Enumerating drives in Python or wxPython (different behavior in PythonWin and python.exe)

[Alan Miller]

David Bolen (db3l at fitlinxx.com) wrote:
[os.path.isdir() to identify drives, and what happens for removables]
>Alan Miller <ajm at enteract.com> writes:
>> Interesting, I tried it out before I posted and didn't have any problems.  
>
>Strange - it did for me for A: (I did it just before posting as well).
>Perhaps it's a Windowism that depends on platform or some strange
>registry setting or something.  I was on NT 4, SP4.  In any event, I
>guess at the very least it means that it's behavior isn't guaranteed
>to be consistent.

Interesting, I'm on NT4WS, SP6a.  It also works on a Win95 laptop with 
the floppy drive missing, though it (Win95, really) goes out to lunch for 
a few seconds waiting for a response from the drive.

>The problem was that isdir() didn't return an error - it returned 1
>(which I guess is ok, given that A:\ is technically valid even without
>a diskette), but the NT error while performing isdir() put up a
>message box that the user would have to answer before isdir() returned
>in the Python script.

I'm running Python 1.5.2 and PythonWin build 132, and I was running the 
test from the PythonWin interactive window.  

I just tried it from a command line Python and got the message you were 
probably seeing - "There is no disk in the drive.  Please insert a disk 
into drive A:.\n[Abort][Retry][Ignore]"

I'm not sure what to do about that - obviously PythonWin is doing 
something differently, but I'm not sure what - it should be using the 
same modules, but I don't know if it has the builtin modules compiled 
into it or if it calls python.exe.  I suspect the entire interpreter and 
the builtins are compiled in, possibly Visual C/Visual C++ does something 
unusual in terms of optimization?

Hopefully the change in the subject will catch Mark's attention....
Mark, any thoughts on this?

ajm