extensions/spellcheck/hunspell/src/affixmgr.hxx

   1 /******* BEGIN LICENSE BLOCK *******
   2  * Version: MPL 1.1/GPL 2.0/LGPL 2.1
   3  *
   4  * The contents of this file are subject to the Mozilla Public License Version
   5  * 1.1 (the "License"); you may not use this file except in compliance with
   6  * the License. You may obtain a copy of the License at
   7  * http://www.mozilla.org/MPL/
   8  *
   9  * Software distributed under the License is distributed on an "AS IS" basis,
  10  * WITHOUT WARRANTY OF ANY KIND, either express or implied. See the License
  11  * for the specific language governing rights and limitations under the
  12  * License.
  13  *
  14  * The Initial Developers of the Original Code are Kevin Hendricks (MySpell)
  15  * and László Németh (Hunspell). Portions created by the Initial Developers
  16  * are Copyright (C) 2002-2005 the Initial Developers. All Rights Reserved.
  17  *
  18  * Contributor(s): Kevin Hendricks (kevin.hendricks@sympatico.ca)
  19  *                 David Einstein (deinst@world.std.com)
  20  *                 László Németh (nemethl@gyorsposta.hu)
  21  *                 Davide Prina
  22  *                 Giuseppe Modugno
  23  *                 Gianluca Turconi
  24  *                 Simon Brouwer
  25  *                 Noll Janos
  26  *                 Biro Arpad
  27  *                 Goldman Eleonora
  28  *                 Sarlos Tamas
  29  *                 Bencsath Boldizsar
  30  *                 Halacsy Peter
  31  *                 Dvornik Laszlo
  32  *                 Gefferth Andras
  33  *                 Nagy Viktor
  34  *                 Varga Daniel
  35  *                 Chris Halls
  36  *                 Rene Engelhard
  37  *                 Bram Moolenaar
  38  *                 Dafydd Jones
  39  *                 Harri Pitkanen
  40  *                 Andras Timar
  41  *                 Tor Lillqvist
  42  *
  43  * Alternatively, the contents of this file may be used under the terms of
  44  * either the GNU General Public License Version 2 or later (the "GPL"), or
  45  * the GNU Lesser General Public License Version 2.1 or later (the "LGPL"),
  46  * in which case the provisions of the GPL or the LGPL are applicable instead
  47  * of those above. If you wish to allow use of your version of this file only
  48  * under the terms of either the GPL or the LGPL, and not to allow others to
  49  * use your version of this file under the terms of the MPL, indicate your
  50  * decision by deleting the provisions above and replace them with the notice
  51  * and other provisions required by the GPL or the LGPL. If you do not delete
  52  * the provisions above, a recipient may use your version of this file under
  53  * the terms of any one of the MPL, the GPL or the LGPL.
  54  *
  55  ******* END LICENSE BLOCK *******/
  56
  57 #ifndef _AFFIXMGR_HXX_
  58 #define _AFFIXMGR_HXX_
  59
  60 #ifdef MOZILLA_CLIENT
  61 #ifdef __SUNPRO_CC // for SunONE Studio compiler
  62 using namespace std;
  63 #endif
  64 #include <stdio.h>
  65 #else
  66 #include <cstdio>
  67 #endif
  68
  69 #include "atypes.hxx"
  70 #include "baseaffix.hxx"
  71 #include "hashmgr.hxx"
  72 #include "phonet.hxx"
  73 #include "replist.hxx"
  74
  75 // check flag duplication
  76 #define dupSFX        (1 << 0)
  77 #define dupPFX        (1 << 1)
  78
  79 class AffixMgr
  80 {
  81
  82   AffEntry *          pStart[SETSIZE];
  83   AffEntry *          sStart[SETSIZE];
  84   AffEntry *          pFlag[SETSIZE];
  85   AffEntry *          sFlag[SETSIZE];
  86   HashMgr *           pHMgr;
  87   HashMgr **          alldic;
  88   int *               maxdic;
  89   char *              keystring;
  90   char *              trystring;
  91   char *              encoding;
  92   struct cs_info *    csconv;
  93   int                 utf8;
  94   int                 complexprefixes;
  95   FLAG                compoundflag;
  96   FLAG                compoundbegin;
  97   FLAG                compoundmiddle;
  98   FLAG                compoundend;
  99   FLAG                compoundroot;
 100   FLAG                compoundforbidflag;
 101   FLAG                compoundpermitflag;
 102   int                 checkcompounddup;
 103   int                 checkcompoundrep;
 104   int                 checkcompoundcase;
 105   int                 checkcompoundtriple;
 106   int                 simplifiedtriple;
 107   FLAG                forbiddenword;
 108   FLAG                nosuggest;
