1.1 --- /dev/null Thu Jan 01 00:00:00 1970 +0000
1.2 +++ b/extensions/spellcheck/hunspell/src/phonet.cpp Wed Dec 31 06:09:35 2014 +0100
1.3 @@ -0,0 +1,309 @@
1.4 +/******* BEGIN LICENSE BLOCK *******
1.5 + * Version: MPL 1.1/GPL 2.0/LGPL 2.1
1.6 + *
1.7 + * The contents of this file are subject to the Mozilla Public License Version
1.8 + * 1.1 (the "License"); you may not use this file except in compliance with
1.9 + * the License. You may obtain a copy of the License at
1.10 + * http://www.mozilla.org/MPL/
1.11 + *
1.12 + * Software distributed under the License is distributed on an "AS IS" basis,
1.13 + * WITHOUT WARRANTY OF ANY KIND, either express or implied. See the License
1.14 + * for the specific language governing rights and limitations under the
1.15 + * License.
1.16 + *
1.17 + * The Initial Developer of the Original Code is Björn Jacke. Portions created
1.18 + * by the Initial Developers are Copyright (C) 2000-2007 the Initial
1.19 + * Developers. All Rights Reserved.
1.20 + *
1.21 + * Contributor(s): Björn Jacke (bjoern.jacke@gmx.de)
1.22 + * László Németh (nemethl@gyorsposta.hu)
1.23 + * Caolan McNamara (caolanm@redhat.com)
1.24 + *
1.25 + * Alternatively, the contents of this file may be used under the terms of
1.26 + * either the GNU General Public License Version 2 or later (the "GPL"), or
1.27 + * the GNU Lesser General Public License Version 2.1 or later (the "LGPL"),
1.28 + * in which case the provisions of the GPL or the LGPL are applicable instead
1.29 + * of those above. If you wish to allow use of your version of this file only
1.30 + * under the terms of either the GPL or the LGPL, and not to allow others to
1.31 + * use your version of this file under the terms of the MPL, indicate your
1.32 + * decision by deleting the provisions above and replace them with the notice
1.33 + * and other provisions required by the GPL or the LGPL. If you do not delete
1.34 + * the provisions above, a recipient may use your version of this file under
1.35 + * the terms of any one of the MPL, the GPL or the LGPL.
1.36 + *
1.37 + * Changelog:
1.38 + * 2000-01-05 Björn Jacke <bjoern.jacke AT gmx.de>
1.39 + * Initial Release insprired by the article about phonetic
1.40 + * transformations out of c't 25/1999
1.41 + *
1.42 + * 2007-07-26 Björn Jacke <bjoern.jacke AT gmx.de>
1.43 + * Released under MPL/GPL/LGPL tri-license for Hunspell
1.44 + *
1.45 + * 2007-08-23 László Németh <nemeth at OOo>
1.46 + * Porting from Aspell to Hunspell using C-like structs
1.47 + *
1.48 + ******* END LICENSE BLOCK *******/
1.49 +
1.50 +#include <stdlib.h>
1.51 +#include <string.h>
1.52 +#include <stdio.h>
1.53 +#include <ctype.h>
1.54 +
1.55 +#include "csutil.hxx"
1.56 +#include "phonet.hxx"
1.57 +
1.58 +void init_phonet_hash(phonetable & parms)
1.59 + {
1.60 + int i, k;
1.61 +
1.62 + for (i = 0; i < HASHSIZE; i++) {
1.63 + parms.hash[i] = -1;
1.64 + }
1.65 +
1.66 + for (i = 0; parms.rules[i][0] != '\0'; i += 2) {
1.67 + /** set hash value **/
1.68 + k = (unsigned char) parms.rules[i][0];
1.69 +
1.70 + if (parms.hash[k] < 0) {
1.71 + parms.hash[k] = i;
1.72 + }
1.73 + }
1.74 + }
1.75 +
1.76 +// like strcpy but safe if the strings overlap
