Merge lp:~chipaca/testscenarios/fix-with-py3 into lp:~testtools-committers/testscenarios/trunk

[John Lenton]

Fix for big justice with python 3.

Fairly straightforward. The only hairy bit is the use of StringIO.StringIO vs io.BytesIO and io.StringIO in the docstring, versus unittest.runner's writing a bare "." to the stream in both python 2 and 3. Resolved with some import shenanigans.

[lp:~chipaca/testscenarios/fix-with-py3]

23. By John Lenton on 2012-05-11

make setup.py work also

[John Vandenberg (jayvdb)]

I think this should be merged. I have tested setup.py and 'make check' with Python 3.2, 3.3 and 3.4.

There are two tests with fail on Python 3, due to the addition of '.<locals>.' in the name of local objects. However that only affects the visual UI, and is arguably correct. I have a separate patch for that; will try to figure out how to upload load that to launchpad.

For Python 3.1.5 (officially retired anyway), unittest2-1.0.1 fails, and testtools depends on that.
NameError: global name 'callable' is not defined

review: Approve

[John Vandenberg (jayvdb)]

Hmm, looking at this more closely, a bit more explanation is needed. This patch is needed to catch up with the released version on https://pypi.python.org/pypi/testscenarios.

pypi lists v0.4 as the current version, and was released on 2013-01-27.

An inspection of the diff between lp:testscenarios and 0.4 indicates that this changeset was essentially included into pypi 0.4, without being merged here in launchpad, and is the only _functional_ difference between lp:testscenarios and pypi 0.4.

So, this code is very mature, having been in the wild for a long time, distributed via pypi.

[Robert Collins (lifeless)]

testscenarios code is maintained on github now, pull requests should be submitted there, bug reports here (preferred) or there.

review: Needs Resubmitting

[Robert Collins (lifeless)]

Huh no sorry, I haven't completed the migration, only intended to. I'll have to check my local tree to see the status.

[Robert Collins (lifeless)]

Huh no sorry, I haven't completed the migration, only intended to. I'll have to check my local tree to see the status.

[John Lenton (chipaca)]

You do realise you're asking me to resubmit a merge proposal from three years ago, yes?

[Robert Collins (lifeless)]

@chipaca sorry, I didn't really process that it was a different John initially, also I was confused about which projects I'd move the code over to github for. I'm not sure what happened with this here MP, since testscenarios is clearly 3 compatible and has been for ages, likely from this code, or dual evolution : EIHAVENOCLUE.

[John Vandenberg (jayvdb)]

I've added a patch to bring lp up to date with pypi: https://code.launchpad.net/~jayvdb/testscenarios/0.4

However, a move to github would be welcome.

As background for reviving this lp project,
I've been building a rather too complex test framework for pywikibot, and want to migrate to a separately maintained framework, and testscenarios looks to be the closest match.<https://phabricator.wikimedia.org/T85899> And I couldnt find it on github...

Any better suggestions would be appreciated.

[Robert Collins (lifeless)]

Code equivalent to this is in trunk.

=== modified file 'README'
--- README	2012-04-04 10:02:46 +0000
+++ README	2012-05-11 07:43:19 +0000
@@ -98,7 +98,10 @@
 ``testscenarios.generate_scenarios()``. For instance::
 
   >>> import unittest
-  >>> import StringIO
+  >>> try:
+  ...     from StringIO import StringIO
+  ... except ImportError:
+  ...     from io import StringIO
   >>> from testscenarios.scenarios import generate_scenarios
 
 This can work with loaders and runners from the standard library, or possibly other
@@ -106,7 +109,7 @@
 
   >>> loader = unittest.TestLoader()
   >>> test_suite = unittest.TestSuite()
-  >>> runner = unittest.TextTestRunner(stream=StringIO.StringIO())
+  >>> runner = unittest.TextTestRunner(stream=StringIO())
 
   >>> mytests = loader.loadTestsFromNames(['doc.test_sample'])
   >>> test_suite.addTests(generate_scenarios(mytests))
@@ -211,7 +214,7 @@
   ...
   >>> suite2 = unittest.TestSuite()
   >>> suite2.addTests(generate_scenarios(suite))
-  >>> print suite2.countTestCases()
+  >>> print(suite2.countTestCases())
   4
 
 Dynamic Scenarios

=== modified file 'lib/testscenarios/scenarios.py'
--- lib/testscenarios/scenarios.py	2012-04-04 10:02:46 +0000
+++ lib/testscenarios/scenarios.py	2012-05-11 07:43:19 +0000
@@ -34,7 +34,7 @@
 from testtools import iterate_tests
 
 
-def apply_scenario((name, parameters), test):
+def apply_scenario(scenario, test):
     """Apply scenario to test.
 
     :param scenario: A tuple (name, parameters) to apply to the test. The test
@@ -43,6 +43,7 @@
     :param test: The test to apply the scenario to. This test is unaltered.
     :return: A new test cloned from test, with the scenario applied.
     """
+    (name, parameters) = scenario
     scenario_suffix = '(' + name + ')'
     newtest = clone_test_with_new_id(test,
         test.id() + scenario_suffix)
@@ -50,7 +51,7 @@
     if test_desc is not None:
         newtest_desc = "%(test_desc)s %(scenario_suffix)s" % vars()
         newtest.shortDescription = (lambda: newtest_desc)
-    for key, value in parameters.iteritems():
+    for key, value in parameters.items():
         setattr(newtest, key, value)
     return newtest
 

=== modified file 'setup.py'
--- setup.py	2012-04-04 10:46:52 +0000
+++ setup.py	2012-05-11 07:43:19 +0000
@@ -3,7 +3,8 @@
 from distutils.core import setup
 import os.path
 
-description = file(os.path.join(os.path.dirname(__file__), 'README'), 'rb').read()
+description = open(os.path.join(os.path.dirname(__file__),
+                                'README'), 'rb').read().decode("utf-8")
 
 setup(name="testscenarios",
       version="0.3",