annotate lisp/soundex.el @ 11002:ff115809a39e

(Fash): Fix previous change.
author Richard M. Stallman <rms@gnu.org>
date Tue, 14 Mar 1995 18:38:28 +0000
parents ba1bbdb8595e
children 83f275dcd93a
Ignore whitespace changes - Everywhere: Within whitespace: At end of lines:
rev   line source
5995
a24f3890171e Initial revision
Richard M. Stallman <rms@gnu.org>
parents:
diff changeset
1 ;;; soundex.el -- implement Soundex algorithm
a24f3890171e Initial revision
Richard M. Stallman <rms@gnu.org>
parents:
diff changeset
2
a24f3890171e Initial revision
Richard M. Stallman <rms@gnu.org>
parents:
diff changeset
3 ;; Copyright (C) 1993 Free Software Foundation, Inc.
a24f3890171e Initial revision
Richard M. Stallman <rms@gnu.org>
parents:
diff changeset
4
a24f3890171e Initial revision
Richard M. Stallman <rms@gnu.org>
parents:
diff changeset
5 ;; Author: Christian Plaunt <chris@bliss.berkeley.edu>
a24f3890171e Initial revision
Richard M. Stallman <rms@gnu.org>
parents:
diff changeset
6 ;; Maintainer: FSF
a24f3890171e Initial revision
Richard M. Stallman <rms@gnu.org>
parents:
diff changeset
7 ;; Keywords: matching
a24f3890171e Initial revision
Richard M. Stallman <rms@gnu.org>
parents:
diff changeset
8 ;; Created: Sat May 15 14:48:18 1993
a24f3890171e Initial revision
Richard M. Stallman <rms@gnu.org>
parents:
diff changeset
9
a24f3890171e Initial revision
Richard M. Stallman <rms@gnu.org>
parents:
diff changeset
10 ;; This file is part of GNU Emacs.
a24f3890171e Initial revision
Richard M. Stallman <rms@gnu.org>
parents:
diff changeset
11
a24f3890171e Initial revision
Richard M. Stallman <rms@gnu.org>
parents:
diff changeset
12 ;; GNU Emacs is free software; you can redistribute it and/or modify
a24f3890171e Initial revision
Richard M. Stallman <rms@gnu.org>
parents:
diff changeset
13 ;; it under the terms of the GNU General Public License as published by
a24f3890171e Initial revision
Richard M. Stallman <rms@gnu.org>
parents:
diff changeset
14 ;; the Free Software Foundation; either version 2, or (at your option)
a24f3890171e Initial revision
Richard M. Stallman <rms@gnu.org>
parents:
diff changeset
15 ;; any later version.
a24f3890171e Initial revision
Richard M. Stallman <rms@gnu.org>
parents:
diff changeset
16
a24f3890171e Initial revision
Richard M. Stallman <rms@gnu.org>
parents:
diff changeset
17 ;; GNU Emacs is distributed in the hope that it will be useful,
a24f3890171e Initial revision
Richard M. Stallman <rms@gnu.org>
parents:
diff changeset
18 ;; but WITHOUT ANY WARRANTY; without even the implied warranty of
a24f3890171e Initial revision
Richard M. Stallman <rms@gnu.org>
parents:
diff changeset
19 ;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
a24f3890171e Initial revision
Richard M. Stallman <rms@gnu.org>
parents:
diff changeset
20 ;; GNU General Public License for more details.
a24f3890171e Initial revision
Richard M. Stallman <rms@gnu.org>
parents:
diff changeset
21
a24f3890171e Initial revision
Richard M. Stallman <rms@gnu.org>
parents:
diff changeset
22 ;; You should have received a copy of the GNU General Public License
a24f3890171e Initial revision
Richard M. Stallman <rms@gnu.org>
parents:
diff changeset
23 ;; along with GNU Emacs; see the file COPYING. If not, write to
a24f3890171e Initial revision
Richard M. Stallman <rms@gnu.org>
parents:
diff changeset
24 ;; the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA.
