Kenneth Reitz
0d8c08cd22
binary data upload bug test
2012-02-15 02:03:29 -05:00
Kenneth Reitz
ab4c03adc3
cleanups
2012-02-15 02:03:13 -05:00
Kenneth Reitz
75b0c4be1b
latin1, remove unicode_literals
...
Fixes #407
2012-02-15 02:01:36 -05:00
Kenneth Reitz
4a8e6415c7
sp
2012-02-15 01:47:08 -05:00
Kenneth Reitz
58bf158840
RuntimeError
2012-02-15 01:46:06 -05:00
Kenneth Reitz
c0f066626e
whitespace
2012-02-15 01:42:57 -05:00
Kenneth Reitz
69c7a8638e
make all the tests
2012-02-15 01:42:52 -05:00
Kenneth Reitz
1b6f1b091c
Merge branch '400-header_expand-unicode' of https://github.com/acdha/requests into develop
2012-02-15 01:25:29 -05:00
Kenneth Reitz
d39cdc45d1
decode response url again
2012-02-15 01:21:38 -05:00
Kenneth Reitz
7f761f3adf
history for 0.10.2
2012-02-15 01:18:41 -05:00
Kenneth Reitz
5e76943f02
Merge branch 'develop' of https://github.com/Lukasa/requests into develop
2012-02-15 01:08:10 -05:00
Kenneth Reitz
b887ea3b60
Merge branch 'mgiuca-uri-cleanup-nonpath' into develop
2012-02-15 01:00:52 -05:00
Matt Giuca
690426ac86
Added Matt Giuca to AUTHORS.
2012-02-15 12:06:34 +11:00
Matt Giuca
b8298b0710
Rename utils.requote_path to requote_uri.
...
It is now applied to a whole URI, not just the path.
2012-02-15 12:03:08 +11:00
Matt Giuca
3bd5406c06
Now requotes the entire URL, not just the path (Issue #429 ).
2012-02-15 12:00:47 +11:00
Matt Giuca
3ae0ff3685
Added test cases for URI cleaning in the query part (Issue #429 ).
2012-02-15 12:00:27 +11:00
Cory Benfield
80e29bfcab
Change the RFC we reference.
2012-02-14 14:11:25 +00:00
Cory Benfield
d7107c2cdc
Make clearer the source of forbidden characters
2012-02-14 12:58:35 +00:00
Cory Benfield
658da765ec
Keep non RFC cookie key characters separate.
...
Moved the characters that we accept in keys but that the RFC forbids
into a separate string to keep track of them.
2012-02-14 10:25:40 +00:00
Kenneth Reitz
9e2f5dd0ed
Merge branch 'py3-uri-encoding' of https://github.com/mgiuca/requests into develop
2012-02-14 04:30:10 -05:00
Matt Giuca
1ffce4f7dc
Simplify requote_path.
...
It no longer needs to split on '/' since '/' will not be encoded.
2012-02-14 12:57:49 +11:00
Matt Giuca
fcac1c3746
Fixed URI encoding of reserved characters (Issue #369 ).
...
Previously, util.requote_path would unquote and requote all characters,
causing reserved characters to become encoded (changing the semantics of the
URI). Now, it has special code for unquoting just the unreserved characters,
then quotes only illegal characters.
This ensures that illegal characters are fixed, and URIs are normalised, but
reserved characters do not erroneously become quoted.
Test case test_session_with_escaped_url now passes.
2012-02-14 12:51:03 +11:00
Matt Giuca
c0763bb8d5
Changed test case test_path_is_not_double_encoded.
...
Uses the space character instead of ~ as a test (since ~ is unreserved, it
shouldn't really be encoded at all).
2012-02-14 12:49:31 +11:00
Matt Giuca
75bd9d0e94
Added test case testing URI normalisation.
...
Tests that percent-encoded unreserved characters get de-encoded.
2012-02-14 12:20:48 +11:00
Matt Giuca
08bc1198d3
Added test case for Issue #369 .
...
Tests that legal reserved and unreserved characters in the path are not
percent-encoded. Currently fails.
2012-02-14 12:15:32 +11:00
Matt Giuca
cf54f60367
Fixed URI re-encoding on Python 3 (Issue #369 ).
...
Request.full_url now performs requoting of the path (like it does in Python 2).
Request.path_url no longer quotes the already-quoted path (double quoting).
