From: "Johnson, Shaunn" <[EMAIL PROTECTED]>
> I have this bit of code (below) and I'm wondering
> if there is a quicker way to remove
> some odd-ball characters from very
> large text files (large would be about the
> 200M or so).
>
> [snip code]
>
> #!/usr/bin/perl
>
> #$_ =~ s/\cM\n/\n/g;
>
> while (<>) {
> $_ =~ s/(\cM\n|\\|\~|\!|\@|\#|\$|\%|\^|\&|\*|\(|\))/\n/g;
> print $_;
> }
>
> [/snip code]
>
> I want to add some variable to pass (and rename INPUT file)
> but before I do, I'd like to know if doing something like open()
> would be any faster than this.
tr/// is quicker than s/// so if that is possible you should use
that. Also your s/// looks a bit strange. You really want to replace
any of those characters with a newline?
Also is there any reason to leave a \cM that's not followed by \n in
the file? If there is not you could use
$_ =~ [EMAIL PROTECTED]&*()}{\n};
Also ... if replacing any \cM is fine it would be more efficient to
read the file in chunks instead of line by line:
open my $IN, "< $filename"
or die "Can't open $filename: $!\n"
open my $OUT, "> $outfilename"
or die "Can't create $outfilename: $!\n"
while (read $IN, $buff, 10*1024) {
$buff =~ [EMAIL PROTECTED]&*()}{\n};
print $OUT $buff;
}
close $IN;
close $OUT;
HTH, Jenda
===== [EMAIL PROTECTED] === http://Jenda.Krynicky.cz =====
When it comes to wine, women and song, wizards are allowed
to get drunk and croon as much as they like.
-- Terry Pratchett in Sourcery
--
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
