[python-win32] makepy creates gen_py/123-ID.py but no gen_py/ 123-ID - Folder

[Simon Dahlbacka]

AFAIK depending on the optional bForDemand parameter to EnsureDispatch
either one big file is generated or a directory containing the needed
stuff as they are needed.


On Wed, 10 Nov 2004 10:41:13 +0100, The Count
<the.count at count0.dyndns.org> wrote:
> On Wed, 10 Nov 2004 09:10:50 -0000, Tim Golden
> <tim.golden at viacom-outdoor.co.uk> wrote:
> 
> > [The Count]
> >
> > [... snip losing makepy folders etc. ...]
> >
> > | What am I missing? I do (sort of) remember they weren't there
> > | from the start but I have no clue as to how I created them,
> > | they're source claims them to be created by makepy ...
> >
> > Don't know why makepy isn't generating them,
> 
> To clarify : I DO get
>   - the $ID$.py / .pyc file
>   - an __init__.py / .pyc and
>   - a dicts.dat file
> in .../win32com/gen_py/ but I'm missing a folder of the same name.
> This contains files for all those items inside the type libraries branch
> displayed via combrowse.py,
> they're only the wrappers but - for example:
> Certain functions inside the application allow to do stuff like this :
> Data("foo") = "bar"
> This is not legal in python.
> Obviously they use a "SetData('foo','bar')" wrap around that, but it's not
> documented,
> those wrapper sources give the game away, though :-)
> 
> > but if you haven't
> > already, try using EnsureDispatch,
> 
> Well, that doesn't seem to change anything for me,
> could you maybe get me up to speed on gencache.EnsureDispatch vs.
> Dispatch?!
> What's the difference?
> 
> > Also, I seem to remember that Mark has moved around the location
> > in which makepy creates its proxy modules. I'm running the latest
> > stable pywin32 and the proxies appear to be being created under
> > my profile in "Local Settings\temp\gen_py\2.3". So, for me, that's
> 
> I checked my system temp folder and my "local settings" folder without
> finding anything,
> also I didn't change any part of my python installation since first
> running makepy on it,
> so the location should still be the same.
> 
> I just remember one day clicking on part of a traceback and landing right
> inside the wrapper around the COM objects function I botched up with.
> 
> --
> ---
> The Count
> 
> 
> 
> _______________________________________________
> Python-win32 mailing list
> Python-win32 at python.org
> http://mail.python.org/mailman/listinfo/python-win32
>