#!/usr/local/bin/perl -W --
# $Id: ungoopspam,v 1.4 2002/01/31 09:20:40 chip Exp $
# $Source: /home/chip/src/ungoopspam/RCS/ungoopspam,v $

use HTML::Entities;
use MIME::QuotedPrint;

# slurp in entire document
$_ = join("", <>);

# undo quoted-printable (based on dumb heuristic)
$_ = decode_qp($_) . "\n"
	if (/=\n/);

# decode &entities;
$_ = decode_entities($_);

# translate %nn hex escapes (this should be limited to URIs)
while (m/%([0-9a-f][0-9a-f])/i) {
	$t = chr(hex($1));
	# print STDERR "changing \"%$1\" -> \"$t\"\n";
	s/%$1/$t/gi;
}

print $_;

__END__

=head1 NAME

ungoopspam - clarify HTML obfuscations popular with spammers

=head1 SYNOPSIS

B<ungoopspam> [file ...]

=head1 DESCRIPTION

The B<ungoopspam> utility reads an HTML document from I<file> (or standard
input if no files are specified), and performs a number of transformations.

=over 4

=item *

The document is converted from quoted-printable encoding to plain 8-bit
text.  This may have happened if the spammer attached the web page
using QP format.  It's actually a function of the mail system, not HTML.

=item *

Any HTML entities, such as "&quot;" or "&#65;" are converted to their
8-bit character equivalent.

=item *

Hex encoding ("%nn") in URIs is converted to its 8-bit character
equivalent.

=back

=head1 BUGS

This utility is a hack.  It is not a rugged tool built for a production
environment.  It exists to assist you in decoding the obfuscated web
pages favored by spammers.

Quoted-printable is recognized by a low-rent heuristic.

The "%nn" hex character decoding is performed throughout the document,
not just in URIs.  This may lead to undesired results.

=head1 SEE ALSO

ascii(7)

=head1 AUTHOR

 Chip Rosenthal
 Unicom Systems Development
 <chip@unicom.com>

 $Id: ungoopspam,v 1.4 2002/01/31 09:20:40 chip Exp $
 See http://www.unicom.com/sw/ungoopspam/ for latest version.