[Distutils] [PATCH] catch egg_base=./ foolishness ? was (distutils data_files and setuptools.pkg_resources are driving me crazy)

Bob Ippolito bob at redivi.com
Mon Jul 17 23:42:54 CEST 2006


On Jul 17, 2006, at 2:30 PM, Phillip J. Eby wrote:

> At 10:06 PM 7/15/2006 +0100, Robin Bryce wrote:
>> On 15/07/06, Robin Bryce <robinbryce at gmail.com> wrote:
>>> egg_base=./ seemed to prevent the installation of console_scripts.
>>
>> Oh fiddle sticks. this is because './' is not equal to os.curdir on
>> *any* system.
>>
>> line 103: egg_info.py does the right thing: if self.egg_base !=
>> os.curdir. This mirrors exactly how distutils.filelist.findall does
>> the same gaurd on files in the current directory.
>>
>> Observing that:
>> os.path.join('.' + os.sep, '') == os.path.join(os.curdir, '') and
>> os.path.join('.', '') == os.path.join(os.curdir, '')
>>
>> how about this patch::
>
> I don't see the point of the patch.  egg_base is supposed to be a  
> directory
> name.  './' is not a directory name, and I don't see a use case for  
> setting
> egg_base to that.  In fact, 99% or more of the time, there's no use  
> case
> for setting --egg-base in the first place.  The option was created  
> mainly
> to support certain kinds of workaround scenarios in the early days of
> setuptools development.  I would deprecate the option, except that it
> doesn't really do any harm, unless somebody messes around with it  
> on the
> theory that it does more than what it actually does.  :)

FYI egg_base what we're using on the PyObjC trunk [1] in order to  
facilitate "python setup.py develop" support without re-organizing  
everything. All of the top-level Python packages are in a Lib  
directory, so we set egg_base="Lib".

[1] http://svn.red-bean.com/pyobjc/trunk/pyobjc/

-bob



More information about the Distutils-SIG mailing list