From schmir at gmail.com  Wed Jun  3 15:33:23 2009
From: schmir at gmail.com (Ralf Schmitt)
Date: Wed, 3 Jun 2009 15:33:23 +0200
Subject: [py-dev] pytest_twisted vs generator based tests
Message-ID: <932f8baf0906030633u3eaef4dk29753b78e44933d2@mail.gmail.com>

Hi,

generator based tests do not work with the pytest_twisted plugin (at
least if you want to pass arguments)

http://bitbucket.org/schmir/py-trunk/changeset/20e64b3893a0/
contains a fix. I'm not quite sure if is correct

I've also fixed some typos in comments:
http://bitbucket.org/schmir/py-trunk/changeset/4796f75ef7c5/

Regards,
- Ralf


From schmir at gmail.com  Fri Jun  5 11:18:51 2009
From: schmir at gmail.com (Ralf Schmitt)
Date: Fri, 5 Jun 2009 11:18:51 +0200
Subject: [py-dev] pytest_twisted vs generator based tests
In-Reply-To: <932f8baf0906030633u3eaef4dk29753b78e44933d2@mail.gmail.com>
References: <932f8baf0906030633u3eaef4dk29753b78e44933d2@mail.gmail.com>
Message-ID: <932f8baf0906050218gc33f338h94df183ebcca1b5e@mail.gmail.com>

On Wed, Jun 3, 2009 at 3:33 PM, Ralf Schmitt<schmir at gmail.com> wrote:
> Hi,
>
> generator based tests do not work with the pytest_twisted plugin (at
> least if you want to pass arguments)

apparently I've been running the pytest_twisted plugin with a slightly
older py lib version.
it looks like the plugin architecture has been changed a bit, and
pytest_twisted must still be adapted to these changes.
I'll try to do that...

Regards,
- Ralf


From schmir at gmail.com  Fri Jun  5 12:14:31 2009
From: schmir at gmail.com (Ralf Schmitt)
Date: Fri, 5 Jun 2009 12:14:31 +0200
Subject: [py-dev] pytest_twisted vs generator based tests
In-Reply-To: <932f8baf0906050218gc33f338h94df183ebcca1b5e@mail.gmail.com>
References: <932f8baf0906030633u3eaef4dk29753b78e44933d2@mail.gmail.com>
	<932f8baf0906050218gc33f338h94df183ebcca1b5e@mail.gmail.com>
Message-ID: <932f8baf0906050314g2575f690yddfb5cf18320be95@mail.gmail.com>

On Fri, Jun 5, 2009 at 11:18 AM, Ralf Schmitt<schmir at gmail.com> wrote:
> apparently I've been running the pytest_twisted plugin with a slightly
> older py lib version.
> it looks like the plugin architecture has been changed a bit, and
> pytest_twisted must still be adapted to these changes.
> I'll try to do that...
>

I got the pytest_twisted plugin working again:

http://bitbucket.org/schmir/py-trunk/changeset/b65184087a6f/


From holger at merlinux.eu  Mon Jun  8 16:49:53 2009
From: holger at merlinux.eu (holger krekel)
Date: Mon, 8 Jun 2009 16:49:53 +0200
Subject: [py-dev] pytest_twisted vs generator based tests
In-Reply-To: <932f8baf0906050314g2575f690yddfb5cf18320be95@mail.gmail.com>
References: <932f8baf0906030633u3eaef4dk29753b78e44933d2@mail.gmail.com>
	<932f8baf0906050218gc33f338h94df183ebcca1b5e@mail.gmail.com>
	<932f8baf0906050314g2575f690yddfb5cf18320be95@mail.gmail.com>
Message-ID: <20090608144953.GQ7437@trillke.net>

Hi Ralf, 

On Fri, Jun 05, 2009 at 12:14 +0200, Ralf Schmitt wrote:
> On Fri, Jun 5, 2009 at 11:18 AM, Ralf Schmitt<schmir at gmail.com> wrote:
> > apparently I've been running the pytest_twisted plugin with a slightly
> > older py lib version.
> > it looks like the plugin architecture has been changed a bit, and
> > pytest_twisted must still be adapted to these changes.
> > I'll try to do that...
> >
> 
> I got the pytest_twisted plugin working again:
> 
> http://bitbucket.org/schmir/py-trunk/changeset/b65184087a6f/

thanks.  I reviewed and pulled your changes.  I think it makes
sense to also pass on kwargs (also called "funcargs" in the docs). 

