Thanks Tom - I've committed your changes.

On Wednesday 18 January 2012 08:42:30 [email protected] 
wrote:
>  Herve, I agree that unbuffered output is fine, but others might have
>  need for it. So, I added the -b option to both and I've been using for a
>  couple of weeks. I've attached them here because I don't have github
>  access.
> 
>  The tokenizer.perl default is still buffered and the new -b option
>  disables Perl buffering. The detokenizer removes the %|++; line, which
>  changes the script's default to buffered. The new -b option disables
>  Perl buffering. The usage help prompt is also updated with the changes.
> 
>  Maybe someone can update the repository.
> 
>  Thanks,
>  Tom
> 
>  On Mon, 09 Jan 2012 12:59:30 +0000, Herve Saint-Amand
> 
>  <[email protected]> wrote:
> > Hi,
> >
> > I added that $|++; a while ago so that a script could fork off
> > detokenizer.perl, keep it running as a child process, feed it input
> > line
> > by line and synchronously read back the corresponding output.
> >
> > As far as I can see it's OK to just always unbuffer output. I don't
> > think it can break anything, but honestly I don't know what effect it
> > has on I/O performance. For what it's worth I always add that line to
> > every Perl script I write.
> >
> > I'm cool with the command-line switch too.
> >
> > Herve
> >
> > On 07/01/12 11:58, Tom Hoar wrote:
> >> I just found that detokenizer.perl has a variant of the proposal
> >> below
> >> on lines 8 & 9.
> >>
> >> # This added by Herve Saint-Amand for compatibility with
> >> translate.cgi
> >> $|++;
> >>
> >> For consistency between the two scripts, I propose adding a
> >> command-line
> >> option to both that disables buffering if the user wishes, and a
> >> line in
> >> the usage help.
> >>
> >> while (@ARGV) {
> >> $_ = shift;
> >> /^-b$/ && ($| = 1, next);
> >>
> >> if ($HELP) {
> >> print " -b ... disable Perl buffering.\n";
> >>
> >> On Thu, 05 Jan 2012 20:28:01 +0700, Tom Hoar
> >>
> >> <[email protected]> wrote:
> >>> The current tokenizer.perl and detokenizer.perl script buffer the
> >>> print output. Adding this line to the beginning of each script:
> >>>
> >>> $| = 1;
> >>>
> >>> disables buffering and forces Perl to flush each line. Can this be
> >>> added?
> >>>
> >>> Thanks,
> >>> Tom
> >>
> >> _______________________________________________
> >> Moses-support mailing list
> >> [email protected]
> >> http://mailman.mit.edu/mailman/listinfo/moses-support
> 
 
--
Barry Haddow
University of Edinburgh
+44 (0) 131 651 3173

-- 
The University of Edinburgh is a charitable body, registered in
Scotland, with registration number SC005336.

_______________________________________________
Moses-support mailing list
[email protected]
http://mailman.mit.edu/mailman/listinfo/moses-support

Reply via email to