Question

I want to extract the size value from a string. The string can be be formatted in one of two ways:

Data-Size: (2000 bytes)

or

file Data-Size: (2082 bytes)

If the string is present in a file, it will appear only once.

So far I have:

#!/usr/bin/perl

use strict;
use warnings;

open FILE, "</tmp/test";
my $input = do { local $/; <FILE> };

my ($length) = $input =~ /(file)?\s*Data-Size: \((\d+) bytes\)/m;                   

$length or die "could not get data length\n";
print "length: $length\n";

The problem seems to be with making the word file optional. I thought I could do this with:

(file)?

But this seems to be stopping matches when the word file is not present. Also when the word file is there it sets $length to the string “file”. I think this is because the parenthesis around file also mean extraction.

So how do I match either of the two strings and extract the size value?

Answer 1

You want the second capture in $length. To do that, you could use

my (undef, $length) = $input =~ /(file)?\s*Data-Size: \((\d+) bytes\)/;

or

my $length = ( $input =~ /(file)?\s*Data-Size: \((\d+) bytes\)/ )[1];

But a much better approach would be to avoid capturing something you’re not interested in capturing.

my ($length) = $input =~ /(?:file)?\s*Data-Size: \((\d+) bytes\)/;

Of course, you’d get the same result from

my ($length) = $input =~ /Data-Size: \((\d+) bytes\)/;

---

By the way, I removed the needless /m. /m changes the meaning of ^ and $, yet neither are present in the pattern.