Re: shlex, unicode, and subprocess.Popen on Windows

2010-08-30 Thread Nobody
On Mon, 30 Aug 2010 10:24:26 -0400, python wrote: >> Kudos for avoiding shell=True > > My understanding is that the only time one needs to use shell=True is > when they are 'executing' a non-executable file whose executable must be > discovered via file association rules? Does that sound accurate

Re: shlex, unicode, and subprocess.Popen on Windows

2010-08-30 Thread Tim Golden
On 30/08/2010 3:24 PM, pyt...@bdurham.com wrote: My understanding is that the only time one needs to use shell=True is when they are 'executing' a non-executable file whose executable must be discovered via file association rules? Does that sound accurate? I'm not entirely sure what you mean by

Re: shlex, unicode, and subprocess.Popen on Windows

2010-08-30 Thread python
Hi Chris, > It's a bug: http://bugs.python.org/issue1170 Thanks for pointing out the shlex bug. My concern was that shlex had Windows specific Unicode limitations because of the way the Windows shell so poorly supports unicode output. > Kudos for avoiding shell=True My understanding is that the

Re: shlex, unicode, and subprocess.Popen on Windows

2010-08-30 Thread Chris Rebert
On Mon, Aug 30, 2010 at 6:54 AM, wrote: > Python 2.6/Windows: shlex.split() does not support unicode strings. Is this > simply a limitation of the current shlex implementation or is this an > intentional design decision that reflects the behavior of how the Windows > shell supports unicode values