#!/usr/bin/perl -w
# 
# ao.pl
# 
# 'Auto-obfuscator', or simple text compressor, for C programs.
# 
# Replaces commonly-used long words with an abbreviation (which is
# the first 1 or more characters from the beginning), and adds
# '#define ABBR WORD' at the top of the file.  Running the output
# through the C preprocessor will give back the same file, although
# whitespace will be different.
# 
# Bugs: has absolutely no idea about strings.
# 
# -- Ed Avis, epa98@doc.ic.ac.uk, 1999-09-12
# 

while (<>) {
    while (length) {
	s/^(\w+)// or s/^(\W+)// or die;
	push @{$l[$.]}, $1;
    }
}

foreach $l (1 .. $#l) {
    foreach (grep /^[a-zA-Z_]/, @{$l[$l]}) {
	$c{$_}++;
	$used{$_}++;
    }
}

while (($word, $freq) = each %c) {
    next if $freq == 1;
#    print "word $word appears $freq times, ";
    if (defined($abbrev = getabbrev($word))) {
	$al = (length "#define $word $abbrev\n") + (length $abbrev) * $freq;
	$l = (length $word) * $freq;
	if ($al < $l) {
	    $used{$abbrev}++;
	    $abbrev{$word} = $abbrev;
	    print "#define $abbrev $word\n";
#           print "abbrev is $abbrev\n";
	}
#       else { print "abbrev is $abbrev, but not used\n"; }
    }
#    else { print "no abbrev\n" }
}

foreach $l (1 .. $#l) {
    foreach (@{$l[$l]}) {
	print defined $abbrev{$_} ? $abbrev{$_} : $_;
    }
}

sub getabbrev($) {
    my ($w) = @_;
    foreach my $l (1 .. (length $w) - 1) {
	my $a = substr($w, 0, $l);
	return $a unless $used{$a};
    }
    return undef;
}