Re: [Distutils] Limit package installation to a specific Python version

[Matthias Bussonnier]

Hi all,

On Fri, Jan 20, 2017 at 2:56 PM, Nathaniel Smith  wrote:
> On Fri, Jan 20, 2017 at 1:56 PM, Lele Gaifax  wrote:
>> Hi all,
>>
>> do installers like pip and conda consult trove classifiers, or more generally
>> is there a way to "mark" a package published on PyPI as installable only in a
>> Python 3 environment?
>>

>  For sdists, historically there was no way to do this, but as of a
> few months ago this was fixed and pip now *does* honor
> python_requires, and PyPI has been updated to expose this information
> so pip can see it before downloading the sdists:
>
> https://github.com/pypa/pip/pull/3846
> https://github.com/python/peps/pull/56
> https://github.com/pypa/pip/pull/3877
>
> Since this is a pretty recent change, you probably want to put a
> runtime check inside your setup.py for old Python versions, so if
> someone uses an old version of pip then they will at least get a nice
> explanation.
>
> IPython's an example of a python-3-only project that uses this
> belt-and-suspenders approach:
>
> https://github.com/ipython/ipython/blob/f060158665189fd456e54a5a864080efe28b4fdf/setup.py#L29
> https://github.com/ipython/ipython/blob/f060158665189fd456e54a5a864080efe28b4fdf/setup.py#L248
>

I'm also in the (slow) process of writing the various step you can
take when python_requires is not honored by older systems and how to
mitigate that:

https://github.com/python3statement/python3statement.github.io/pull/55

Feedback (and help) would be welcome.

In particular only system where pip 9.0+ is installed will understand
the metadata now exposed by PyPI,
so when/if you fail in your setup.py, a useful tip is to ask user to
upgrade pip.

There should be some snipets of code you copy and past in your
setup.py in this PR as well.

Thanks.
-- 
M
___
Distutils-SIG maillist  -  Distutils-SIG@python.org
https://mail.python.org/mailman/listinfo/distutils-sig

Re: [Distutils] Limit package installation to a specific Python version

[Lele Gaifax]

Great, thank you Nathaniel!
-- 
nickname: Lele Gaifax | Quando vivrò di quello che ho pensato ieri
real: Emanuele Gaifas | comincerò ad aver paura di chi mi copia.
l...@metapensiero.it  | -- Fortunato Depero, 1929.

___
Distutils-SIG maillist  -  Distutils-SIG@python.org
https://mail.python.org/mailman/listinfo/distutils-sig

Re: [Distutils] Limit package installation to a specific Python version

[Nathaniel Smith]

On Fri, Jan 20, 2017 at 1:56 PM, Lele Gaifax  wrote:
> Hi all,
>
> do installers like pip and conda consult trove classifiers, or more generally
> is there a way to "mark" a package published on PyPI as installable only in a
> Python 3 environment?
>
> An user proposed[1] to change the classifiers of a package to get that result,
> so I tried looking at pip and setuptools sources, but didn't find evidence of
> that. I see both use python_requires[2] in their setup that seem closer to the
> goal, but again, is that actually honored by pip and/or other installers?
>
> Thanks for any hint,
> ciao, lele.
>
> [1] https://github.com/python-rapidjson/python-rapidjson/pull/47
> [2] https://www.python.org/dev/peps/pep-0345/#requires-python

Neither pip nor conda do anything automatically with trove
classifiers. Conda uses its own metadata added when the conda package
is built; whoever runs your conda-build can probably figure something
out. Wheels have always had this metadata and pip has always honored
it. For sdists, historically there was no way to do this, but as of a
few months ago this was fixed and pip now *does* honor
python_requires, and PyPI has been updated to expose this information
so pip can see it before downloading the sdists:

https://github.com/pypa/pip/pull/3846
https://github.com/python/peps/pull/56
https://github.com/pypa/pip/pull/3877

Since this is a pretty recent change, you probably want to put a
runtime check inside your setup.py for old Python versions, so if
someone uses an old version of pip then they will at least get a nice
explanation.

IPython's an example of a python-3-only project that uses this
belt-and-suspenders approach:

https://github.com/ipython/ipython/blob/f060158665189fd456e54a5a864080efe28b4fdf/setup.py#L29
https://github.com/ipython/ipython/blob/f060158665189fd456e54a5a864080efe28b4fdf/setup.py#L248

-n

-- 
Nathaniel J. Smith -- https://vorpus.org
___
Distutils-SIG maillist  -  Distutils-SIG@python.org
https://mail.python.org/mailman/listinfo/distutils-sig

[Distutils] Limit package installation to a specific Python version

[Lele Gaifax]

Hi all,

do installers like pip and conda consult trove classifiers, or more generally
is there a way to "mark" a package published on PyPI as installable only in a
Python 3 environment?