1.77 +// but only if dest < src
1.78 +static inline void strmove(char * dest, char * src) {
1.79 + while (*src)
1.80 + *dest++ = *src++;
1.81 + *dest = '\0';
1.82 +}
1.83 +
1.84 +static int myisalpha(char ch) {
1.85 + if ((unsigned char) ch < 128) return isalpha(ch);
1.86 + return 1;
1.87 +}
1.88 +
1.89 +/* phonetic transcription algorithm */
1.90 +/* see: http://aspell.net/man-html/Phonetic-Code.html */
1.91 +/* convert string to uppercase before this call */
1.92 +int phonet (const char * inword, char * target,
1.93 + int len,
1.94 + phonetable & parms)
1.95 + {
1.96 + /** Do phonetic transformation. **/
1.97 + /** "len" = length of "inword" incl. '\0'. **/
1.98 +
1.99 + /** result: >= 0: length of "target" **/
1.100 + /** otherwise: error **/
1.101 +
1.102 + int i,j,k=0,n,p,z;
1.103 + int k0,n0,p0=-333,z0;
1.104 + char c, c0;
1.105 + const char * s;
1.106 + typedef unsigned char uchar;
1.107 + char word[MAXPHONETUTF8LEN + 1];
1.108 + if (len == -1) len = strlen(inword);
1.109 + if (len > MAXPHONETUTF8LEN) return 0;
1.110 + strcpy(word, inword);
1.111 +
1.112 + /** check word **/
1.113 + i = j = z = 0;
1.114 + while ((c = word[i]) != '\0') {
1.115 + n = parms.hash[(uchar) c];
1.116 + z0 = 0;
1.117 +
1.118 + if (n >= 0) {
1.119 + /** check all rules for the same letter **/
1.120 + while (parms.rules[n][0] == c) {
1.121 +
1.122 + /** check whole string **/
1.123 + k = 1; /** number of found letters **/
1.124 + p = 5; /** default priority **/
1.125 + s = parms.rules[n];
1.126 + s++; /** important for (see below) "*(s-1)" **/
1.127 +
1.128 + while (*s != '\0' && word[i+k] == *s
1.129 + && !isdigit ((unsigned char) *s) && strchr ("(-<^$", *s) == NULL) {
1.130 + k++;
1.131 + s++;
1.132 + }
1.133 + if (*s == '(') {
1.134 + /** check letters in "(..)" **/
1.135 + if (myisalpha(word[i+k]) // ...could be implied?
1.136 + && strchr(s+1, word[i+k]) != NULL) {
1.137 + k++;
1.138 + while (*s != ')')
1.139 + s++;
1.140 + s++;
1.141 + }
1.142 + }
1.143 + p0 = (int) *s;
1.144 + k0 = k;
1.145 + while (*s == '-' && k > 1) {
1.146 + k--;
1.147 + s++;
1.148 + }
1.149 + if (*s == '<')
1.150 + s++;
1.151 + if (isdigit ((unsigned char) *s)) {
1.152 + /** determine priority **/
1.153 + p = *s - '0';
1.154 + s++;
1.155 + }
1.156 + if (*s == '^' && *(s+1) == '^')
1.157 + s++;
1.158 +
1.159 + if (*s == '\0'
1.160 + || (*s == '^'
1.161 + && (i == 0 || ! myisalpha(word[i-1]))
1.162 + && (*(s+1) != '$'
1.163 + || (! myisalpha(word[i+k0]) )))
1.164 + || (*s == '$' && i > 0
1.165 + && myisalpha(word[i-1])
1.166 + && (! myisalpha(word[i+k0]) )))
1.167 + {
1.168 + /** search for followup rules, if: **/
1.169 + /** parms.followup and k > 1 and NO '-' in searchstring **/
1.170 + c0 = word[i+k-1];
1.171 + n0 = parms.hash[(uchar) c0];
1.172 +
1.173 +// if (parms.followup && k > 1 && n0 >= 0
1.174 + if (k > 1 && n0 >= 0
1.175 + && p0 != (int) '-' && word[i+k] != '\0') {
1.176 + /** test follow-up rule for "word[i+k]" **/
1.177 + while (parms.rules[n0][0] == c0) {
1.178 +
1.179 + /** check whole string **/
1.180 + k0 = k;
1.181 + p0 = 5;
1.182 + s = parms.rules[n0];
1.183 + s++;
1.184 + while (*s != '\0' && word[i+k0] == *s
1.185 + && ! isdigit((unsigned char) *s) && strchr("(-<^$",*s) == NULL) {
