How about File::Copy? It comes with your Perl unless it's a fairly old Perl.

Preferred solution: Heed Corion's note above.

FTR, while your code is "close," it won't do quite what you want unless your file.ext is a single, monolithic line. If, as is likely, your file.ext contains newlines, your copy will contain only the last line of the original.

One (TIMTOWTDI) cure for that if your log is plain text):

my $data;    # ugly global; don't do this

open(LOG, "file.ext") or die ("oops, could not open source", $!);
local $/;             # slurp mode (see perlvar Record Separator)
$data2 = <LOG>;
close(LOG);

open(LOG2, ">","path/to/copy_of_file.ext") or die ("oops - could not o
+pen target", $!);
print LOG2 $data2;
close(LOG2);
[download]

For non-text files, see binmode; note also the modest changes in the die messages, which IMO (YMMV) are good practice, just to make sure that even if I glance only carelessly at the message, I'm still apt to note which file failed to open. Also note that you could do this by calling the shell's cp or copy method.

open OLD "<", $oldfile or die "Can't open $oldfile ($!)\n";
open NEW ">", $newfile or die "Can't open $newfile ($!)\n";

print NEW while(<OLD>);

close OLD;
close NEW;
[download]

File::Copy will at least copy more than a line at a time...by default 2Mb at a time for files, 1k for non-files. Which is likely preferable to one line at a time. But I think I'd go with system("cp ...")

I have used this approach numerous times (though after I discovered File::Copy I've used it much more). The only thing I missed (typical of my not seeing all the power one can get into a single line) is the 'print NEW while(<OLD>)' single line.

I am curious that in the original Anonymous Monk's code the input from OLD is to a scalar, doesn't that just read in a single record from the OLD document?

ack Albuquerque, NM

$rc = `cp /path/to/old/file/file.xt /path/to/new/file/file.txt`;
[download]