Todd Wade <[EMAIL PROTECTED]> wrote:
> "Gary Stainburn" <[EMAIL PROTECTED]> wrote in message
> news:[EMAIL PROTECTED]
[...]
> > One thing you forgot was to close the file. Also, don't
> > forget that you can do it with less typing:
Closing files is optional in Perl ;-) Filehandles
will either be closed when the program terminates
when the filehandle is opened the next time.
> > $file = "/location/of/header.incl";
> > open (FH, "$file") or die "Cannot open file $file $!";
If you wanted to pipe other stuff than just plain
text, then
binmode(FH);
would be a good idea for portability.
> > print while(<FH>);
This is bad because it first pulls in the file to
build the list.
> > close(FH);
> print <FH>;
>
> print takes a list of arguments:
Just to be nit-picking (and to repeat what I learned
from a M.J. Dominus talk at YAPC::EU this year ;-))
"print'ing a list is slower than printing a single
(large) scalar."
In this particular case it might be worthwhile to
use "slurping": We create a scope in which the input
record separator $/ is temporarily set to undef.
Reading from <IN> then returns just a single record.
The payoff - printing is nearly twice as fast.
#!/usr/bin/perl -w
use strict;
use Benchmark;
my $file = "out.txt";
my $count = 20000;
open( OUT, ">$file") or die "Can't open '$file':$!";
print "print using a scalar:\n";
timethis( $count, 'open(IN, "$0"); { local $/; undef $/; print OUT <IN>; };
close( IN)' );
close( OUT );
unlink $file;
open( OUT, ">$file") or die "Can't open '$file':$!";
print "print using a list:\n";
timethis( $count, 'open(IN, "$0"); print OUT <IN>; close( IN)' );
close( OUT );
unlink $file;
__END__
print using a scalar:
timethis 20000: 6 wallclock secs ( 2.85 usr + 2.44 sys = 5.30 CPU) @
3775.72/s (n=20000)
print using a list:
timethis 20000: 11 wallclock secs ( 7.88 usr + 2.67 sys = 10.56 CPU) @
1894.84/s (n=20000)
Cheers,
Thomas
--
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]