1.1 --- /dev/null Thu Jan 01 00:00:00 1970 +0000
1.2 +++ b/extensions/spellcheck/hunspell/src/hunspell.hxx Wed Dec 31 06:09:35 2014 +0100
1.3 @@ -0,0 +1,229 @@
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 Developers of the Original Code are Kevin Hendricks (MySpell)
1.18 + * and László Németh (Hunspell). Portions created by the Initial Developers
1.19 + * are Copyright (C) 2002-2005 the Initial Developers. All Rights Reserved.
1.20 + *
1.21 + * Contributor(s): Kevin Hendricks (kevin.hendricks@sympatico.ca)
1.22 + * David Einstein (deinst@world.std.com)
1.23 + * László Németh (nemethl@gyorsposta.hu)
1.24 + * Caolan McNamara (caolanm@redhat.com)
1.25 + * Davide Prina
1.26 + * Giuseppe Modugno
1.27 + * Gianluca Turconi
1.28 + * Simon Brouwer
1.29 + * Noll Janos
1.30 + * Biro Arpad
1.31 + * Goldman Eleonora
1.32 + * Sarlos Tamas
1.33 + * Bencsath Boldizsar
1.34 + * Halacsy Peter
1.35 + * Dvornik Laszlo
1.36 + * Gefferth Andras
1.37 + * Nagy Viktor
1.38 + * Varga Daniel
1.39 + * Chris Halls
1.40 + * Rene Engelhard
1.41 + * Bram Moolenaar
1.42 + * Dafydd Jones
1.43 + * Harri Pitkanen
1.44 + * Andras Timar
1.45 + * Tor Lillqvist
1.46 + *
1.47 + * Alternatively, the contents of this file may be used under the terms of
1.48 + * either the GNU General Public License Version 2 or later (the "GPL"), or
1.49 + * the GNU Lesser General Public License Version 2.1 or later (the "LGPL"),
1.50 + * in which case the provisions of the GPL or the LGPL are applicable instead
1.51 + * of those above. If you wish to allow use of your version of this file only
1.52 + * under the terms of either the GPL or the LGPL, and not to allow others to
1.53 + * use your version of this file under the terms of the MPL, indicate your
1.54 + * decision by deleting the provisions above and replace them with the notice
1.55 + * and other provisions required by the GPL or the LGPL. If you do not delete
1.56 + * the provisions above, a recipient may use your version of this file under
1.57 + * the terms of any one of the MPL, the GPL or the LGPL.
1.58 + *
1.59 + ******* END LICENSE BLOCK *******/
1.60 +
1.61 +#include "hunvisapi.h"
1.62 +
1.63 +#include "hashmgr.hxx"
1.64 +#include "affixmgr.hxx"
1.65 +#include "suggestmgr.hxx"
1.66 +#include "langnum.hxx"
1.67 +
1.68 +#define SPELL_XML "<?xml?>"
1.69 +
1.70 +#define MAXDIC 20
1.71 +#define MAXSUGGESTION 15
1.72 +#define MAXSHARPS 5
1.73 +
1.74 +#define HUNSPELL_OK (1 << 0)
1.75 +#define HUNSPELL_OK_WARN (1 << 1)
1.76 +
1.77 +#ifndef _MYSPELLMGR_HXX_
1.78 +#define _MYSPELLMGR_HXX_
1.79 +
1.80 +class LIBHUNSPELL_DLL_EXPORTED Hunspell
1.81 +{
1.82 + AffixMgr* pAMgr;
1.83 + HashMgr* pHMgr[MAXDIC];
1.84 + int maxdic;
1.85 + SuggestMgr* pSMgr;
1.86 + char * affixpath;
1.87 + char * encoding;
1.88 + struct cs_info * csconv;
1.89 + int langnum;
1.90 + int utf8;
1.91 + int complexprefixes;
1.92 + char** wordbreak;
1.93 +
1.94 +public:
1.95 +
1.96 + /* Hunspell(aff, dic) - constructor of Hunspell class
1.97 + * input: path of affix file and dictionary file
1.98 + */
1.99 +
1.100 + Hunspell(const char * affpath, const char * dpath, const char * key = NULL);
1.101 + ~Hunspell();
1.102 +
1.103 + /* load extra dictionaries (only dic files) */
1.104 + int add_dic(const char * dpath, const char * key = NULL);
1.105 +
1.106 + /* spell(word) - spellcheck word
1.107 + * output: 0 = bad word, not 0 = good word
1.108 + *
1.109 + * plus output:
1.110 + * info: information bit array, fields:
1.111 + * SPELL_COMPOUND = a compound word
1.112 + * SPELL_FORBIDDEN = an explicit forbidden word
1.113 + * root: root (stem), when input is a word with affix(es)
1.114 + */
1.115 +
1.116 + int spell(const char * word, int * info = NULL, char ** root = NULL);
1.117 +
1.118 + /* suggest(suggestions, word) - search suggestions
1.119 + * input: pointer to an array of strings pointer and the (bad) word
1.120 + * array of strings pointer (here *slst) may not be initialized
1.121 + * output: number of suggestions in string array, and suggestions in
1.122 + * a newly allocated array of strings (*slts will be NULL when number
1.123 + * of suggestion equals 0.)