An user proposed[1] to change the classifiers of a package to get that result,
so I tried looking at pip and setuptools sources, but didn't find evidence of
that. I see both use python_requires[2] in their setup that seem closer to the
goal, but again, is that actually honored by pip and/or other installers?

Thanks for any hint,
ciao, lele.

[1] https://github.com/python-rapidjson/python-rapidjson/pull/47
[2] https://www.python.org/dev/peps/pep-0345/#requires-python
-- 
nickname: Lele Gaifax | Quando vivrò di quello che ho pensato ieri
real: Emanuele Gaifas | comincerò ad aver paura di chi mi copia.
l...@metapensiero.it  | -- Fortunato Depero, 1929.

___
Distutils-SIG maillist  -  Distutils-SIG@python.org
https://mail.python.org/mailman/listinfo/distutils-sig

Re: [Distutils] Has anyone looked at bug 29225?

[Elvis Stansvik]

Den 20 jan. 2017 6:32 em skrev "Brett Cannon" :
>
> Oh, over the 26 years of Python's lifetime you would be surprised what
people come to rely on (including broken semantics which they have written
fixes for themselves which then break when a proper fix is introduced).

I can imagine, perhaps I shouldn't have been so confident :) Hopefully if
people have put in their own workarounds, they'll be happy to be able to
drop them, but who knows, heh.

Elvis

>
> On Thu, 19 Jan 2017 at 10:01 Elvis Stansvik 
wrote:
>>
>> 2017-01-19 18:52 GMT+01:00 Brett Cannon :
>>>
>>> It's probably a combination of people being busy and the fact that it
involves distutils which not very many people feel comfortable reviewing
patches for and changing its semantics.
>>
>>
>> Alright, that's understandable and sort of what I suspected.
>>
>> I hope (and believe) my patch preserves the semantics for all cases
except the specific one it fixes. The wrong behavior which it fixes is so
obviously wrong that I can't imagine someone relying on it somehow :)
>>
>> Elvis
>>
>>
>>>
>>>
>>> Regardless of what happens to your patch, though, thanks for taking the
time to report the problem you ran into and trying to fix it!
>>>
>>> On Wed, 18 Jan 2017 at 23:21 Elvis Stansvik <
elvis.stans...@orexplore.com> wrote:

 Hi distutils folks,

 Just wanted to ask if anyone has had time to look at my

 http://bugs.python.org/issue29225

 ? It's my first bug report _and_ first patch for Python, so it's quite
possible I screwed something up. Or people are just busy :)

 Cheers,
 Elvis
 ___
 Distutils-SIG maillist  -  Distutils-SIG@python.org
 https://mail.python.org/mailman/listinfo/distutils-sig
___
Distutils-SIG maillist  -  Distutils-SIG@python.org
https://mail.python.org/mailman/listinfo/distutils-sig

Re: [Distutils] Has anyone looked at bug 29225?

[Brett Cannon]

Oh, over the 26 years of Python's lifetime you would be surprised what
people come to rely on (including broken semantics which they have written
fixes for themselves which then break when a proper fix is introduced).

On Thu, 19 Jan 2017 at 10:01 Elvis Stansvik 
wrote:

> 2017-01-19 18:52 GMT+01:00 Brett Cannon :
>
> It's probably a combination of people being busy and the fact that it
> involves distutils which not very many people feel comfortable reviewing
> patches for and changing its semantics.
>
>
> Alright, that's understandable and sort of what I suspected.
>
> I hope (and believe) my patch preserves the semantics for all cases except
> the specific one it fixes. The wrong behavior which it fixes is so
> obviously wrong that I can't imagine someone relying on it somehow :)
>
> Elvis
>
>
>
>
> Regardless of what happens to your patch, though, thanks for taking the
> time to report the problem you ran into and trying to fix it!
>
> On Wed, 18 Jan 2017 at 23:21 Elvis Stansvik 
> wrote:
>
> Hi distutils folks,
>
> Just wanted to ask if anyone has had time to look at my
>
> http://bugs.python.org/issue29225
>
> ? It's my first bug report _and_ first patch for Python, so it's quite
> possible I screwed something up. Or people are just busy :)
>
> Cheers,
> Elvis
> ___
> Distutils-SIG maillist  -  Distutils-SIG@python.org
> https://mail.python.org/mailman/listinfo/distutils-sig
>
>
___
Distutils-SIG maillist  -  Distutils-SIG@python.org
https://mail.python.org/mailman/listinfo/distutils-sig