| Commit message (Collapse) | Author | Age | Files | Lines |
| |
|
|
|
|
|
|
|
|
|
| |
In unittest2, assertDictEqual() is implemented by
using != operator to compare two dicts. So is
assertEqual() in testtools. assertEqual() in
testtools is able to handle dict, list, set and
so on. So we just call assertEqual() to make the
unit tests simpler.
Change-Id: Ice343b2ce468acae39d2ad79f7121503e3627656
|
| |
|
|
|
|
|
| |
Leave parseractions.py and test_parseractions.py as a sanity check during the
deprecation period.
Change-Id: I1a7469b6d872284e0276502a1a287bc0b87f8f83
|
| |
|
|
|
|
|
|
|
|
|
|
| |
The set --property command requires that the input match
the "key=value" type, but if the type don't match, the return
value will be None, and the command still can be implemented
successfully, this may confuse the users. I think we should
raise exception if the argument type don't match "key=value".
So I make some changes in KeyValueAction class in this patch.
Change-Id: I14e64922faa7e083bc8b5e7e1cac41ef8117c224
Closes-Bug: #1589935
|
| |
|
|
|
|
|
|
| |
assertListEqual() is order sensitive. So we need to sort
the lists before we compare them. Use assertItemsEqual()
instead is better.
Change-Id: I9eaa98716c7401f5b099b007438acc916dae619b
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
| |
Class MultiKeyValueAction will be used to parse arguments like this:
--route destination=xxx,gateway=xxx --route destination=yyy,gateway=yyy
The result is a list like this:
[{destination:xxx, gateway:xxx}, {destination:yyy, gateway:yyy}]
This action also contain validation of the parameters.
Change-Id: Ie3aa8635c6a13fc2e429fe6922acd681dc7244cf
|
| |
|
|
| |
Change-Id: I12846acc4450d31d19897bbdfc6846bde8c8f2ce
|
| |
|
|
|
|
|
|
| |
No need to initialize parser in each test case. Do it in setUp().
Also remove the test_default_values case because it could be tested
in the test_good_values case.
Change-Id: Ia2ed7c9e46bf6baabbd62b9d50511c5e8103e5e2
|
| |
|
|
|
|
|
|
| |
This patch is going to replace assertEqual(None, *) with
assertIsNone(*) in unit test code to have more clear messages
in case of failure.
Change-Id: I6f85498347e8fc7cad5ea7afb832b9acda7daafc
|
| |
|
|
|
|
|
|
| |
The behaviors are inconsistent while different
negative line numbers specified.
Change-Id: I2573f3e789f5603c896758971830ffc0b94c5e2b
Closes-Bug: #1512263
|
| |
|
|
|
|
| |
Let's fix them thoroughly.
Change-Id: I8a1f042fb614b05c9836a49041f3883638b870b0
|
| |
|
|
| |
Change-Id: I048ee857fc1215fea7f60978364894e1b5abdf66
|
|
|
There are several reasons for this. One is that the majority of
OpenStack packages behave this way. The second is that it makes writing
software that extends something easier to test (which is a clear usecase
for openstackclient) And third, tests/__init__.py implies a global
package named "tests" - which I'm pretty sure we're not providing.
Change-Id: Ic708ffd92aea78c2ffc1a8579af0587af4fca4ff
|