#!/usr/bin/env perl
use strict;
use CGI qw/param/;
use URI;

my $comic = param("c");
my $title;
my $url = "";
my $images = "";

sub get {
	my ($url) = @_;
	my $ua = "Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.4b) Gecko/20030525 Mozilla Firebird/0.6";
	return scalar `/opt/bin/curl -s -A "$ua" "$url"`;
}

sub page {
	$url = $_[0];
	$_ = get($url);
	die "can't get page" unless defined $_;
}

my %imgs_seen = ();
sub img {
	my ($img, $tag) = @_;
	return if defined $imgs_seen{$img};
	$imgs_seen{$img} = 1;
	$img = URI->new_abs($img, $url)->as_string;
	my $a = "";
	$a = " alt=\"$tag\"" if defined $tag;
	$images .= "<img src=\"$img\"$a />";
}

die "no comic specified" unless defined $comic;
if ($comic eq "userfriendly") {
	# UF are now doing Referer detection.
	$title = "User Friendly";
	page("http://www.userfriendly.org/");
	/<IMG ALT="Latest Strip"[^>]* SRC="([^"]*)"/is or die "no match";
	img($1);
} elsif ($comic eq "sluggy") {
	$title = "Sluggy Freelance";
	page("http://www.sluggy.com/");
	while (/<img alt="Sluggy Freelance &copy; Pete Abrams[^>]* src="([^"]*)"/gis) {
		img($1);
	}
} elsif ($comic eq "angsttech") {
	$title = "Angst Technology";
	page("http://www.inktank.com/AT/index.cfm");
	/<IMG SRC="(\/images\/AT\/cartoons\/[^"]*\.gif)"/is or die "no match";
	img($1);
} elsif ($comic eq "redmeat") {
	$title = "Red Meat";
	page("http://www.redmeat.com/redmeat/current/index.html");
	/<a href="..\/(....-..-..)\/index\.html">previous<\/a>/is or die "no match";
	img("index-1.gif", $1);
} elsif ($comic eq "megatokyo") {
	$title = "Megatokyo";
	page("http://www.megatokyo.com/");
	/<IMG SRC="(\/strips\/[^"]*\.gif)"/is or die "no match";
	img($1);
} elsif ($comic eq "bruno") {
	$title = "Bruno";
	page("http://www.brunostrip.com/bruno.html");
	/<img src="(\d+pics\/\d+\.gif)">/is or die "no match";
	img($1);
} elsif ($comic eq "helen") {
	# This doesn't work.
	$title = "Helen, Sweetheart of the Internet";
	page("http://www.comicspage.com/comicspage/main.jsp?custid=69&catid=1242&dir=/helen");
	/<img src="(http:\/\/www.tmsfeatures.com\/tmsfeatures\/servlet\/com.featureserv.util.Download\?file=[^"]*code=cshl)"/is or die "no match";
	img($1);
} elsif ($comic eq "absurdnotions") {
	$title = "Absurd Notions";
	page("http://www.absurdnotions.org/");
	/<IMG SRC="(an[^"]+\.gif)"[^>]* ALT="Comic for/is or die "no match";
	img($1);
} elsif ($comic eq "longface") {
	$title = "Why The Long Face";
	page("http://www.absurdnotions.org/");
	/<IMG SRC="(lf[^"]+\.gif)"[^>]* ALT="Comic for/is and img($1);
} elsif ($comic eq "boxjam") {
	$title = "Boxjam's Doodle";
	page("http://www.boxjamsdoodle.com/");
	/<IMG [^>]* SRC="(\/comics\/[^"]+)"/is or die "no match";
	img($1);
} elsif ($comic eq "sinfest") {
	$title = "Sinfest";
	page("http://www.sinfest.net/");
	/<IMG ALT="comic"[^>]* SRC="(\/comics\/[^"]+)"/is or die "no match";
	img($1);
} elsif ($comic eq "dilbert") {
	$title = "Dilbert";
	page("http://www.unitedmedia.com/comics/dilbert/index.html");
	/<IMG SRC="([^"]+)"[^>]* ALT="Today's Dilbert Comic">/is or die "no match";
	img($1);
} elsif ($comic eq "sheldon") {
	$title = "Sheldon";
	page("http://www.sheldoncomics.com/");
	/<IMG SRC="([^"]+)"[^>]* ALT="Today's Comic"/is or die "no match";
	img($1);
} elsif ($comic eq "nukees") {
	$title = "Nukees";
	page("http://www.nukees.com/");
	/<IMG ALT="comic"[^>]* SRC="(\/comics\/[^"]+)"/is or die "no match";
	img($1);
} elsif ($comic eq "gpf") {
	$title = "General Protection Fault";
	page("http://www.gpf-comics.com/");
	/<IMG [^>]* SRC="(\/comics\/[^"]+)"/is or die "no match";
	img($1);
} elsif ($comic eq "errantstory") {
	$title = "Errant Story";
	page("http://www.errantstory.com/");
	if (/<img .*src="([^"]*\/images\/[^"]*cover[^"]*)"/is) {
	} elsif (/<IMG ALT="comic"[^>]* SRC="(\/comics\/[^"]+)"/is) {
	} else {
		die "no match";
	}
	img($1);
} elsif ($comic eq "reallife") {
	$title = "Real Life";
	page("http://www.reallifecomics.com/");
	/src='(\/index\.php\?do_command=show_strip[^']+)'/is or die "no match";
	# Growling a bit at this, but it's more work than I'm willing to put in
	# to make it possible to directly link to their image.
	img("http://www.reallifecomics.com/aqua/logo.gif", $1);
} elsif ($comic eq "wapsisquare") {
	$title = "Wapsi Square";
	page("http://wapsisquare.com/");
	/<IMG [^>]* SRC="(\/comics\/[^"]+)"/is or die "no match";
	img($1);
} elsif ($comic eq "phd") {
	$title = "PhD";
	page("http://www.phdcomics.com/comics/comics.php");
	/<img src=(archive\/phd[^ ]+) border=/is or die "no match";
	img($1);
} elsif ($comic eq "flea") {
	$title = "Flea";
	page("http://www.flea-toon.com/");
	/<image src="(flea\d+\.gif)"/ or die "no match";
	img($1);
} elsif ($comic eq "freefall") {
	$title = "Freefall";
	page("http://freefall.purrsia.com/default.htm");
	/<img src="(\/ff...\/fv......gif)"/ or die "no match";
	img($1);
} else {
	die "unknown comic";
}

print "Content-type: application/rdf+xml\r\n\r\n";
print <<EOF;
<?xml version="1.0" encoding="iso-8859-1"?>
<rss version="0.91">
<channel>
<title>$title</title>
<link>$url</link>
<description>$title</description>
<item>
<title>$title</title>
<link>$url</link>
<description>$images</description>
</item>
</rss>
EOF