()
+
+
 .. toctree::
    :glob:
 

=== modified file 'src/wadllib/application.py'
--- src/wadllib/application.py	2012-03-12 20:41:45 +0000
+++ src/wadllib/application.py	2013-02-25 21:26:22 +0000
@@ -1,4 +1,4 @@
-# Copyright 2008 Canonical Ltd.  All rights reserved.
+# Copyright 2008-2013 Canonical Ltd.  All rights reserved.
 
 # This file is part of wadllib.
 #
@@ -732,7 +732,7 @@
                 append = '&'
             else:
                 append = '?'
-            url += append + urlencode(validated_values)
+            url += append + urlencode(sorted(validated_values.items()))
         return url
 
 
@@ -834,7 +834,7 @@
             params, param_values, **kw_param_values)
         media_type = self.media_type
         if media_type == 'application/x-www-form-urlencoded':
-            doc = urlencode(validated_values)
+            doc = urlencode(sorted(validated_values.items()))
         elif media_type == 'multipart/form-data':
             outer = MIMEMultipart()
             outer.set_type('multipart/form-data')