Btw, I installed twisted and greenlets and can run the test 
that is contained in the __init__ file with "py.test ../__init__.py".
Probably it'd just be fine to have have a pytest_twisted.py file
instead of a package.   Feel free to ping me for more changes. 

best,
holger


From simon at arrowtheory.com  Tue Jun 16 22:27:34 2009
From: simon at arrowtheory.com (Simon Burton)
Date: Tue, 16 Jun 2009 16:27:34 -0400
Subject: [py-dev] py.test TypeError
Message-ID: <20090616162734.5fc29213.simon@arrowtheory.com>

____________________________________________________________________ FAILURES _____________________________________________________________________

self = <CollectorRunner colitem=<Instance '()'>>

    def run(self):
        """ return result of running setup, execution, teardown procedures. """
        excinfo = None
        res = NORESULT
        capture = self.getcapture()
        try:
            try:
                when = "setup"
                self.setup()
                try:
>                   res = self.execute()

/home/simon/local/pypy/py/test/runner.py:37: 
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

self = <CollectorRunner colitem=<Instance '()'>>

    def execute(self):
>       return self.colitem._memocollect()

/home/simon/local/pypy/py/test/runner.py:76: 
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

self = <Instance '()'>

    def _memocollect(self):
        """ internal helper method to cache results of calling collect(). """
>       return self._memoizedcall('_collected', self.collect)

/home/simon/local/pypy/py/test/collect.py:350: 
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

self = <Instance '()'>, attrname = '_collected', function = <bound method Instance.collect of <Instance '()'>>

    def _memoizedcall(self, attrname, function):
        exattrname = "_ex_" + attrname
        failure = getattr(self, exattrname, None)
        if failure is not None:
            raise failure[0], failure[1], failure[2]
        if hasattr(self, attrname):
            return getattr(self, attrname)
        try:
>           res = function()

/home/simon/local/pypy/py/test/collect.py:146: 
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

self = <Instance '()'>

    def collect(self):
        l = self._deprecated_collect()
        if l is not None:
            return l
>       name2items = self._buildname2items()

/home/simon/local/pypy/py/test/pycollect.py:115: 
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

self = <Instance '()'>

    def _buildname2items(self):
        # NB. we avoid random getattrs and peek in the __dict__ instead
        d = {}
>       dicts = [getattr(self.obj, '__dict__', {})]

/home/simon/local/pypy/py/test/pycollect.py:123: 
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

self = <Instance '()'>

    def fget(self):
        try:
            return self._obj
        except AttributeError:
>           self._obj = obj = self._getobj()

/home/simon/local/pypy/py/test/pycollect.py:30: 
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

self = <Instance '()'>

    def _getobj(self):
>       return self.parent.obj()
E       TypeError: function takes exactly 1 argument (0 given)

/home/simon/local/pypy/py/test/pycollect.py:219: TypeError


From holger at merlinux.eu  Wed Jun 17 21:00:06 2009
From: holger at merlinux.eu (holger krekel)
Date: Wed, 17 Jun 2009 21:00:06 +0200
Subject: [py-dev] py.test TypeError
In-Reply-To: <20090616162734.5fc29213.simon@arrowtheory.com>
References: <20090616162734.5fc29213.simon@arrowtheory.com>
Message-ID: <20090617190006.GE7437@trillke.net>

Hi Simon,

can you post a bit more context like the py lib version the test 
file/function this refers to? 

