Hello all, i just uploaded a 1.0.0b3 release up to PyPI. This is still The reason being that installing command line scripts on But there is no deep dependency - i.e. you can easily tweak the setup.py file to run without setuptools. What's new with 1.0.0b3, compared to 1.0.0b1? * much improved py.test documentation (http://pytest.org) * Plugins now define hooks directly instead of in a class * there is a new pytest_generate_tests test parametrization hook, see my blog Hope things work for everyone - please let me know of any problems. Next week i'd like to do a 1.0 final. thanks & best, holger -- Metaprogramming, Python, Testing: http://tetamap.wordpress.com Python, PyPy, pytest contracting: http://merlinux.eu
n Fri, Jun 19, 2009 at 13:21 +0200, holger krekel wrote:
Hello all,
i just uploaded a 1.0.0b3 release up to PyPI. This is still The reason being that installing command line scripts on
editing accident. what i meant is: i just uploaded a 1.0.0b3 release to PyPI. This is still setuptools based because of some issues with installing command line scripts on win32.
But there is no deep dependency - i.e. you can easily tweak the setup.py file to run without setuptools.
What's new with 1.0.0b3, compared to 1.0.0b1?
* much improved py.test documentation (http://pytest.org) * Plugins now define hooks directly instead of in a class * there is a new pytest_generate_tests test parametrization hook, see my blog
Hope things work for everyone - please let me know of any problems. Next week i'd like to do a 1.0 final.
thanks & best, holger
-- Metaprogramming, Python, Testing: http://tetamap.wordpress.com Python, PyPy, pytest contracting: http://merlinux.eu _______________________________________________ py-dev mailing list py-dev@codespeak.net http://codespeak.net/mailman/listinfo/py-dev
-- Metaprogramming, Python, Testing: http://tetamap.wordpress.com Python, PyPy, pytest contracting: http://merlinux.eu
On Fri, Jun 19, 2009 at 1:21 PM, holger krekel<holger@merlinux.eu> wrote:
Hope things work for everyone - please let me know of any problems. Next week i'd like to do a 1.0 final.
py.test.mark isn't there: xfail = py.test.mark.xfail("expected failure") E AttributeError: mark Is this intended? Regards, - Ralf
Hi Ralf, On Mon, Jun 22, 2009 at 18:42 +0200, Ralf Schmitt wrote:
On Fri, Jun 19, 2009 at 1:21 PM, holger krekel<holger@merlinux.eu> wrote:
Hope things work for everyone - please let me know of any problems. Next week i'd like to do a 1.0 final.
py.test.mark isn't there:
xfail = py.test.mark.xfail("expected failure") E AttributeError: mark
Is this intended?
it's now py.test.xfail("expected failure"). Did you find some rogue link/documentation about this Or was it just that i missed pointing that out in the CHANGELOG? FYI, plugins can implement the pytest_namespace(config) hook returning a dict that will thus extend the py.test.* namespace. best, holger
On Tue, Jun 23, 2009 at 10:29 AM, holger krekel<holger@merlinux.eu> wrote:
Hi Ralf,
On Mon, Jun 22, 2009 at 18:42 +0200, Ralf Schmitt wrote:
On Fri, Jun 19, 2009 at 1:21 PM, holger krekel<holger@merlinux.eu> wrote:
Hope things work for everyone - please let me know of any problems. Next week i'd like to do a 1.0 final.
py.test.mark isn't there:
xfail = py.test.mark.xfail("expected failure") E AttributeError: mark
Is this intended?
it's now py.test.xfail("expected failure"). Did you find some rogue link/documentation about this Or was it just that i missed pointing that out in the CHANGELOG?
I didn't even notice there is a CHANGELOG :) Anyway, this only gives me py.test.xfail and not mark. BTW, py.test.xfail doesn't work for me too; In [2]: py.test.xfail --------------------------------------------------------------------------- AttributeError Traceback (most recent call last) /home/ralf/strtotime/<ipython console> in <module>() /home/ralf/py26/lib/python2.6/site-packages/py-1.0.0b3-py2.6.egg/py/initpkg.pyc in __getattr__(self, name) 182 except KeyError: 183 __tracebackhide__ = True --> 184 raise AttributeError(name) 185 else: 186 result = self.__pkg__._resolve(extpy) AttributeError: xfail
FYI, plugins can implement the
pytest_namespace(config)
hook returning a dict that will thus extend the py.test.* namespace.