1.124 + */
1.125 +
1.126 + int suggest(char*** slst, const char * word);
1.127 +
1.128 + /* deallocate suggestion lists */
1.129 +
1.130 + void free_list(char *** slst, int n);
1.131 +
1.132 + char * get_dic_encoding();
1.133 +
1.134 + /* morphological functions */
1.135 +
1.136 + /* analyze(result, word) - morphological analysis of the word */
1.137 +
1.138 + int analyze(char*** slst, const char * word);
1.139 +
1.140 + /* stem(result, word) - stemmer function */
1.141 +
1.142 + int stem(char*** slst, const char * word);
1.143 +
1.144 + /* stem(result, analysis, n) - get stems from a morph. analysis
1.145 + * example:
1.146 + * char ** result, result2;
1.147 + * int n1 = analyze(&result, "words");
1.148 + * int n2 = stem(&result2, result, n1);
1.149 + */
1.150 +
1.151 + int stem(char*** slst, char ** morph, int n);
1.152 +
1.153 + /* generate(result, word, word2) - morphological generation by example(s) */
1.154 +
1.155 + int generate(char*** slst, const char * word, const char * word2);
1.156 +
1.157 + /* generate(result, word, desc, n) - generation by morph. description(s)
1.158 + * example:
1.159 + * char ** result;
1.160 + * char * affix = "is:plural"; // description depends from dictionaries, too
1.161 + * int n = generate(&result, "word", &affix, 1);
1.162 + * for (int i = 0; i < n; i++) printf("%s\n", result[i]);
1.163 + */
1.164 +
1.165 + int generate(char*** slst, const char * word, char ** desc, int n);
1.166 +
1.167 + /* functions for run-time modification of the dictionary */
1.168 +
1.169 + /* add word to the run-time dictionary */
1.170 +
1.171 + int add(const char * word);
1.172 +
1.173 + /* add word to the run-time dictionary with affix flags of
1.174 + * the example (a dictionary word): Hunspell will recognize
1.175 + * affixed forms of the new word, too.
1.176 + */
1.177 +
1.178 + int add_with_affix(const char * word, const char * example);
1.179 +
1.180 + /* remove word from the run-time dictionary */
1.181 +
1.182 + int remove(const char * word);
1.183 +
1.184 + /* other */
1.185 +
1.186 + /* get extra word characters definied in affix file for tokenization */
1.187 + const char * get_wordchars();
1.188 + unsigned short * get_wordchars_utf16(int * len);
1.189 +
1.190 + struct cs_info * get_csconv();
1.191 + const char * get_version();
1.192 +
1.193 + int get_langnum() const;
1.194 +
1.195 + /* experimental and deprecated functions */
1.196 +
1.197 +#ifdef HUNSPELL_EXPERIMENTAL
1.198 + /* suffix is an affix flag string, similarly in dictionary files */
1.199 + int put_word_suffix(const char * word, const char * suffix);
1.200 + char * morph_with_correction(const char * word);
1.201 +
1.202 + /* spec. suggestions */
1.203 + int suggest_auto(char*** slst, const char * word);
1.204 + int suggest_pos_stems(char*** slst, const char * word);
1.205 +#endif
1.206 +
1.207 +private:
1.208 + int cleanword(char *, const char *, int * pcaptype, int * pabbrev);
1.209 + int cleanword2(char *, const char *, w_char *, int * w_len, int * pcaptype, int * pabbrev);
1.210 + void mkinitcap(char *);
1.211 + int mkinitcap2(char * p, w_char * u, int nc);
1.212 + int mkinitsmall2(char * p, w_char * u, int nc);
1.213 + void mkallcap(char *);
1.214 + int mkallcap2(char * p, w_char * u, int nc);
1.215 + void mkallsmall(char *);
1.216 + int mkallsmall2(char * p, w_char * u, int nc);
1.217 + struct hentry * checkword(const char *, int * info, char **root);
1.218 + char * sharps_u8_l1(char * dest, char * source);
1.219 + hentry * spellsharps(char * base, char *, int, int, char * tmp, int * info, char **root);
1.220 + int is_keepcase(const hentry * rv);
1.221 + int insert_sug(char ***slst, char * word, int ns);
1.222 + void cat_result(char * result, char * st);
1.223 + char * stem_description(const char * desc);
1.224 + int spellml(char*** slst, const char * word);
1.225 + int get_xml_par(char * dest, const char * par, int maxl);
1.226 + const char * get_xml_pos(const char * s, const char * attr);
1.227 + int get_xml_list(char ***slst, char * list, const char * tag);
1.228 + int check_xml_par(const char * q, const char * attr, const char * value);
1.229 +
1.230 +};
1.231 +
1.232 +#endif
