This issue tracker has been migrated to GitHub, and is currently read-only.
For more information, see the GitHub FAQs in the Python's Developer Guide.

classification
Title: windows sys.path contains nonexistant directory
Type: behavior Stage: resolved
Components: Versions: Python 3.1, Python 3.2, Python 2.7
process
Status: closed Resolution: fixed
Dependencies: Superseder:
Assigned To: tim.golden Nosy List: BreamoreBoy, JosephArmbruster, logistix, quentin.gallet-gilles, tim.golden, tim.peters
Priority: low Keywords:

Created on 2003-09-02 01:31 by logistix, last changed 2022-04-10 16:10 by admin. This issue is now closed.

Messages (8)
msg18042 - (view) Author: Grant Olson (logistix) Date: 2003-09-02 01:31
Starting in python 2.3, sys.path on Windows contains 
the nonexistant directory "c:\python23\lib\plat-win".

I don't know if this is really a bug, since we can no 
longer assume that an entry in sys.path exists or is a 
physical directory.  Still, it seems like each entry in 
sys.path should be able to resolve to SOMETHING.

I would've flagged this low priority if I had the capability.
msg18043 - (view) Author: Tim Peters (tim.peters) * (Python committer) Date: 2003-09-02 01:40
Logged In: YES 
user_id=31435

Strange.  I see it under Python 2.1.3 too.  So it's not a new 
thing, but somehow 2.2.3 managed to get rid of it temporarily 
<wink>.  Boosting the priority to 4, since it's at least a 
mystery!
msg18044 - (view) Author: Grant Olson (logistix) Date: 2003-09-16 01:01
Logged In: YES 
user_id=699438

There's a wierd bug on line 255 of getpathp.c.  The malloc 
fails when trying to allocate 42 bytes, which is the length 
of "\Software\Python\PythonCore\X.X\PythonPath" and it 
causes the function to abort.

Just on a hunch, I bumped the malloc size up to 44 since 
that's an even number of dwords, and the malloc magically 
succeeded.

Also, Since I'm running XP, shouldn't the 
sizeof "\Software\Python\PythonCore\X.X\PythonPath" be 84?
msg18045 - (view) Author: Grant Olson (logistix) Date: 2003-09-16 02:01
Logged In: YES 
user_id=699438

Disregard last comment.  I ran a clean and the problem 
disappeared.
msg58215 - (view) Author: Quentin Gallet-Gilles (quentin.gallet-gilles) Date: 2007-12-05 15:06
I'm able to reproduce it on 2.4, 2.5 and it's most likely still here in
the trunk. Is it still considered a bug ?


>>> import sys

>>> import os

>>> for file in sys.path:

...     print "%s - %s" % (file, os.path.exists(file))

...

: False

C:\WINNT\system32\python25.zip : False

C:\PYTHON25 : True

C:\Python25\DLLs : True

C:\Python25\lib : True

C:\Python25\lib\plat-win : False

C:\Python25\lib\lib-tk : True

C:\Python25\lib\site-packages : True
msg59238 - (view) Author: Joseph Armbruster (JosephArmbruster) Date: 2008-01-04 16:29
If I look at Lib/plat-win back at rev 16910, winreg.py could be found
there.  Everything I see in the tree referencing plat-win is
documentation, except the PYTONPATH define itself.  It looks like
Lib/plat-win was removed back in revision 16911 [Aug 20,2000].

Log output:
Remove the winreg module from the project.  I don't believe any
docs changes are needed (only reference to winreg I could find
was in libwinreg.tex, which is documenting _winreg, and merely
mentions that a higher-level winreg module *may* appear someday;
that's still true).

And made its way back in, in 27697 [July 19, 2002].

Log output:
Land Patch [ 566100 ] Rationalize DL_IMPORT and DL_EXPORT.


I'm thinking this is stale code.

As per the test app, I know from a previous inquiry of mine that
C:\WINNT\system32\python25.zip (and the like on different systems /
version) is in fact necessary.

Your thoughts?
msg114266 - (view) Author: Mark Lawrence (BreamoreBoy) * Date: 2010-08-18 19:08
Still an issue in 2.6, 2.7, 3.1 and 3.2.
msg164357 - (view) Author: Tim Golden (tim.golden) * (Python committer) Date: 2012-06-29 21:56
THis was fixed almost two years ago in Py3K. Won't fix for 2.7
History
Date User Action Args
2022-04-10 16:10:58adminsetgithub: 39170
2012-06-29 21:56:20tim.goldensetstatus: open -> closed
resolution: fixed
messages: + msg164357

stage: test needed -> resolved
2010-08-20 09:04:32tim.goldensetassignee: tim.golden

nosy: + tim.golden
components: - Interpreter Core
versions: + Python 3.1, Python 3.2
2010-08-18 19:08:48BreamoreBoysetnosy: + BreamoreBoy

messages: + msg114266
versions: + Python 2.7, - Python 2.6
2009-02-13 03:39:37ajaksu2setstage: test needed
type: behavior
versions: - Python 2.5
2008-01-04 16:29:55JosephArmbrustersetnosy: + JosephArmbruster
messages: + msg59238
2008-01-04 00:42:10christian.heimessetversions: + Python 2.6, - Python 2.4, Python 2.3
2007-12-05 15:06:43quentin.gallet-gillessetnosy: + quentin.gallet-gilles
messages: + msg58215
versions: + Python 2.5, Python 2.4
2003-09-02 01:31:14logistixcreate