Here is a patch that moves the source files of the _io module (and
closely related modules) into subdirectory of Modules. The new file
hierarchy proposed is the following:

  Modules/
    _io/
      _iomodule.h
      _iomodule.c
      iobase.c
      fileio.c
      bufferedio.c
      textio.c
      bytesio.c
      stringio.c

I believe that moving the io module into its own subdirectory will make
clear whether the different I/O implementations share common bugs.
Currently, it is quite easy to miss a module when doing a bug fix.

The patch was not tested on Windows. I tried my best to configure by
hand PCbuild and PC/{VC, VS7.1, VS8.0}. However, I would be surprised if
my changes were flawless.

Where's the patch?

Some comments:
- there are some unrelated pickle changes in your patch
- in Makefile.pre.in, it seems the IO_OBJS definition might need updating

Otherwise I don't have any objections.
Please note, when committing, it would be good to use "svn mv" for the
actual moving of the files, so that their history can be followed.

Antoine, you're too quick. :-)

By the way, the patch is fairly large due to the renaming. So, here's
quick-and-dirty summary of the important changes.

  PCbuild/pythoncore.vcproj: Renamed the paths and added a <Filter/> tag
for the _io directory.
  PC/VS7.1/pythoncore.vcproj: Same as above.
  PC/VS8.0/pythoncore.vcproj: Same as above.
  PC/VC6/pythoncore.dsp: Renamed the paths.
  Makefile.pre.in: Updated the path to _iomodule.h
  Modules/Setup.dist: Renamed the paths.

Oh, my last patch had some junk for pickle. Here's a clean patch.

Antoine wrote:
> - in Makefile.pre.in, it seems the IO_OBJS definition might need updating

I updated the patch to also fix IO_OBJS. Also, I added stringio.c to
IO_OBJS. 

> Please note, when committing, it would be good to use "svn mv" for the
> actual moving of the files, so that their history can be followed.

Yes, I used "svn move" to rename the file in my local checkout, but
unfortunately "svn diff" doesn't support renames.

I think you can commit!

Have you tested PCbuild changes? As, this is the part I am the most
unsure about.

> Have you tested PCbuild changes? As, this is the part I am the most
> unsure about.

Ok, unless someone beats me to it, I'm gonna fire a Windows VM and test
this.

Some failures:

_iomodule.c
c1 : fatal error C1083: Impossible d'ouvrir le fichier source :
'..\..\Modules\_io\_iomodule.c' : No such file or directory
textio.c
c1 : fatal error C1083: Impossible d'ouvrir le fichier source :
'..\..\Modules\_io\textio.c' : No such file or directory
iobase.c
c1 : fatal error C1083: Impossible d'ouvrir le fichier source :
'..\..\Modules\_io\iobase.c' : No such file or directory
bufferedio.c
c1 : fatal error C1083: Impossible d'ouvrir le fichier source :
'..\..\Modules\_io\bufferedio.c' : No such file or directory
stringio.c
c1 : fatal error C1083: Impossible d'ouvrir le fichier source :
'..\..\Modules\_io\stringio.c' : No such file or directory
bytesio.c
c1 : fatal error C1083: Impossible d'ouvrir le fichier source :
'..\..\Modules\_io\bytesio.c' : No such file or directory
fileio.c
c1 : fatal error C1083: Impossible d'ouvrir le fichier source :
'..\..\Modules\_io\fileio.c' : No such file or directory

It looks like a mistake in file names (missing leading underscores, and
a non-existent _iomodule.c instead of io.c).

Or perhaps these are really meant as the new file names? Initially I
thought the purpose was initially to move the files into a subdir.

After also renaming the files, there is a slight bug in
PCbuild/pythoncore.vcproj. Here is a subpatch incorporating the changes
to this only file.

Thanks! Here is the updated patch.

You should commit. If there is a problem with other Windows build files,
it will certainly be noticed and fixed by a true Windows developer.

Committed in r71185.

Thanks Antoine for the review!