Am 9. September 2011 13:46 schrieb René Nussbaumer <[email protected]>:
> On Fri, Sep 9, 2011 at 13:14, Michael Hanselmann <[email protected]> wrote:
>> --- a/lib/rapi/baserlib.py
>> +++ b/lib/rapi/baserlib.py
>> +class _MetaOpcodeResource(type):
>> + """Meta class for RAPI resources.
>> +
>> + """
>
> Can you please elaborate the changes made by this metaclass to the
> underlying class in the docstring? Maybe based on the attributes. Or
> what attributes are possible.
See interdiff at the end.
>> +class OpcodeResource(ResourceBase):
>> + """Base class for opcode-based RAPI resources.
>> +
>> + Instances of this class automatically gain handler functions for any
>> method
>> + for which a C{$METHOD$_OPCODE} variable is defined at class level.
>> Subclasses
>> + can define a C{Get$Method$OpInput} method to do their own opcode input
>> + processing (e.g. for static values). The C{$METHOD$_RENAME} variable
>> defines
>> + which values are renamed (see L{FillOpCode}).
>> +
>> + """
>
> I would move that to the metaclass docstring.
Disagreed. If you click through the generated docs it's easier to find
the base class. I added a reference to the metaclass, though.
>> --- a/test/ganeti.rapi.baserlib_unittest.py
>> +++ b/test/ganeti.rapi.baserlib_unittest.py
>> +class TestOpcodeResource(unittest.TestCase):
>> + def testDoubleDefinition(self):
>> + class _TClass(baserlib.OpcodeResource):
>> + GET_OPCODE = opcodes.OpTestDelay
>> + def GET(self): pass
>> +
>> + self.assertRaises(AssertionError, _TClass, None, None, None)
>> +
>> + def testNoOpCode(self):
>> + class _TClass(baserlib.OpcodeResource):
>> + POST_OPCODE = None
>> + def POST(self): pass
>> +
>> + self.assertRaises(AssertionError, _TClass, None, None, None)
>
> Is this not just due to double definition? (See method above)
Nope, this one tests whether setting “None” as the opcode also raises
an exception when the HTTP method's direct handler is also defined.
>> + def testEmpty(self):
>> + class _Empty(baserlib.OpcodeResource):
>> + pass
>> +
>> + obj = _Empty(None, None, None)
>> + for attr in ["GetPostOpInput", "GetPutOpInput", "GetGetOpInput",
>> + "GetDeleteOpInput"]:
>> + self.assertFalse(hasattr(obj, attr))
>
> Could you also test a working set?
All other RAPI unittests (see the bazillion patches) are working sets.
Michael