holger

On Tue, Jun 16, 2009 at 16:27 -0400, Simon Burton wrote:
> ____________________________________________________________________ FAILURES _____________________________________________________________________
> 
> self = <CollectorRunner colitem=<Instance '()'>>
> 
>     def run(self):
>         """ return result of running setup, execution, teardown procedures. """
>         excinfo = None
>         res = NORESULT
>         capture = self.getcapture()
>         try:
>             try:
>                 when = "setup"
>                 self.setup()
>                 try:
> >                   res = self.execute()
> 
> /home/simon/local/pypy/py/test/runner.py:37: 
> _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
> 
> self = <CollectorRunner colitem=<Instance '()'>>
> 
>     def execute(self):
> >       return self.colitem._memocollect()
> 
> /home/simon/local/pypy/py/test/runner.py:76: 
> _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
> 
> self = <Instance '()'>
> 
>     def _memocollect(self):
>         """ internal helper method to cache results of calling collect(). """
> >       return self._memoizedcall('_collected', self.collect)
> 
> /home/simon/local/pypy/py/test/collect.py:350: 
> _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
> 
> self = <Instance '()'>, attrname = '_collected', function = <bound method Instance.collect of <Instance '()'>>
> 
>     def _memoizedcall(self, attrname, function):
>         exattrname = "_ex_" + attrname
>         failure = getattr(self, exattrname, None)
>         if failure is not None:
>             raise failure[0], failure[1], failure[2]
>         if hasattr(self, attrname):
>             return getattr(self, attrname)
>         try:
> >           res = function()
> 
> /home/simon/local/pypy/py/test/collect.py:146: 
> _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
> 
> self = <Instance '()'>
> 
>     def collect(self):
>         l = self._deprecated_collect()
>         if l is not None:
>             return l
> >       name2items = self._buildname2items()
> 
> /home/simon/local/pypy/py/test/pycollect.py:115: 
> _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
> 
> self = <Instance '()'>
> 
>     def _buildname2items(self):
>         # NB. we avoid random getattrs and peek in the __dict__ instead
>         d = {}
> >       dicts = [getattr(self.obj, '__dict__', {})]
> 
> /home/simon/local/pypy/py/test/pycollect.py:123: 
> _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
> 
> self = <Instance '()'>
> 
>     def fget(self):
>         try:
>             return self._obj
>         except AttributeError:
> >           self._obj = obj = self._getobj()
> 
> /home/simon/local/pypy/py/test/pycollect.py:30: 
> _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
> 
> self = <Instance '()'>
> 
>     def _getobj(self):
> >       return self.parent.obj()
> E       TypeError: function takes exactly 1 argument (0 given)
> 
> /home/simon/local/pypy/py/test/pycollect.py:219: TypeError
> _______________________________________________
> py-dev mailing list
> py-dev at codespeak.net
> http://codespeak.net/mailman/listinfo/py-dev
> 

-- 
Metaprogramming, Python, Testing: http://tetamap.wordpress.com
Python, PyPy, pytest contracting: http://merlinux.eu 


From simon at arrowtheory.com  Thu Jun 18 00:35:00 2009
From: simon at arrowtheory.com (Simon Burton)
Date: Wed, 17 Jun 2009 18:35:00 -0400
Subject: [py-dev] py.test TypeError
In-Reply-To: <20090617190006.GE7437@trillke.net>
References: <20090616162734.5fc29213.simon@arrowtheory.com>
	<20090617190006.GE7437@trillke.net>
Message-ID: <20090617183500.917c49d3.simon@arrowtheory.com>


It was failing for an old version of py lib, and then i ran top of tree
and that fails too in the same way. I'm sorry I can't help with the test
file (it is a large proprietary blob).

On Wed, 17 Jun 2009 21:00:06 +0200
holger krekel <holger at merlinux.eu> wrote:

> 
> Hi Simon,
> 
> can you post a bit more context like the py lib version the test 
> file/function this refers to? 
> 
> holger


From holger at merlinux.eu  Fri Jun 19 13:21:28 2009
From: holger at merlinux.eu (holger krekel)
Date: Fri, 19 Jun 2009 13:21:28 +0200
Subject: [py-dev] 1.0.0b3 released
Message-ID: <20090619112128.GI7437@trillke.net>

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 


From holger at merlinux.eu  Fri Jun 19 13:37:48 2009
From: holger at merlinux.eu (holger krekel)
Date: Fri, 19 Jun 2009 13:37:48 +0200
Subject: [py-dev] 1.0.0b3 released
In-Reply-To: <20090619112128.GI7437@trillke.net>
References: <20090619112128.GI7437@trillke.net>
Message-ID: <20090619113748.GK7437@trillke.net>

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 at codespeak.net
> http://codespeak.net/mailman/listinfo/py-dev
> 

-- 
Metaprogramming, Python, Testing: http://tetamap.wordpress.com
Python, PyPy, pytest contracting: http://merlinux.eu 


From schmir at gmail.com  Mon Jun 22 18:42:11 2009
From: schmir at gmail.com (Ralf Schmitt)
Date: Mon, 22 Jun 2009 18:42:11 +0200
Subject: [py-dev] 1.0.0b3 released
In-Reply-To: <20090619112128.GI7437@trillke.net>
References: <20090619112128.GI7437@trillke.net>
Message-ID: <932f8baf0906220942kce50f39ncdaa2ed75b2822f2@mail.gmail.com>

