6 @EXPORT = qw(&soundex $soundex_nocode);
10 # $Id: soundex.pl,v 1.2 1994/03/24 00:30:27 mike Exp $
12 # Implementation of soundex algorithm as described by Knuth in volume
13 # 3 of The Art of Computer Programming, with ideas stolen from Ian
14 # Phillips <ian@pipex.net>.
16 # Mike Stok <Mike.Stok@meiko.concord.ma.us>, 2 March 1994.
18 # Knuth's test cases are:
20 # Euler, Ellery -> E460
21 # Gauss, Ghosh -> G200
22 # Hilbert, Heilbronn -> H416
25 # Lukasiewicz, Lissajous -> L222
27 # $Log: soundex.pl,v $
28 # Revision 1.2 1994/03/24 00:30:27 mike
29 # Subtle bug (any excuse :-) spotted by Rich Pinder <rpinder@hsc.usc.edu>
30 # in the way I handles leasing characters which were different but had
31 # the same soundex code. This showed up comparing it with Oracle's
34 # Revision 1.1 1994/03/02 13:01:30 mike
38 ##############################################################################
40 # $soundex_nocode is used to indicate a string doesn't have a soundex
41 # code, I like undef other people may want to set it to 'Z000'.
43 $soundex_nocode = undef;
47 local (@s, $f, $fc, $_) = @_;
49 push @s, '' unless @s; # handle no args as a single empty string
63 tr/AEHIOUWYBFPVCGJKQSXZDTLMNR/00000000111122222222334556/;
73 wantarray ? @s : shift @s;
82 Text::Soundex - Implementation of the Soundex Algorithm as Described by Knuth
88 $code = soundex $string; # get soundex code for a string
89 @codes = soundex @list; # get list of codes for list of strings
91 # set value to be returned for strings without soundex code
93 $soundex_nocode = 'Z000';
97 This module implements the soundex algorithm as described by Donald Knuth
98 in Volume 3 of B<The Art of Computer Programming>. The algorithm is
99 intended to hash words (in particular surnames) into a small space using a
100 simple model which approximates the sound of the word when spoken by an English
101 speaker. Each word is reduced to a four character string, the first
102 character being an upper case letter and the remaining three being digits.
104 If there is no soundex code representation for a string then the value of
105 C<$soundex_nocode> is returned. This is initially set to C<undef>, but
106 many people seem to prefer an I<unlikely> value like C<Z000>
107 (how unlikely this is depends on the data set being dealt with.) Any value
108 can be assigned to C<$soundex_nocode>.
110 In scalar context C<soundex> returns the soundex code of its first
111 argument, and in list context a list is returned in which each element is the
112 soundex code for the corresponding argument passed to C<soundex> e.g.
114 @codes = soundex qw(Mike Stok);
116 leaves C
<@codes> containing C
<('M200', 'S320')>.
120 Knuth's examples of various names and the soundex codes they map to
123 Euler, Ellery -> E460
125 Hilbert, Heilbronn -> H416
128 Lukasiewicz, Lissajous -> L222
132 $code = soundex 'Knuth'; # $code contains 'K530'
133 @list = soundex qw(Lloyd Gauss); # @list contains 'L300', 'G200'
137 As the soundex algorithm was originally used a B<long> time ago in the US
138 it considers only the English alphabet and pronunciation.
140 As it is mapping a large space (arbitrary length strings) onto a small
141 space (single letter plus 3 digits) no inference can be made about the
142 similarity of two strings which end up with the same soundex code. For
143 example, both C<Hilbert> and C<Heilbronn> end up with a soundex code
148 This code was implemented by Mike Stok (C<stok@cybercom.net>) from the
149 description given by Knuth. Ian Phillips (C<ian@pipex.net>) and Rich Pinder
150 (C<rpinder@hsc.usc.edu>) supplied ideas and spotted mistakes.