Parse out all image tags and attributes in Perl

[DarkWater]

Just decided to experiment with it really quickly (kinda bored, to be honest).  Anyone have any advice for parts I could rewrite?

 

#!/usr/bin/perl

$_ = <<HTML; #set up test HTML, would be from file normally
<img src="foo" height="1" />
<img src="testing!" height="50" width="200" />
<img height="20" src="lol" />
HTML

my @imgs;

while (/<img (.+?) \/>/g) {
push @imgs, {$1 =~ /(\w+)="([^"]+)"/g}; #@imgs becomes an array of anonymous hashes
}

for $loc (0 .. $#imgs) { #loop through indices to for easy position tracking (might change in real script to just a for loop)
print "Image at location $loc:\n";
$hash = $imgs[$loc]; #get the hash out!
for $key (keys %$hash) { #and deference it out of the scalar.  made THAT mistake a few times
	print "$key = $hash->{$key}\n"; #print out the pairs
}
}

 

Anyone have any suggestions?

[DarkWater]

Just changed it a tiny bit to use strict and warning pragmas, so I also needed to change that my declaration:

 

#!/usr/bin/perl
use strict;
use warnings;

$_ = <<HTML; #set up test HTML, would be from file normally
<img src="foo" height="1" />
<img src="testing!" height="50" width="200" />
<img height="20" src="lol" />
HTML

my (@imgs, $loc, $hash, $key);

while (/<img (.+?) \/>/g) {
push @imgs, {$1 =~ /(\w+)="([^"]+)"/g}; #@imgs becomes an array of anonymous hashes
}

for $loc (0 .. $#imgs) { #loop through indices to for easy position tracking (might change in real script to just a for loop)
print "Image $loc:\n";
$hash = $imgs[$loc]; #get the hash out!
for $key (keys %$hash) { #and deference it out of the scalar.  made THAT mistake a few times
	print "$key = $hash->{$key}\n"; #print out the pairs
}
}

 

No big difference.

[effigy]

1. You're only going to catch double-quoted attributes.

2. You haven't accounted for rogue white space.

3. There are various modules on CPAN for things like this; my take is below:

 

use strict;
use warnings;
use XML::Twig;
use Data::Dumper;

my $data = <<HTML;
<img src="foo" height="1" />
<img src="testing!" height="50" width="200" />
<img height="20" src="lol" />
HTML

my @atts;
my $twig = XML::Twig->new(
twig_roots => {
	'img' => sub {
		push @atts, $_->atts();
	}
}
);
$twig->parse_html($data);
print Data::Dumper->Dump([\@atts]);

[DarkWater]

1. You're only going to catch double-quoted attributes.

2. You haven't accounted for rogue white space.

3. There are various modules on CPAN for things like this; my take is below:

 

1. I was only planning on using this on valid XHTML anyway, so the double quoted attributes aren't a problem.

2. Do you mean like:

$1 =~ /(\w+)\s*=\s*"([^"]+)"/g

(notice the addition of \s*)

3. I know, I love CPAN, I was just wanting to try this on my own just to experiment with.

[effigy]

2. Do you mean like:

$1 =~ /(\w+)\s*=\s*"([^"]+)"/g

(notice the addition of \s*)

 

Yes.

[DarkWater]

Alright, thanks.