/usr/share/festival/tokenize.scm is in festival-freebsoft-utils 0.10-3.
This file is owned by root:root, with mode 0o644.
The actual contents of the file can be viewed below.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 | ;;; Tokenization and utterance chunking
;; Copyright (C) 2004, 2005, 2006 Brailcom, o.p.s.
;; Author: Milan Zamazal <pdm@brailcom.org>
;; COPYRIGHT NOTICE
;; This program is free software; you can redistribute it and/or modify
;; it under the terms of the GNU General Public License as published by
;; the Free Software Foundation; either version 2 of the License, or
;; (at your option) any later version.
;; This program is distributed in the hope that it will be useful, but
;; WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY
;; or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
;; for more details.
;; You should have received a copy of the GNU General Public License
;; along with this program; if not, write to the Free Software
;; Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA.
(require 'tts)
(require 'spell-mode)
(require 'util)
(defvar max-number-of-tokens 100)
(defvar max-number-of-token-chars 100)
(define (token-utterance)
(let ((utt (Utterance Tokens)))
(utt.relation.create utt 'Token)))
(define (token-utterance-append utt name whitespace punc prepunc)
(token-utterance-append* utt name whitespace punc prepunc 'Token))
(define (token-utterance-append* utt name whitespace punc prepunc relation)
(utt.relation.append utt relation
`(,name
((name ,name) (whitespace ,whitespace)
(punc ,punc) (prepunctuation ,prepunc)))))
(define (get-regular-token utt text)
(let* ((l-whitespace (symbolexplode token.whitespace))
(l-punc (symbolexplode token.punctuation))
(l-prepunctuation (symbolexplode token.prepunctuation))
(r-whitespace (lambda (c) (member_string c l-whitespace)))
(r-punc (lambda (c) (member_string c l-punc)))
(r-prepunctuation (lambda (c) (member_string c l-prepunctuation)))
(r-name (lambda (c)
(not (or (r-whitespace c) (r-punc c)))))
(name "")
(whitespace "")
(punc "")
(prepunctuation "")
(oversized nil))
(let ((read-part (lambda (matcher)
(let ((result-chars '())
(nchars 0)
(char (substring text 0 1)))
(while (and (not (string-equal char ""))
(not oversized)
(matcher char))
(set! result-chars (cons char result-chars))
(set! text (substring text 1 (- (length text) 1)))
(set! nchars (+ nchars 1))
(when (>= nchars max-number-of-token-chars)
(set! oversized t))
(set! char (substring text 0 1)))
(apply string-append (reverse result-chars))))))
(set! whitespace (read-part r-whitespace))
(set! prepunctuation (read-part r-prepunctuation))
(set! name (read-part r-name))
(set! punc (read-part r-punc))
(while (and (not oversized)
(not (string-equal text ""))
(not (r-whitespace (substring text 0 1))))
(set! name (string-append name punc (read-part r-name)))
(set! punc (read-part r-punc))
(when (>= (length name) max-number-of-token-chars)
(set! oversized t)))
;; Presence of empty words is questionable and the built-in tokenization
;; process doesn't generate them
(when (string-equal name "")
(cond
(oversized
nil)
((not (string-equal punc ""))
(set! name (substring punc 0 1))
(set! punc (substring punc 1 (- (length punc) 1))))
((not (string-equal prepunctuation ""))
(set! name (substring prepunctuation 0 1))
(set! prepunctuation
(substring prepunctuation 1 (- (length prepunctuation) 1))))
(t ; final whitespace -- ignored
(set! text "")))))
(unless (string-equal name "")
(token-utterance-append utt name whitespace punc prepunctuation))
text))
(define (get-single-char-token utt text)
(let ((name (substring text 0 1)))
(set! text (substring text 1 (length text)))
(unless (string-equal name "")
(token-utterance-append utt name "" "" "")))
text)
(define (get-token utt text)
((if spell-mode get-single-char-token get-regular-token) utt text))
(define (next-chunk text)
(let ((utt (token-utterance))
(finished nil)
(ntokens 1))
(while (not finished)
(let* ((new-text (get-token utt text))
(token (utt.relation.last utt 'Token))
(ptoken (item.prev token)))
(cond
((string-equal new-text "")
(set! text "")
(set! finished t))
((and ptoken (string-equal (wagon_predict ptoken eou_tree) 1))
(item.delete token)
(set! finished t))
((>= ntokens max-number-of-tokens)
;; This could be handled (probably less efficiently) by eou_tree, but
;; we don't want to touch the tree just because of this
(set! text new-text)
(set! finished t))
(t
(set! text new-text)
(set! ntokens (+ ntokens 1))))))
(list utt text)))
(define (append-token-utterances . utts)
(let ((u (token-utterance)))
(while utts
(let ((i (utt.relation (car utts) 'Token)))
(while i
(token-utterance-append u (item.name i)
(item.feat i 'whitespace)
(item.feat i 'punc)
(item.feat i 'prepunctuation))
(set! i (item.next i))))
(set! utts (cdr utts))))
u)
(provide 'tokenize)
|