/ 
Text-Rhyme-0.01
/ Text::Rhymes

NAME

Text::Rhyme - very basic test for rhymes across two words.

SYNOPSIS

use Text::Rhyme;

my @words = (
	'stocks','box', 'commotion','dennotion',
	'spoke','spake', 'foo','bar'
);

for (my $i=0; $i<=$#words; $i+=2) {
	my $rhyme_val = Text::Rhyme::test($words[$i],$words[$i+1]);
	print "$words[$i] and $words[$i+1] = $rhyme_val - ";
	if ($rhyme_val == 0) { print "as good as it gets, I'm afraid." }
	elsif ($rhyme_val<3) { print "well...not good...." }
	else { print "terrible!" }
	print "\n";
}

DESCRIPTION

Mike Stok, who implemented perl's Text::Soundex writes in that modules POD:

As [I<Text::Soundex>] is mapping a large space (arbitrary length strings) onto a small
space (single letter plus 3 digits) no inference can be made about the
similarity of two strings which end up with the same soundex code. ...

It would seem that the smaller the space mapped, the greater the chance of the soundex code output reflecting the input string. This module therefore considers only a small part of a word when attempting to find a rhyme.

Common endings that produce glib rhymes ('~tion') are cut if shared by both words.

Only the final groups of vowels (preceding trailing consonants) are considered.

Each is passed to the Text::Soundex routine, which returns a code of one letter and three digits.

The module returns the difference between the soundex values of the extracted vowels, as a positive integer.

LIMITATIONS

They're huge, I'm afraid: when tested, 'spoke' and 'spake' return 0, which is plainly wrong.

Any thoughts/suggestions/mods much appreciated.

EXPANSION

Any thoughts/suggestions/mods much appreciated.

Word stress patterns could help to produce a better analysis, but we need a means of identifying syllables, and sadly Text::Syllable isn't up to it at the date of writing.

Any thoughts/suggestions/mods much appreciated....

SEE ALSO

L<Text::Soundex>

AUTHOR

Lee Goddard <lgoddard@cpan.org>

COPYRIGHT

Copyright (C) Lee Goddard, 2001. All Rights Reserved.

This is free software which you may use under the same terms as Perl itself.