- 11 Feb, 2015 4 commits
-
-
Chris Dent authored
Otherwise tests that provide data for subsequent tests that are marked xfail for minor problems would not have their data available.
-
Chris Dent authored
-
Chris Dent authored
Add support for skipping or expecting failure on a test
-
Chris Dent authored
-
- 10 Feb, 2015 2 commits
-
-
Chris Dent authored
This is needed because during the test writing process there are often _many_ things that ought to succeed but don't that are beyond the scope of the immediate project. For example a framework not properly returning a 405 when a method is not allowed, or returning 405 but then not returning the Allow header. This feature required switching the base test case from unittest to testtools. The former has changed (in Python 3.4) the internals of how expecting failure is handled and as a result making it work with dynamically generated tests was proving extremely difficult. An additional shell driven test is provided for coverage of skip and fail.
-
Chris Dent authored
It wasn't counting the number of tests and it should to say it is just running one test.
-
- 05 Feb, 2015 2 commits
-
-
Chris Dent authored
-
Chris Dent authored
In some APIs it is useful to refer to the response of a previous request in the response of this current request to, for example, verify that an ID continues to be used. Up to now $RESPONSE handling worked in url and data, this change adds it to response_strings and response_json_paths. As with all of the $RESPONSE handling, this is power that should really only be used when testing existing APIs. If you need a lot of this when you are developing new APIs YOU ARE DOING IT WRONG.
-
- 02 Feb, 2015 6 commits
-
-
Chris Dent authored
-
Chris Dent authored
-
Chris Dent authored
Having coverage information will help improve future testing. Also add some POST tests to enhance coverage of the the SimpleWsgi test backend.
-
Chris Dent authored
The limit tox target confirms that interception still works when limiting the tests that will be run.
-
Chris Dent authored
If a pattern is passed to the test runner such that only some tests in a YAML file are to be run, the unrequested tests are represented in the TestSuite as empty TestSuites. These do not have a 'fixtures' attribute. This can result in the WSGI-intercept not being discovered and installed. Instead of asking for the first test when looking for fixtures, we traverse the suite until we find the first test which has it.
-
Chris Dent authored
If a test specifies a header which is not present in the response provide a verbose message of what headers are present. Fixes #10.
-
- 26 Jan, 2015 3 commits
-
-
Chris Dent authored
Prompted by flake8 and pylint.
-
Chris Dent authored
Every test needs its own name or otherwise we can't track it.
-
Chris Dent authored
HttpTestCase can inherit from unittest not testtools.
-
- 22 Jan, 2015 2 commits
-
-
Chris Dent authored
-
Chris Dent authored
Some WSGI applications require a lot of prior setup before they can be use. Since we want most of that setup to happen in fixtures, wsgi-interception is delayed until after all extant fixtures have been started. This is done with a built in default fixture that is run as the most deeply nested of the context managers. The intecept callable is passed on along to the tests so that suite can be made aware of and use it when using the intercept context manager. If the callable is a function, when it is passed into the class builder, it can become bound to the TestCase. This binding is unbound in the caller.
-
- 21 Jan, 2015 1 commit
-
-
Chris Dent authored
If a fixture class raises unittest.case.SkipTest then all the tests in the YAML file associated with that test will be skipped.
-
- 16 Jan, 2015 1 commit
-
-
Chris Dent authored
For clarity and readability. This is starting to point out that there really ought to be unit tests for the various modules. The gabbits confirm things end to end but not the details.
-
- 14 Jan, 2015 7 commits
-
-
Chris Dent authored
README.rst: Nicer formatting for commands
-
Marc Abramowitz authored
-
Chris Dent authored
An error meant that the name of the current test included the name of the prior test. Each test name getting longer and longer.
-
Chris Dent authored
To keep things moving.
-
Chris Dent authored
-
Chris Dent authored
-
Chris Dent authored
This separates the loading of the data from turning that data into tests, making things a bit more readable as well as opening the eventual possibility of loading from other formats.
-
- 09 Jan, 2015 2 commits
-
-
Chris Dent authored
* Better jsonpath error messages
-
Chris Dent authored
When the YAML files uses JSONpaths that do not or cannot match we need to know which patch it was that didn't match, otherwise we just get a rather lame error message that is not useful information for fixing the problem.
-
- 08 Jan, 2015 3 commits
-
-
Chris Dent authored
* regexp matches for response headers
-
Chris Dent authored
Returning the / / makes the regex stand out better.
-
Chris Dent authored
If the expected value for a response_header is wrapped in '/' the value is treated as a regular expression and passed to `testtools.TestCase.assertRegexpMatches`. Fixes #6
-
- 06 Jan, 2015 2 commits
-
-
Chris Dent authored
-
Chris Dent authored
This is rather limited for now, as fixtures are not yet sufficiently fleshed out to have a clear idea of how they are going to work.
-
- 05 Jan, 2015 1 commit
-
-
Chris Dent authored
-
- 04 Jan, 2015 1 commit
-
-
Chris Dent authored
It just makes more sense. It's likely there will need some tuneups with regard to how exceptions are handled, once some real fixtures exist.
-
- 01 Jan, 2015 3 commits
-
-
Chris Dent authored
-
Chris Dent authored
-
Chris Dent authored
data can handle a string, a string indicating a file, and a data structure.
-