reddit-get - Grabs image files from reddit (./get.pl <subreddit>).

get.pl

#!/usr/bin/env perl

use strict;
use warnings;
use JSON;
use WWW::Mechanize;

die "Usage: $0 subreddit [...]\n" unless @ARGV;

my $mech = WWW::Mechanize->new(autocheck=>0);

foreach my $arg (@ARGV) {

    # Create the directory if necessary.
    mkdir "downloads/$arg" unless -d "downloads/$arg";

    my $url = "http://www.reddit.com/r/$arg/.json?limit=1000";
    print "Downloading from http://www.reddit.com/r/$arg\n";
    $mech->get($url);

    my $json_text = JSON->new->allow_nonref->utf8->relaxed->decode($mech->text);
    my $posts = $json_text->{data}->{children}; 
    foreach my $post (@{$posts}) {
        my $url = $post->{data}->{url};
        download_image($url, $arg) if $url =~ m@imgur.com/[^\.]+\.@i;
        scrape_album($url, $arg) if $url =~ m@imgur.com/(?:a/)?[^\.]+$@i;
        }
}

sub download_image {
    my ( $url, $arg ) = @_;
    my ($file_name) = ($url =~ m@imgur\.com/(?:a/)?(.+)$@i);
    ( -e "downloads/$arg/$file_name" )
      ? print "     Skipping $url (copy found)\n"
      : print "     Downloading $url => downloads/$arg/$file_name\n";
    $mech->get( $url, ':content_file' => "downloads/$arg/$file_name" );
    print "      Done.\n";
}

sub scrape_album {
    my ( $url, $arg ) = @_;
    print "     Found album $url\n";

    $mech->get($url);
    if ( my ($json) = ( $mech->content =~ m@images:\s+(.+?),\s*cover:@is ) ) {
        my $struct = JSON->new->allow_nonref->utf8->relaxed->decode($json);
        map {
            download_image( 'http://imgur.com/' . $_->{hash} . $_->{ext}, $arg )
        } @{ $struct->{items} };
    }
    else {    #otherwise, we've just found a page like http://imgur.com/aOLy2
        my ($img) = ($url =~ m@imgur\.com/([^\.#]+)@i);
        return unless $img;
        $mech->follow_link( url_regex => qr/$img/i );
        download_image( $mech->uri, $arg );
    }
}