When pydoc is run from the command line and unable to write a file for a
module because of a syntax error, it would be helpful to return a
non-zero status code as a result.

What it does currently:
    C:\>C:\python26\Lib\pydoc.py -w mymod
    problem in .\mymod.py - <type 'exceptions.SyntaxError'>: invalid
syntax (mymod.py, line 1)

    C:\>echo %ERRORLEVEL%
    0

What it should do if an error is found:
    C:\>type doesnt_exist.txt
    The system cannot find the file specified.

    C:\>echo %ERRORLEVEL%
    1


This would allow us to make our build scripts able to react to error
conditions, etc without having to use another script.  Thanks.

This seems like a resonable thing for pydoc to do.  However, I took a
quick look, and pydoc has no notion of error status. Adding one would
require a non-trival refactoring of the code (at least to do it
right...a hackish version might not be too difficult).  

So, I doubt this is going to happen unless you can contribute a patch.

On the other hand, your test suite should be catching any errors pydoc
would catch, so I can't say I really understand your stated reason for
the request.  Unless you are building documentation for packages you
don't maintain?

Ahh, I see ... I didn't explain my issue with much depth in haste.

I don't usually run the test suite when I am in the process of updating
code or docs, only when I'm done, then I check it in.

So I was updating the docs and running pydoc with a local batch file and
proofreading the html output.  If there was an error, the file seem to
update.  

You are right it won't happen in the build process after all, but rather
when I am making repetitive changes locally and using the command line.

I might look into doing it myself, although not sure when I'll have time
to get to it.