On Fri, Jun 19, 2009 at 1:21 PM, holger krekel<holger at 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


From holger at merlinux.eu  Tue Jun 23 10:29:57 2009
From: holger at merlinux.eu (holger krekel)
Date: Tue, 23 Jun 2009 10:29:57 +0200
Subject: [py-dev] 1.0.0b3 released
In-Reply-To: <932f8baf0906220942kce50f39ncdaa2ed75b2822f2@mail.gmail.com>
References: <20090619112128.GI7437@trillke.net>
	<932f8baf0906220942kce50f39ncdaa2ed75b2822f2@mail.gmail.com>
Message-ID: <20090623082957.GM7437@trillke.net>

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 at 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


From schmir at gmail.com  Tue Jun 23 12:40:42 2009
From: schmir at gmail.com (Ralf Schmitt)
Date: Tue, 23 Jun 2009 12:40:42 +0200
Subject: [py-dev] 1.0.0b3 released
In-Reply-To: <20090623082957.GM7437@trillke.net>
References: <20090619112128.GI7437@trillke.net>
	<932f8baf0906220942kce50f39ncdaa2ed75b2822f2@mail.gmail.com>
	<20090623082957.GM7437@trillke.net>
Message-ID: <932f8baf0906230340q1027d2a8id6f16dd380ef53a0@mail.gmail.com>

On Tue, Jun 23, 2009 at 10:29 AM, holger krekel<holger at 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 at 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


From holger at merlinux.eu  Tue Jun 23 15:12:33 2009
From: holger at merlinux.eu (holger krekel)
Date: Tue, 23 Jun 2009 15:12:33 +0200
Subject: [py-dev] 1.0.0b3 released
In-Reply-To: <932f8baf0906230340q1027d2a8id6f16dd380ef53a0@mail.gmail.com>
References: <20090619112128.GI7437@trillke.net>
	<932f8baf0906220942kce50f39ncdaa2ed75b2822f2@mail.gmail.com>
	<20090623082957.GM7437@trillke.net>
	<932f8baf0906230340q1027d2a8id6f16dd380ef53a0@mail.gmail.com>
Message-ID: <20090623131233.GP7437@trillke.net>

On Tue, Jun 23, 2009 at 12:40 +0200, Ralf Schmitt wrote:
> On Tue, Jun 23, 2009 at 10:29 AM, holger krekel<holger at 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 at 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


From schmir at gmail.com  Wed Jun 24 00:02:29 2009
From: schmir at gmail.com (Ralf Schmitt)
Date: Wed, 24 Jun 2009 00:02:29 +0200
Subject: [py-dev] 1.0.0b3 released
In-Reply-To: <20090623131233.GP7437@trillke.net>
References: <20090619112128.GI7437@trillke.net>
	<932f8baf0906220942kce50f39ncdaa2ed75b2822f2@mail.gmail.com>
	<20090623082957.GM7437@trillke.net>
	<932f8baf0906230340q1027d2a8id6f16dd380ef53a0@mail.gmail.com>
	<20090623131233.GP7437@trillke.net>
Message-ID: <932f8baf0906231502w178befb8i2b0ac7cc046006af@mail.gmail.com>

On Tue, Jun 23, 2009 at 3:12 PM, holger krekel<holger at 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


From holger at merlinux.eu  Wed Jun 24 18:22:54 2009
From: holger at merlinux.eu (holger krekel)
Date: Wed, 24 Jun 2009 18:22:54 +0200
Subject: [py-dev] 1.0.0b5 with some fixes uploaded
In-Reply-To: <20090619112128.GI7437@trillke.net>
References: <20090619112128.GI7437@trillke.net>
Message-ID: <20090624162254.GS7437@trillke.net>

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 


From sridharr at activestate.com  Thu Jun 25 21:47:02 2009
From: sridharr at activestate.com (Sridhar Ratnakumar)
Date: Thu, 25 Jun 2009 12:47:02 -0700
Subject: [py-dev] py.test - suppressing stdout
Message-ID: <4A43D436.7050102@activestate.com>

In py.test, is there a way to suppress stdout at all?

In my program, if a test fails .. py.test prints a huge amount of text 
(that the program prints to stdout) .. and I had to scroll back a lot to 
see the traceback. This is very inconvenient. In most case, I just want 
to see the traceback along with locals values.



