Subprocess module

[Christian Heimes]

Dominique.Holzwarth at ch.delarue.com schrieb:
> Hello all
> 
> I want to convert a tex file into a pdf by using pdflatex. For that, I thought the 'subprocess' module might be a good option. My code doesn't work at all tho:
> 
> Import os, subprocess
> 
> def main():
>         scriptpath = os.path.dirname(__file__)
> 
>         p = subprocess.Popen("pdflatex --include-directory=%s --output-directory=%s/output --aux-directory=%s/temp --interaction=nonstopmode                                                    myFile.tex"
>                              % (scriptpath, scriptpath, scriptpath),
>                              stdin=subprocess.PIPE,
>                              stdout=subprocess.PIPE,
>                              stderr=subprocess.PIPE,
>                              shell=True,
>                              cwd=scriptpath)
>     (child_stdin,
>      child_stdout,
>      child_stderr) = (p.stdin, p.stdout, p.stderr)
>     print 'stdin'
>     print child_stdin
>     print 'stdout'
>     print child_stdout
>     print 'stderr'
>     print child_stderr
> 
> When I run that code I get the following printouts:
> 
> stdin
> <open file '<fdopen>', mode 'wb' at 0x009E7968>
> stdout
> <open file '<fdopen>', mode 'rb' at 0x009E7A40>
> stderr
> <open file '<fdopen>', mode 'rb' at 0x009E79F8>
> Done

The pdflatex job stales when the standard stream buffers are full. Why
do you need stdin anyway? The community method should do the trick for you:

    out, err = p.communicate()

I also suggest you use the form ["pdflatex",
"--include-directory="+scriptpath ...] instead of a single string +
shell=True.

Christian