 109   FLAG                needaffix;
 110   int                 cpdmin;
 111   int                 numrep;
 112   replentry *         reptable;
 113   RepList *           iconvtable;
 114   RepList *           oconvtable;
 115   int                 nummap;
 116   mapentry *          maptable;
 117   int                 numbreak;
 118   char **             breaktable;
 119   int                 numcheckcpd;
 120   patentry *          checkcpdtable;
 121   int                 simplifiedcpd;
 122   int                 numdefcpd;
 123   flagentry *         defcpdtable;
 124   phonetable *        phone;
 125   int                 maxngramsugs;
 126   int                 nosplitsugs;
 127   int                 sugswithdots;
 128   int                 cpdwordmax;
 129   int                 cpdmaxsyllable;
 130   char *              cpdvowels;
 131   w_char *            cpdvowels_utf16;
 132   int                 cpdvowels_utf16_len;
 133   char *              cpdsyllablenum;
 134   const char *        pfxappnd; // BUG: not stateless
 135   const char *        sfxappnd; // BUG: not stateless
 136   FLAG                sfxflag;  // BUG: not stateless
 137   char *              derived;  // BUG: not stateless
 138   AffEntry *          sfx;      // BUG: not stateless
 139   AffEntry *          pfx;      // BUG: not stateless
 140   int                 checknum;
 141   char *              wordchars;
 142   unsigned short *    wordchars_utf16;
 143   int                 wordchars_utf16_len;
 144   char *              ignorechars;
 145   unsigned short *    ignorechars_utf16;
 146   int                 ignorechars_utf16_len;
 147   char *              version;
 148   char *              lang;
 149   int                 langnum;
 150   FLAG                lemma_present;
 151   FLAG                circumfix;
 152   FLAG                onlyincompound;
 153   FLAG                keepcase;
 154   FLAG                substandard;
 155   int                 checksharps;
 156   int                 fullstrip;
 157
 158   int                 havecontclass; // boolean variable
 159   char                contclasses[CONTSIZE]; // flags of possible continuing classes (twofold affix)
 160
 161 public:
 162
 163   AffixMgr(const char * affpath, HashMgr** ptr, int * md,
 164     const char * key = NULL);
 165   ~AffixMgr();
 166   struct hentry *     affix_check(const char * word, int len,
 167             const unsigned short needflag = (unsigned short) 0,
 168             char in_compound = IN_CPD_NOT);
 169   struct hentry *     prefix_check(const char * word, int len,
 170             char in_compound, const FLAG needflag = FLAG_NULL);
 171   inline int isSubset(const char * s1, const char * s2);
 172   struct hentry *     prefix_check_twosfx(const char * word, int len,
 173             char in_compound, const FLAG needflag = FLAG_NULL);
 174   inline int isRevSubset(const char * s1, const char * end_of_s2, int len);
 175   struct hentry *     suffix_check(const char * word, int len, int sfxopts,
 176             AffEntry* ppfx, char ** wlst, int maxSug, int * ns,
 177             const FLAG cclass = FLAG_NULL, const FLAG needflag = FLAG_NULL,
 178             char in_compound = IN_CPD_NOT);
 179   struct hentry *     suffix_check_twosfx(const char * word, int len,
 180             int sfxopts, AffEntry* ppfx, const FLAG needflag = FLAG_NULL);
 181
 182   char * affix_check_morph(const char * word, int len,
 183             const FLAG needflag = FLAG_NULL, char in_compound = IN_CPD_NOT);
 184   char * prefix_check_morph(const char * word, int len,
 185             char in_compound, const FLAG needflag = FLAG_NULL);
 186   char * suffix_check_morph (const char * word, int len, int sfxopts,
 187             AffEntry * ppfx, const FLAG cclass = FLAG_NULL,
 188             const FLAG needflag = FLAG_NULL, char in_compound = IN_CPD_NOT);
 189
 190   char * prefix_check_twosfx_morph(const char * word, int len,
 191             char in_compound, const FLAG needflag = FLAG_NULL);
 192   char * suffix_check_twosfx_morph(const char * word, int len,
 193             int sfxopts, AffEntry * ppfx, const FLAG needflag = FLAG_NULL);
 194
 195   char * morphgen(char * ts, int wl, const unsigned short * ap,
 196             unsigned short al, char * morph, char * targetmorph, int level);
 197
 198   int    expand_rootword(struct guessword * wlst, int maxn, const char * ts,
 199             int wl, const unsigned short * ap, unsigned short al, char * bad,
 200             int, char *);