From holger at merlinux.eu  Thu Jun 25 22:42:05 2009
From: holger at merlinux.eu (holger krekel)
Date: Thu, 25 Jun 2009 22:42:05 +0200
Subject: [py-dev] py.test - suppressing stdout
In-Reply-To: <4A43D436.7050102@activestate.com>
References: <4A43D436.7050102@activestate.com>
Message-ID: <20090625204204.GT7437@trillke.net>

Hi Sridhar, 

On Thu, Jun 25, 2009 at 12:47 -0700, Sridhar Ratnakumar wrote:
> In py.test, is there a way to suppress stdout at all?
> 
> In my program, if a test fails .. py.test prints a huge amount of text 
> (that the program prints to stdout) .. and I had to scroll back a lot to 
> see the traceback. This is very inconvenient. In most case, I just want 
> to see the traceback along with locals values.

yes, makes sense - i had this need myself as well, actually.
I'd like to have this and other report customizations be
doable through a --report option (and thus also through ENV
vars and conftest.py settings). 

For now, as a kind of workaround, you could put the following lines 
into a conftest plugin: 

    # example content of conftest.py 
    import py

    def pytest_runtest_call(__call__, item):
        cap = py.io.StdCapture()
        try:
            return __call__.execute() # call all other hook implementations
        finally:
            outerr = cap.reset() # forget about capture

if you run your tests now, all capturing should be dropped.
does it work for you and is the effect roughly what you want? 
best,

holger

 
> _______________________________________________
> py-dev mailing list
> py-dev at codespeak.net
> http://codespeak.net/mailman/listinfo/py-dev
> 

-- 
Metaprogramming, Python, Testing: http://tetamap.wordpress.com
Python, PyPy, pytest contracting: http://merlinux.eu 


From holger at merlinux.eu  Mon Jun 29 13:01:29 2009
From: holger at merlinux.eu (holger krekel)
Date: Mon, 29 Jun 2009 13:01:29 +0200
Subject: [py-dev] 1.0.0b6 and xfail/py.test.mark
Message-ID: <20090629110129.GC7437@trillke.net>

Hi all, 

i just released a 1.0.0b6 which is hopefully
the last beta.  Main change to B5: 

* py.test.xfail is now py.test.mark.xfail 

* and py.test.mark is back for setting keywords
  on functions as documented. 

Andreas Kloeckner pointed me to the fact that py.test.mark
*was* documented and he had provided (for b5) a patch which 
is also in that makes it work correctly with "py.test -k". 
see http://tinyurl.com/m6wc3b


we apologize for any inconvenience. 
(this is the london underground).

holger

-- 
Metaprogramming, Python, Testing: http://tetamap.wordpress.com
Python, PyPy, pytest contracting: http://merlinux.eu 


From SridharR at activestate.com  Mon Jun 29 23:08:14 2009
From: SridharR at activestate.com (Sridhar Ratnakumar)
Date: Mon, 29 Jun 2009 14:08:14 -0700
Subject: [py-dev] py.test - suppressing stdout
In-Reply-To: <20090625204204.GT7437@trillke.net>
References: <4A43D436.7050102@activestate.com>
	<20090625204204.GT7437@trillke.net>
Message-ID: <op.uwa0z0t6brrvlq@double>

On Thu, 25 Jun 2009 13:42:05 -0700, holger krekel <holger at merlinux.eu>  
wrote:

> On Thu, Jun 25, 2009 at 12:47 -0700, Sridhar Ratnakumar wrote:
>> In py.test, is there a way to suppress stdout at all?
>>
>> In my program, if a test fails .. py.test prints a huge amount of text
>> (that the program prints to stdout) .. and I had to scroll back a lot to
>> see the traceback. This is very inconvenient. In most case, I just want
>> to see the traceback along with locals values.
> yes, makes sense - i had this need myself as well, actually.
> I'd like to have this and other report customizations be
> doable through a --report option (and thus also through ENV
> vars and conftest.py settings).
> For now, as a kind of workaround, you could put the following lines
> into a conftest plugin:
>    # example content of conftest.py
>     import py
>    def pytest_runtest_call(__call__, item):
>         cap = py.io.StdCapture()
>         try:
>             return __call__.execute() # call all other hook  
> implementations
>         finally:
>             outerr = cap.reset() # forget about capture
> if you run your tests now, all capturing should be dropped.
> does it work for you and is the effect roughly what you want?

Yes, this does the trick for me (although I would probably modify the code  
to log stdout to a file insead).

Thanks.

-srid