#! /usr/bin/perl
use warnings;
use strict;
use feature qw{ say };

use XML::LibXML;

my $string = q~...~;
my $xml = 'XML::LibXML'->load_html(string => $string);
say for $xml->findnodes('//a');
[download]

So you want to keep everything that looks like a link?

I would use something like HTML::TreeBuilder::XPath and the appropriate XPath query (//a). Other candidates are XML::Twig.

Personally, I would use App::scrape, which is a tiny command line wrapper around HTML::TreeBuilder::XPath.

More specifically, the links are all just news affiliate webites, like newsok.com, etc. I have no idea which news affiliate websites they will be, but there are hundreds of them. He changes some of them regularly, so I am building this to check on them once per day, to pick up any new ones and add them to our database.

My friend that does that said to check it daily, so I am just writing a script to go do that. The part I'm having a problem with is getting the full html link, I've been using strip and striping down every part, but that is just too much, I know there is an expr that will work. I just cannot recall how to write it.

Rich

Have you looked at the modules I linked? They will all happily extract the links.

Alternatively, you might want to (re)read perlre, but I would use an existing HTML parser instead of trying my own.

I did not mean expr, I meant a regex...

The URL's will always be different, I won't know what they are, it is based upon unique links, a friend of mine always changes and he said I could always get them, I am building a script that will check them for me, to see if I already have them, I don't want to check everyday, manually.

thanks,
Rich

a regex like this: <((?!a[ ]).|\n)*?>

Except one that leaves the trailing </a> in it.

Can you find one that is like that that works?

But in there, you know the base:



my $base = 'http://perlmonks.org/';
[download]

I will have no idea what they are, could be any news affiliate website in the world.

I just want to remove the other stuff and leave what is in the html link:



Link: <a href="http://example.com">and Anchor</a>
[download]

If that above were the string, it would remove Link: and leave the rest.



my $string = q~Link: <a href="http://example.com">and Anchor</a>~;
$string =~ s/<[a href.... # I cannot remember this string. There was o
+ne that worked perfect, even if the link had target="_blank" it did n
+ot matter what else it had... but I cannot find it in any of my files
+ or remember who to write it.
[download]



Also, I've at this point already downloaded the one page they are all on, and I've parsed it down to just one table cell, that has other data in it and I've gotten out of that table cell the information I need, all that is left is the remnants including the html link with anchor... so I want to just use that string to remove everything left, except the link and anchor.

For production code that will be used regularly or by other people, I would use one of the HTML-parsing modules mentioned earlier. For a one-time grab, a regex may be good enough to get the job done. If there's only one A link in the block of text:

$text =~ s|^.*(<a .+?</a>).*$|$1|s;
[download]