1.186 + k0++;
1.187 + s++;
1.188 + }
1.189 + if (*s == '(') {
1.190 + /** check letters **/
1.191 + if (myisalpha(word[i+k0])
1.192 + && strchr (s+1, word[i+k0]) != NULL) {
1.193 + k0++;
1.194 + while (*s != ')' && *s != '\0')
1.195 + s++;
1.196 + if (*s == ')')
1.197 + s++;
1.198 + }
1.199 + }
1.200 + while (*s == '-') {
1.201 + /** "k0" gets NOT reduced **/
1.202 + /** because "if (k0 == k)" **/
1.203 + s++;
1.204 + }
1.205 + if (*s == '<')
1.206 + s++;
1.207 + if (isdigit ((unsigned char) *s)) {
1.208 + p0 = *s - '0';
1.209 + s++;
1.210 + }
1.211 +
1.212 + if (*s == '\0'
1.213 + /** *s == '^' cuts **/
1.214 + || (*s == '$' && ! myisalpha(word[i+k0])))
1.215 + {
1.216 + if (k0 == k) {
1.217 + /** this is just a piece of the string **/
1.218 + n0 += 2;
1.219 + continue;
1.220 + }
1.221 +
1.222 + if (p0 < p) {
1.223 + /** priority too low **/
1.224 + n0 += 2;
1.225 + continue;
1.226 + }
1.227 + /** rule fits; stop search **/
1.228 + break;
1.229 + }
1.230 + n0 += 2;
1.231 + } /** End of "while (parms.rules[n0][0] == c0)" **/
1.232 +
1.233 + if (p0 >= p && parms.rules[n0][0] == c0) {
1.234 + n += 2;
1.235 + continue;
1.236 + }
1.237 + } /** end of follow-up stuff **/
1.238 +
1.239 + /** replace string **/
1.240 + s = parms.rules[n+1];
1.241 + p0 = (parms.rules[n][0] != '\0'
1.242 + && strchr (parms.rules[n]+1,'<') != NULL) ? 1:0;
1.243 + if (p0 == 1 && z == 0) {
1.244 + /** rule with '<' is used **/
1.245 + if (j > 0 && *s != '\0'
1.246 + && (target[j-1] == c || target[j-1] == *s)) {
1.247 + j--;
1.248 + }
1.249 + z0 = 1;
1.250 + z = 1;
1.251 + k0 = 0;
1.252 + while (*s != '\0' && word[i+k0] != '\0') {
1.253 + word[i+k0] = *s;
1.254 + k0++;
1.255 + s++;
1.256 + }
1.257 + if (k > k0)
1.258 + strmove (&word[0]+i+k0, &word[0]+i+k);
1.259 +
1.260 + /** new "actual letter" **/
1.261 + c = word[i];
1.262 + }
1.263 + else { /** no '<' rule used **/
1.264 + i += k - 1;
1.265 + z = 0;
1.266 + while (*s != '\0'
1.267 + && *(s+1) != '\0' && j < len) {
1.268 + if (j == 0 || target[j-1] != *s) {
1.269 + target[j] = *s;
1.270 + j++;
1.271 + }
1.272 + s++;
1.273 + }
1.274 + /** new "actual letter" **/
1.275 + c = *s;
1.276 + if (parms.rules[n][0] != '\0'
1.277 + && strstr (parms.rules[n]+1, "^^") != NULL) {
1.278 + if (c != '\0') {
1.279 + target[j] = c;
1.280 + j++;
1.281 + }
1.282 + strmove (&word[0], &word[0]+i+1);
1.283 + i = 0;
1.284 + z0 = 1;
1.285 + }
1.286 + }
1.287 + break;
1.288 + } /** end of follow-up stuff **/
1.289 + n += 2;
1.290 + } /** end of while (parms.rules[n][0] == c) **/
1.291 + } /** end of if (n >= 0) **/
1.292 + if (z0 == 0) {
1.293 +// if (k && (assert(p0!=-333),!p0) && j < len && c != '\0'
1.294 +// && (!parms.collapse_result || j == 0 || target[j-1] != c)){
1.295 + if (k && !p0 && j < len && c != '\0'
1.296 + && (1 || j == 0 || target[j-1] != c)){
1.297 + /** condense only double letters **/
1.298 + target[j] = c;
1.299 + ///printf("\n setting \n");
1.300 + j++;
1.301 + }
1.302 +
1.303 + i++;
1.304 + z = 0;
1.305 + k=0;
1.306 + }
1.307 + } /** end of while ((c = word[i]) != '\0') **/
1.308 +
1.309 + target[j] = '\0';
1.310 + return (j);
1.311 +
1.312 + } /** end of function "phonet" **/
