From 85ed9fdbad6de406fc117ff900e4f274840ab19e Mon Sep 17 00:00:00 2001 From: Eli Collins Date: Sun, 25 Jan 2015 12:42:23 -0500 Subject: bugfix: passlib.ext.django: clarified & tweaked logic for when we pass make_password()'s salt parameter on to the hash (fixes issue 52). old behavior would incorrectly pass explicit salt provided for hash that didn't need one; a situation that only occurs in django 1.4.0-1.4.5's unittests, and was corrected in django 1.4.6's. new behavior separates out a couple of the cases we were trying to handle, handles them separately, has better comment explaining what's going on. --- tox.ini | 10 ++++++++++ 1 file changed, 10 insertions(+) (limited to 'tox.ini') diff --git a/tox.ini b/tox.ini index e52df3b..6b47cc9 100644 --- a/tox.ini +++ b/tox.ini @@ -133,6 +133,16 @@ deps = commands = nosetests {posargs:passlib.tests.test_ext_django passlib.tests.test_handlers_django} +[testenv:django145] +# NOTE: doing this to detect regression of issue 52, since django < 1.4.6 +# has some slight differences in the hasher classes & tests. +deps = + django==1.4.5 + bcrypt + {[testenv]deps} +commands = + nosetests {posargs:passlib.tests.test_ext_django passlib.tests.test_handlers_django} + [testenv:django14] deps = django<1.5 -- cgit v1.2.1