[pypy-dev] Re: even more import analysis

[Michael Hudson]

Nicolas Chauvat <nicolas.chauvat at logilab.fr> writes:

> On Fri, Oct 28, 2005 at 10:01:55AM +0100, Michael Hudson wrote:
>> Currently, 'not very' -- in principle, it should work with any python
>> project that doesn't play too sneaky import games.  It had occurred to
>> me that I should probably try to clean things up and make the tool
>> more generally useful, yes :)
>
> I just added to our tracker the following issue for pylint:
>
> coordinate with mwh to see if his import-analysis tool can be
> plugged as an extension/feature to pylint.
>
> it's number #10043 and Sylvain will probably get in touch with you
> next week :)

Nicolas Chauvat <nicolas.chauvat at logilab.fr> writes:

> On Thu, Oct 27, 2005 at 07:39:32PM +0100, Michael Hudson wrote:
>> If you've been reading pypy-svn you'll know I've written an import
>> analysis tool (pypy/tool/importfun.py).  While its implementation is
>> currently completely inscrutable, here's some of its results.
>
> Michael, if you are jumping into code-base analysis of PyPy, I
> suggest you give a try to pylint and maybe define a config file for
> pylint that would be in accordance with pypy's coding standard.

Yeah, if I'd realized what I was getting myself into I'd probably have
looked at pylint or pyflakes a little earlier :) My focus has been
reporting/documenting and maybe, maybe automated refactoring rather
than warning though -- in which light, have you used Bicycle Repair
Man much?


Cheers,
mwh

-- 
  Java is a WORA language! (Write Once, Run Away)
                	-- James Vandenberg (on progstone at egroups.com)
                           & quoted by David Rush on comp.lang.scheme