nice. Regards, - Ralf
On Tue, Jun 23, 2009 at 12:40 +0200, Ralf Schmitt wrote:
On Tue, Jun 23, 2009 at 10:29 AM, holger krekel<holger@merlinux.eu> wrote:
Hi Ralf,
On Mon, Jun 22, 2009 at 18:42 +0200, Ralf Schmitt wrote:
On Fri, Jun 19, 2009 at 1:21 PM, holger krekel<holger@merlinux.eu> wrote:
Hope things work for everyone - please let me know of any problems. Next week i'd like to do a 1.0 final.
py.test.mark isn't there:
xfail = py.test.mark.xfail("expected failure") E AttributeError: mark
Is this intended?
it's now py.test.xfail("expected failure"). Did you find some rogue link/documentation about this Or was it just that i missed pointing that out in the CHANGELOG?
I didn't even notice there is a CHANGELOG :)
Anyway, this only gives me py.test.xfail and not mark.
1. mark is not there, indeed. We could do a plugin that provides py.test.mark - as it was undocumented and not much advertised yet so i didn't think there would be users :) 2. py.test.xfail does not take arguments anymore - see http://tinyurl.com/nxqta2 I basically stripped the py.test.xfail decorator to the bare minimum of marking a function as expected-to-fail. do you have a need for it to take a message argument?
BTW, py.test.xfail doesn't work for me too;
In [2]: py.test.xfail --------------------------------------------------------------------------- AttributeError Traceback (most recent call last)
/home/ralf/strtotime/<ipython console> in <module>()
/home/ralf/py26/lib/python2.6/site-packages/py-1.0.0b3-py2.6.egg/py/initpkg.pyc in __getattr__(self, name) 182 except KeyError: 183 __tracebackhide__ = True --> 184 raise AttributeError(name) 185 else: 186 result = self.__pkg__._resolve(extpy)
AttributeError: xfail
Hum, this happens in an interactive session because plugins are not activated and there is no py.test.xfail entry. If you issue
py.test.config.parse([]) py.test.config.pluginmanager.do_configure(py.test.config)
the default plugins will be activated and py.test.xfail becomes visible. Would you like this incantation to live in a "py.test.init()" helper?!
FYI, plugins can implement the
pytest_namespace(config)
hook returning a dict that will thus extend the py.test.* namespace.
nice.
with the above "interactive" caveat, though. holger
On Tue, Jun 23, 2009 at 3:12 PM, holger krekel<holger@merlinux.eu> wrote:
Anyway, this only gives me py.test.xfail and not mark.
1. mark is not there, indeed. We could do a plugin that provides py.test.mark - as it was undocumented and not much advertised yet so i didn't think there would be users :)
I didn't actually use it. so feel free to drop it.
2. py.test.xfail does not take arguments anymore - see http://tinyurl.com/nxqta2 I basically stripped the py.test.xfail decorator to the bare minimum of marking a function as expected-to-fail. do you have a need for it to take a message argument?
no.
BTW, py.test.xfail doesn't work for me too;
Hum, this happens in an interactive session because plugins are not activated and there is no py.test.xfail entry. If you issue
py.test.config.parse([]) py.test.config.pluginmanager.do_configure(py.test.config)
the default plugins will be activated and py.test.xfail becomes visible. Would you like this incantation to live in a "py.test.init()" helper?!
no, the interactive session was just a quick test... regards, - ralf
Hi again, FYI i addressed a few issues from the tracker and uploaded a new 1.0.0b5 release: * remove scope-argument from request.addfinalizer() because request.cached_setup has the scope arg. TOOWTDI. * perform setup finalization before reporting failures * apply modified patches from Andreas Kloeckner to allow test functions to have no func_code (#22) and to make "-k" and function keywords work (#20) * apply isatty patch from Daniel Peolzleithner (issue #23) * resolve issue #18, multiprocessing.Manager() and redirection clash best, holger On Fri, Jun 19, 2009 at 13:21 +0200, holger krekel wrote:
Hello all,
i just uploaded a 1.0.0b3 release up to PyPI. This is still The reason being that installing command line scripts on But there is no deep dependency - i.e. you can easily tweak the setup.py file to run without setuptools.
What's new with 1.0.0b3, compared to 1.0.0b1?
* much improved py.test documentation (http://pytest.org) * Plugins now define hooks directly instead of in a class * there is a new pytest_generate_tests test parametrization hook, see my blog
Hope things work for everyone - please let me know of any problems. Next week i'd like to do a 1.0 final.
thanks & best, holger
-- Metaprogramming, Python, Testing: http://tetamap.wordpress.com Python, PyPy, pytest contracting: http://merlinux.eu
participants (2)
-
holger krekel -
Ralf Schmitt