a24f3890171e Initial revision
Richard M. Stallman <rms@gnu.org>
parents:
diff changeset
25
a24f3890171e Initial revision
Richard M. Stallman <rms@gnu.org>
parents:
diff changeset
26 ;;; Commentary:
a24f3890171e Initial revision
Richard M. Stallman <rms@gnu.org>
parents:
diff changeset
27
a24f3890171e Initial revision
Richard M. Stallman <rms@gnu.org>
parents:
diff changeset
28 ;; The Soundex algorithm maps English words into representations of
a24f3890171e Initial revision
Richard M. Stallman <rms@gnu.org>
parents:
diff changeset
29 ;; how they sound. Words with vaguely similar sound map to the same string.
a24f3890171e Initial revision
Richard M. Stallman <rms@gnu.org>
parents:
diff changeset
30
a24f3890171e Initial revision
Richard M. Stallman <rms@gnu.org>
parents:
diff changeset
31 ;;; Code:
a24f3890171e Initial revision
Richard M. Stallman <rms@gnu.org>
parents:
diff changeset
32
7534
9b82dae27c01 (soundex-alist): Put variable before fn that uses it.
Richard M. Stallman <rms@gnu.org>
parents: 5995
diff changeset
33 (defvar soundex-alist
8028
ba1bbdb8595e (soundex-alist): Delete the elements that mapped into nil.
Richard M. Stallman <rms@gnu.org>
parents: 7534
diff changeset
34 '((?B . "1") (?F . "1") (?P . "1") (?V . "1")
7534
9b82dae27c01 (soundex-alist): Put variable before fn that uses it.
Richard M. Stallman <rms@gnu.org>
parents: 5995
diff changeset
35 (?C . "2") (?G . "2") (?J . "2") (?K . "2") (?Q . "2") (?S . "2")
9b82dae27c01 (soundex-alist): Put variable before fn that uses it.
Richard M. Stallman <rms@gnu.org>
parents: 5995
diff changeset
36 (?X . "2") (?Z . "2") (?D . "3") (?T . "3") (?L . "4") (?M . "5")
9b82dae27c01 (soundex-alist): Put variable before fn that uses it.
Richard M. Stallman <rms@gnu.org>
parents: 5995
diff changeset
37 (?N . "5") (?R . "6"))
9b82dae27c01 (soundex-alist): Put variable before fn that uses it.
Richard M. Stallman <rms@gnu.org>
parents: 5995
diff changeset
38 "Alist of chars-to-key-code for building Soundex keys.")
9b82dae27c01 (soundex-alist): Put variable before fn that uses it.
Richard M. Stallman <rms@gnu.org>
parents: 5995
diff changeset
39
5995
a24f3890171e Initial revision
Richard M. Stallman <rms@gnu.org>
parents:
diff changeset
40 (defun soundex (word)
a24f3890171e Initial revision
Richard M. Stallman <rms@gnu.org>
parents:
diff changeset
41 "Return a Soundex key for WORD.
a24f3890171e Initial revision
Richard M. Stallman <rms@gnu.org>
parents:
diff changeset
42 Implemented as described in:
a24f3890171e Initial revision
Richard M. Stallman <rms@gnu.org>
parents:
diff changeset
43 Knuth, Donald E. \"The Art of Computer Programming, Vol. 3: Sorting
a24f3890171e Initial revision
Richard M. Stallman <rms@gnu.org>
parents:
diff changeset
44 and Searching\", Addison-Wesley (1973), pp. 391-392."
a24f3890171e Initial revision
Richard M. Stallman <rms@gnu.org>
parents:
diff changeset
45 (let* ((word (upcase word)) (length (length word))
a24f3890171e Initial revision
Richard M. Stallman <rms@gnu.org>
parents:
diff changeset
46 (code (cdr (assq (aref word 0) soundex-alist)))
a24f3890171e Initial revision
Richard M. Stallman <rms@gnu.org>
parents:
diff changeset
47 (key (substring word 0 1)) (index 1) (prev-code code))
8028
ba1bbdb8595e (soundex-alist): Delete the elements that mapped into nil.