Fixed utils.requote_path so it works properly in Python 3.
2012-02-14 11:50:02 +11:00
Matt Giuca
329efcae6e
Added test case for quoted and illegal characters in the URL.
...
This is a test case for issue #369 .
It passes in Python 2, but fails in Python 3 (though the test suite doesn't
actually work in Python 3, it can be shown to fail with some modification).
2012-02-14 11:47:14 +11:00
Kenneth Reitz
5a2a78f7df
Merge pull request #425 from Lukasa/develop
...
Issue #335 : Accept unusual keys in cookies.
2012-02-13 16:14:39 -08:00
Cory Benfield
f72c13ffda
Accept unusual keys in cookies.
...
The Oreos module throws CookieErrors when it catches keys that have
characters that shouldn't be in them, like colons and braces. Other
services are happy to send out such keys, however. This commit makes
requests non-standards compliant, just like everyone else is.
2012-02-14 00:10:26 +00:00
Chris Adams
0a17badb21
header_expand: handle Unicode strings ( closes #400 )
2012-02-13 11:10:05 -05:00
Kenneth Reitz
12f9aa3669
Merge pull request #410 from eriolv/develop
...
Fix SyntaxError on Python3 because "except Error, e" is not supported anymore
2012-02-12 02:16:48 -08:00
Kenneth Reitz
afa2e86a4d
Merge pull request #413 from bancek/patch-1
...
Typo fix in oreos/structures.py
2012-02-06 09:08:03 -08:00
Luka Zakrajšek
b7842ab34b
Typo fix in oreos/structures.py
2012-02-06 13:42:34 +01:00
Greg Taylor
f8b3141204
Use basestring instead of str to determine when to return headers as-is. This makes sure that unicode instances aren't enumerated.
2012-02-05 13:57:35 -05:00
Daniele Tricoli
c40d6d0509
Since requests 0.10.1 dropped Python5 support, to catch an exception and
...
store the exception object in a variable the "as" keyword can be used so
Python3 will not raise a SyntaxError
2012-02-05 04:18:22 +01:00
Kenneth Reitz
a74481a1d4
Update docs/community/faq.rst
2012-02-03 13:24:00 -05:00
Kenneth Reitz
b898415d0d
:trollface:
...
because i can.
2012-02-02 17:23:47 -05:00
Kenneth Reitz
b47effe5d8
test #2
2012-02-01 13:35:34 -05:00
Kenneth Reitz
4f10605389
quick test
2012-02-01 13:33:59 -05:00
Kenneth Reitz
120e62900b
Native Instruments!
2012-01-31 02:15:55 -05:00
Kenneth Reitz
986704dc62
v0.10.2
2012-01-30 09:41:09 -05:00
Kenneth Reitz
2f13956357
Merge branch 'unicode_missing_self' of https://github.com/Telofy/requests into develop
2012-01-30 09:35:53 -05:00
Kenneth Reitz
8d700dd1e5
Merge branch 'pr-urlfix-1' of https://github.com/brendoncrawford/requests into develop
2012-01-30 09:33:31 -05:00
Kenneth Reitz
5be412db13
Brendon Crawford
2012-01-30 09:33:14 -05:00
Telofy
a61a861cfb
The self was missing again, and I think the two blocks were redundant.
2012-01-30 12:23:43 +01:00
Brendon Crawford
c363ca476f
Added ability to check if missing trailing slash in path.
...
Previously this URL would not work:
http://example.com?foo=bar
But this URL would work:
http://example.com/?foo=bar
Even though a slash is required for a valid URL, it is a common use case
for the slash to be left out, so it would be a good idea to account for this.
2012-01-29 23:24:54 -08:00
Kenneth Reitz
319b1ed551
Merge branch '239-socket-leak' of https://github.com/acdha/requests into develop
2012-01-29 19:06:41 -05:00
Chris Adams
67f50a0fe5
Less redundant freeing of resources on redirects
...
With 495646bb6a the double flush is no longer
needed.
2012-01-29 18:45:14 -05:00
Chris Adams
b98cc8b4e9
Revert "Initial stop-the-bleeding fix"
...
This reverts commit 730c0e2e2b .
2012-01-29 18:43:13 -05:00
Chris Adams
495646bb6a
async: force prefetch=True on send
...
… otherwise we can run out of sockets before any of the responses are ever
processed
2012-01-29 18:42:48 -05:00