Pipe problem

[Doug]

This works with a small string, but not a large one (read returns an
empty string if i pass a large html file to tidy):

[color=blue][color=green][color=darkred]
>>> iin , iiout = popen2("tidy.ex e -asxml")
>>> iin.write(str)
>>> iin.close()
>>> iiout.read()[/color][/color][/color]
''

I am using windows and tried the standard pipe and win32pipe as well.

Doug

[Albert Hofkamp]

Re: Pipe problem

On Sat, 13 Sep 2003 16:34:13 GMT, Doug <no.email.you.m ust.phone.me@fu .bar.ws> wrote:[color=blue]
> This works with a small string, but not a large one (read returns an
> empty string if i pass a large html file to tidy):
>
>[color=green][color=darkred]
> >>> iin , iiout = popen2("tidy.ex e -asxml")
> >>> iin.write(str)
> >>> iin.close()
> >>> iiout.read()[/color][/color]
> ''
>
> I am using windows and tried the standard pipe and win32pipe as well.[/color]

Probably the standard pipe buffering problem.

As you know computer systems have a finite amount of memory, which is
used for many things, including buffering data written to a pipe.

The write(str) returns when you have written the entire contents out.
Since it is big, and tidy.exe also uses finite buffering, the latter
starts processing and returning data while you are writing. Since you
are not reading incoming data, the pipe in the other direction gets
full, tidy.exe becomes blocked, which in turn means your write() becomes
blocked.

Solution: read and write at the same time.


Albert
--
Unlike popular belief, the .doc format is not an open publically available format.

[John J. Lee]

Re: Pipe problem

Albert Hofkamp <hat@se-126.se.wtb.tue. nl> writes:
[color=blue]
> On Sat, 13 Sep 2003 16:34:13 GMT, Doug <no.email.you.m ust.phone.me@fu .bar.ws> wrote:[color=green]
> > This works with a small string, but not a large one (read returns an
> > empty string if i pass a large html file to tidy):
> >
> >[color=darkred]
> > >>> iin , iiout = popen2("tidy.ex e -asxml")
> > >>> iin.write(str)
> > >>> iin.close()
> > >>> iiout.read()[/color][/color][/color]
[...]

Didn't see the OP, but (guessing tidy.exe is HTMLTidy): do you know
about mxTidy and uTidylib?


John