Richard M. Stallman <rms@gnu.org>
parents: 7534
diff changeset
48 ;; once we have a four char key, we're done
5995
a24f3890171e Initial revision
Richard M. Stallman <rms@gnu.org>
parents:
diff changeset
49 (while (and (> 4 (length key)) (< index length))
a24f3890171e Initial revision
Richard M. Stallman <rms@gnu.org>
parents:
diff changeset
50 ;; look up the code for each letter in word at index
a24f3890171e Initial revision
Richard M. Stallman <rms@gnu.org>
parents:
diff changeset
51 (setq code (cdr (assq (aref word index) soundex-alist))
a24f3890171e Initial revision
Richard M. Stallman <rms@gnu.org>
parents:
diff changeset
52 index (1+ index)
a24f3890171e Initial revision
Richard M. Stallman <rms@gnu.org>
parents:
diff changeset
53 ;; append code to key unless the same codes belong to
a24f3890171e Initial revision
Richard M. Stallman <rms@gnu.org>
parents:
diff changeset
54 ;; adjacent letters in the original string
a24f3890171e Initial revision
Richard M. Stallman <rms@gnu.org>
parents:
diff changeset
55 key (concat key (if (or (null code) (string= code prev-code))
a24f3890171e Initial revision
Richard M. Stallman <rms@gnu.org>
parents:
diff changeset
56 ()
a24f3890171e Initial revision
Richard M. Stallman <rms@gnu.org>
parents:
diff changeset
57 code))
a24f3890171e Initial revision
Richard M. Stallman <rms@gnu.org>
parents:
diff changeset
58 prev-code code))
a24f3890171e Initial revision
Richard M. Stallman <rms@gnu.org>
parents:
diff changeset
59 ;; return a key that is 4 chars long and padded by "0"s if needed
a24f3890171e Initial revision
Richard M. Stallman <rms@gnu.org>
parents:
diff changeset
60 (if (> 4 (length key))
a24f3890171e Initial revision
Richard M. Stallman <rms@gnu.org>
parents:
diff changeset
61 (substring (concat key "000") 0 4)
a24f3890171e Initial revision
Richard M. Stallman <rms@gnu.org>
parents:
diff changeset
62 key)))
a24f3890171e Initial revision
Richard M. Stallman <rms@gnu.org>
parents:
diff changeset
63
a24f3890171e Initial revision
Richard M. Stallman <rms@gnu.org>
parents:
diff changeset
64 ;(defvar soundex-test
a24f3890171e Initial revision
Richard M. Stallman <rms@gnu.org>
parents:
diff changeset
65 ; '("Euler" "Gauss" "Hilbert" "Knuth" "Lloyd" "Lukasiewicz"
a24f3890171e Initial revision
Richard M. Stallman <rms@gnu.org>
parents:
diff changeset
66 ; "Ellery" "Ghosh" "Heilbronn" "Kant" "Ladd" "Lissajous")
a24f3890171e Initial revision
Richard M. Stallman <rms@gnu.org>
parents:
diff changeset
67 ; "\n Knuth's names to demonstrate the Soundex algorithm.")
a24f3890171e Initial revision
Richard M. Stallman <rms@gnu.org>
parents:
diff changeset
68 ;
a24f3890171e Initial revision
Richard M. Stallman <rms@gnu.org>
parents:
diff changeset
69 ;(mapcar 'soundex soundex-test)
a24f3890171e Initial revision
Richard M. Stallman <rms@gnu.org>
parents:
diff changeset
70 ;("E460" "G200" "H416" "K530" "L300" "L222"
a24f3890171e Initial revision
Richard M. Stallman <rms@gnu.org>
parents:
diff changeset
71 ; "E460" "G200" "H416" "K530" "L300" "L222")
a24f3890171e Initial revision
Richard M. Stallman <rms@gnu.org>
parents:
diff changeset
72
a24f3890171e Initial revision
Richard M. Stallman <rms@gnu.org>
parents:
diff changeset
73 ;; soundex.el ends here