[PATCH] setup.py: Workaround for missing bz2 module in IronPython

[Augie Fackler]

On Mon, May 9, 2011 at 11:02 AM, Zachary Gramana
<zgramana at pottsconsultinggroup.com> wrote:
> That's a good question, and one that does not seem to have been discussed on
> the IronPython dev list, though there is outstanding work item on it:
> http://ironpython.codeplex.com/workitem/21412. The only other reference I
> can find is in the cmodule.py IronPython unit test script:
>
>  "#Modules we don't implement found in DLLs or Lib and
>  #the reason why:
>  # bz2 - TODO?"
>
> Just a guess: the only OSS C# bzip2 library that I could find is licensed
> under GPL2, which conflicts with IronPython's MS-PL license, and nobody has
> felt like re-implementing it.
>
> In addition, there is a project called IronClad which allows you to load
> CPython's bz2.pyd. Unfortunately, IronClad is not yet available for
> IronPython 2.7 and hg flat out cannot run on IronPython 2.6, which lacks
> some very basic built-in modules (e.g. signal).  Moreover, IronClad's
> developer list has gone quiet as of late, so I'm not anticipating any new
> development soon.
>
> An ipy bz2 implementation is probably just low priority given that
> IronPython is now community driven. The main contributors are turning their
> attention to 3.0 support.
>
> That's all just conjecture, though.

A simple Google search pointed me to http://www.digistar.com/bzip2/,
which seems like it might be helpful.

>
>
> On 5/9/2011 10:52 AM, Augie Fackler wrote:
>>
>> On Mon, May 9, 2011 at 9:01 AM, Zachary Gramana
>> <zgramana at pottsconsultinggroup.com>  wrote:
>>>
>>> # HG changeset patch
>>> # User Zachary Gramana<zgramana at pottsconsultinggroup.com>
>>> # Date 1304949224 14400
>>> # Node ID 1f6b3f2388e5e029999a0b17014f0eaa8dd1641b
>>> # Parent  7733e9df43db0a011c866446350c68529e8ecd7c
>>> setup.py: Workaround for missing bz2 module in IronPython
>>>
>>> IronPython does not provide the bz2 module on its own.  This patch skips
>>> importing the bz2 module.
>>
>> OOC, Is there any particular reason IronPython is broken in this respect?
>>
>>>
>>> diff -r 7733e9df43db -r 1f6b3f2388e5 setup.py
>>> --- a/setup.py  Mon May 09 09:38:37 2011 -0400
>>> +++ b/setup.py  Mon May 09 09:53:44 2011 -0400
>>> @@ -4,7 +4,7 @@
>>>  # 'python setup.py install', or
>>>  # 'python setup.py --help' for more options
>>>
>>> -import sys
>>> +import sys, os
>>>  if not hasattr(sys, 'version_info') or sys.version_info<  (2, 4, 0,
>>> 'final'):
>>>     raise SystemExit("Mercurial requires Python 2.4 or later.")
>>>
>>> @@ -36,13 +36,21 @@
>>>     raise SystemExit(
>>>         "Couldn't import standard zlib (incomplete Python install).")
>>>
>>> -try:
>>> -    import bz2
>>> -except:
>>> -    raise SystemExit(
>>> -        "Couldn't import standard bz2 (incomplete Python install).")
>>> +# The base IronPython distribution (as of 2.7.1) doesn't support bz2
>>> + try:
>>> +    isIronPython =  os.path.split(sys.executable)[1] == "ipy.exe"
>>> +    pass
>>> +else:
>>> +    if isIronPython:
>>> +        print "(IronPython detected. Skipping bz2 import.)"
>>> +    else:
>>> +        try:
>>> +            import bz2
>>> +        except:
>>> +            raise SystemExit(
>>> +                "Couldn't import standard bz2 (incomplete Python
>>> install).")
>>>
>>> -import os, subprocess, time
>>> +import subprocess, time
>>>  import shutil
>>>  import tempfile
>>>  from distutils import log
>>>
>>> _______________________________________________
>>> Mercurial-devel mailing list
>>> Mercurial-devel at selenic.com
>>> http://selenic.com/mailman/listinfo/mercurial-devel
>>>
>>>
>