Mercurial > p > roundup > code
view test/mocknull.py @ 5034:4ad465b09cbe
Update test/test_actions.py to work with py.test
The test_actions tests use a FieldStorage object to mock form
submissions. FieldStorage would usually use the 'QUERY_STRING'
environment variable to populate the initial list of MiniFieldStorage
values stored within the FieldStorage object, but because there is no
'QUERY_STRING' environment variable when running the test from the
commandline it tries to parse sys.argv instead. If any py.test options
are used (ie. '--tb=short') then they may end up in the MiniFieldStorage
list causing some tests to fail because they were not expecting the
extra MiniFieldStorage values.
To fix this we explicitly pass an dict with an empty 'QUERY_STRING'
value as the FieldStorage environ argument to ensure that the initial
MiniFieldStorage list is empty.
| author | John Kristensen <john@jerrykan.com> |
|---|---|
| date | Thu, 20 Aug 2015 16:30:50 +1000 |
| parents | 79fd8537ae3b |
| children | 3757449e00c4 |
line wrap: on
line source
class MockNull: def __init__(self, **kwargs): for key, value in kwargs.items(): self.__dict__[key] = value def __call__(self, *args, **kwargs): return MockNull() def __getattr__(self, name): # This allows assignments which assume all intermediate steps are Null # objects if they don't exist yet. # # For example (with just 'client' defined): # # client.db.config.TRACKER_WEB = 'BASE/' self.__dict__[name] = MockNull() return getattr(self, name) def __getitem__(self, key): return self def __nonzero__(self): return 0 def __str__(self): return '' def __repr__(self): return '<MockNull 0x%x>'%id(self) def gettext(self, str): return str _ = gettext