 201
 202   short       get_syllable (const char * word, int wlen);
 203   int         cpdrep_check(const char * word, int len);
 204   int         cpdpat_check(const char * word, int len, hentry * r1, hentry * r2);
 205   int         defcpd_check(hentry *** words, short wnum, hentry * rv,
 206                     hentry ** rwords, char all);
 207   int         cpdcase_check(const char * word, int len);
 208   inline int  candidate_check(const char * word, int len);
 209   void        setcminmax(int * cmin, int * cmax, const char * word, int len);
 210   struct hentry * compound_check(const char * word, int len, short wordnum,
 211             short numsyllable, short maxwordnum, short wnum, hentry ** words,
 212             char hu_mov_rule, char is_sug);
 213
 214   int compound_check_morph(const char * word, int len, short wordnum,
 215             short numsyllable, short maxwordnum, short wnum, hentry ** words,
 216             char hu_mov_rule, char ** result, char * partresult);
 217
 218   struct hentry * lookup(const char * word);
 219   int                 get_numrep();
 220   struct replentry *  get_reptable();
 221   RepList *           get_iconvtable();
 222   RepList *           get_oconvtable();
 223   struct phonetable * get_phonetable();
 224   int                 get_nummap();
 225   struct mapentry *   get_maptable();
 226   int                 get_numbreak();
 227   char **             get_breaktable();
 228   char *              get_encoding();
 229   int                 get_langnum();
 230   char *              get_key_string();
 231   char *              get_try_string();
 232   const char *        get_wordchars();
 233   unsigned short *    get_wordchars_utf16(int * len);
 234   char *              get_ignore();
 235   unsigned short *    get_ignore_utf16(int * len);
 236   int                 get_compound();
 237   FLAG                get_compoundflag();
 238   FLAG                get_compoundbegin();
 239   FLAG                get_forbiddenword();
 240   FLAG                get_nosuggest();
 241   FLAG                get_needaffix();
 242   FLAG                get_onlyincompound();
 243   FLAG                get_compoundroot();
 244   FLAG                get_lemma_present();
 245   int                 get_checknum();
 246   char *              get_possible_root();
 247   const char *        get_prefix();
 248   const char *        get_suffix();
 249   const char *        get_derived();
 250   const char *        get_version();
 251   const int           have_contclass();
 252   int                 get_utf8();
 253   int                 get_complexprefixes();
 254   char *              get_suffixed(char );
 255   int                 get_maxngramsugs();
 256   int                 get_nosplitsugs();
 257   int                 get_sugswithdots(void);
 258   FLAG                get_keepcase(void);
 259   int                 get_checksharps(void);
 260   char *              encode_flag(unsigned short aflag);
 261   int                 get_fullstrip();
 262
 263 private:
 264   int  parse_file(const char * affpath, const char * key);
 265   int  parse_flag(char * line, unsigned short * out, FileMgr * af);
 266   int  parse_num(char * line, int * out, FileMgr * af);
 267   int  parse_cpdsyllable(char * line, FileMgr * af);
 268   int  parse_reptable(char * line, FileMgr * af);
 269   int  parse_convtable(char * line, FileMgr * af, RepList ** rl, const char * keyword);
 270   int  parse_phonetable(char * line, FileMgr * af);
 271   int  parse_maptable(char * line, FileMgr * af);
 272   int  parse_breaktable(char * line, FileMgr * af);
 273   int  parse_checkcpdtable(char * line, FileMgr * af);
 274   int  parse_defcpdtable(char * line, FileMgr * af);
 275   int  parse_affix(char * line, const char at, FileMgr * af, char * dupflags);
 276
 277   void reverse_condition(char *);
 278   void debugflag(char * result, unsigned short flag);
 279   int condlen(char *);
 280   int encodeit(struct affentry * ptr, char * cs);
 281   int build_pfxtree(AffEntry* pfxptr);
 282   int build_sfxtree(AffEntry* sfxptr);
 283   int process_pfx_order();
 284   int process_sfx_order();
 285   AffEntry * process_pfx_in_order(AffEntry * ptr, AffEntry * nptr);
 286   AffEntry * process_sfx_in_order(AffEntry * ptr, AffEntry * nptr);
 287   int process_pfx_tree_to_list();
 288   int process_sfx_tree_to_list();
 289   int redundant_condition(char, char * strip, int stripl,
 290       const char * cond, int);
 291 };
 292
 293 #endif