src/casetab.c

   1 /* GNU Emacs routines to deal with case tables.
   2    Copyright (C) 1993-1994, 2001-2012  Free Software Foundation, Inc.
   3
   4 Author: Howard Gayle
   5
   6 This file is part of GNU Emacs.
   7
   8 GNU Emacs is free software: you can redistribute it and/or modify
   9 it under the terms of the GNU General Public License as published by
  10 the Free Software Foundation, either version 3 of the License, or
  11 (at your option) any later version.
  12
  13 GNU Emacs is distributed in the hope that it will be useful,
  14 but WITHOUT ANY WARRANTY; without even the implied warranty of
  15 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  16 GNU General Public License for more details.
  17
  18 You should have received a copy of the GNU General Public License
  19 along with GNU Emacs.  If not, see <http://www.gnu.org/licenses/>.  */
  20
  21 #include <config.h>
  22
  23 #include "lisp.h"
  24 #include "character.h"
  25 #include "buffer.h"
  26
  27 static Lisp_Object Qcase_table_p, Qcase_table;
  28 Lisp_Object Vascii_downcase_table;
  29 static Lisp_Object Vascii_upcase_table;
  30 Lisp_Object Vascii_canon_table;
  31 static Lisp_Object Vascii_eqv_table;
  32
  33 static void set_canon (Lisp_Object case_table, Lisp_Object range, Lisp_Object elt);
  34 static void set_identity (Lisp_Object table, Lisp_Object c, Lisp_Object elt);
  35 static void shuffle (Lisp_Object table, Lisp_Object c, Lisp_Object elt);
  36
  37 DEFUN ("case-table-p", Fcase_table_p, Scase_table_p, 1, 1, 0,
  38        doc: /* Return t if OBJECT is a case table.
  39 See `set-case-table' for more information on these data structures.  */)
  40   (Lisp_Object object)
  41 {
  42   Lisp_Object up, canon, eqv;
  43
  44   if (! CHAR_TABLE_P (object))
  45     return Qnil;
  46   if (! EQ (XCHAR_TABLE (object)->purpose, Qcase_table))
  47     return Qnil;
  48
  49   up = XCHAR_TABLE (object)->extras[0];
  50   canon = XCHAR_TABLE (object)->extras[1];
  51   eqv = XCHAR_TABLE (object)->extras[2];
  52
  53   return ((NILP (up) || CHAR_TABLE_P (up))
  54           && ((NILP (canon) && NILP (eqv))
  55               || (CHAR_TABLE_P (canon)
  56                   && (NILP (eqv) || CHAR_TABLE_P (eqv))))
  57           ? Qt : Qnil);
  58 }
  59
  60 static Lisp_Object
  61 check_case_table (Lisp_Object obj)
  62 {
  63   CHECK_TYPE (!NILP (Fcase_table_p (obj)), Qcase_table_p, obj);
  64   return (obj);
  65 }
  66
  67 DEFUN ("current-case-table", Fcurrent_case_table, Scurrent_case_table, 0, 0, 0,
  68        doc: /* Return the case table of the current buffer.  */)
  69   (void)
  70 {
  71   return BVAR (current_buffer, downcase_table);
  72 }
  73
  74 DEFUN ("standard-case-table", Fstandard_case_table, Sstandard_case_table, 0, 0, 0,
  75        doc: /* Return the standard case table.
  76 This is the one used for new buffers.  */)
  77   (void)
  78 {
  79   return Vascii_downcase_table;
  80 }
  81
  82 static Lisp_Object set_case_table (Lisp_Object, bool);
  83
  84 DEFUN ("set-case-table", Fset_case_table, Sset_case_table, 1, 1, 0,
  85        doc: /* Select a new case table for the current buffer.
  86 A case table is a char-table which maps characters
  87 to their lower-case equivalents.  It also has three \"extra\" slots
  88 which may be additional char-tables or nil.
  89 These slots are called UPCASE, CANONICALIZE and EQUIVALENCES.
  90 UPCASE maps each non-upper-case character to its upper-case equivalent.
  91  (The value in UPCASE for an upper-case character is never used.)
  92  If lower and upper case characters are in 1-1 correspondence,
  93  you may use nil and the upcase table will be deduced from DOWNCASE.
  94 CANONICALIZE maps each character to a canonical equivalent;
  95  any two characters that are related by case-conversion have the same
  96  canonical equivalent character; it may be nil, in which case it is
  97  deduced from DOWNCASE and UPCASE.
  98 EQUIVALENCES is a map that cyclically permutes each equivalence class
  99  (of characters with the same canonical equivalent); it may be nil,
 100  in which case it is deduced from CANONICALIZE.  */)
 101   (Lisp_Object table)
 102 {
 103   return set_case_table (table, 0);
 104 }
 105
 106 DEFUN ("set-standard-case-table", Fset_standard_case_table,
 107        Sset_standard_case_table, 1, 1, 0,
 108        doc: /* Select a new standard case table for new buffers.
 109 See `set-case-table' for more info on case tables.  */)
 110   (Lisp_Object table)
 111 {
 112   return set_case_table (table, 1);
 113 }
 114
 115 static Lisp_Object
 116 set_case_table (Lisp_Object table, bool standard)
 117 {
 118   Lisp_Object up, canon, eqv;
 119
 120   check_case_table (table);
 121
 122   up = XCHAR_TABLE (table)->extras[0];
 123   canon = XCHAR_TABLE (table)->extras[1];
 124   eqv = XCHAR_TABLE (table)->extras[2];
 125
 126   if (NILP (up))
 127     {
 128       up = Fmake_char_table (Qcase_table, Qnil);
 129       map_char_table (set_identity, Qnil, table, up);
 130       map_char_table (shuffle, Qnil, table, up);
 131       set_char_table_extras (table, 0, up);
 132     }
 133
 134   if (NILP (canon))
 135     {
 136       canon = Fmake_char_table (Qcase_table, Qnil);
 137       set_char_table_extras (table, 1, canon);
 138       map_char_table (set_canon, Qnil, table, table);
 139     }
 140
 141   if (NILP (eqv))
 142     {
 143       eqv = Fmake_char_table (Qcase_table, Qnil);
 144       map_char_table (set_identity, Qnil, canon, eqv);
 145       map_char_table (shuffle, Qnil, canon, eqv);
 146       set_char_table_extras (table, 2, eqv);
 147     }
 148
 149   /* This is so set_image_of_range_1 in regex.c can find the EQV table.  */
 150   set_char_table_extras (canon, 2, eqv);
 151
 152   if (standard)
 153     {
 154       Vascii_downcase_table = table;
 155       Vascii_upcase_table = up;
 156       Vascii_canon_table = canon;
 157       Vascii_eqv_table = eqv;
 158     }
 159   else
 160     {
 161       bset_downcase_table (current_buffer, table);
 162       bset_upcase_table (current_buffer, up);
 163       bset_case_canon_table (current_buffer, canon);
 164       bset_case_eqv_table (current_buffer, eqv);
 165     }
 166
 167   return table;
 168 }
 169 \f
 170 /* The following functions are called in map_char_table.  */
 171
 172 /* Set CANON char-table element for characters in RANGE to a
 173    translated ELT by UP and DOWN char-tables.  This is done only when
 174    ELT is a character.  The char-tables CANON, UP, and DOWN are in
 175    CASE_TABLE.  */
 176
 177 static void
 178 set_canon (Lisp_Object case_table, Lisp_Object range, Lisp_Object elt)
 179 {
 180   Lisp_Object up = XCHAR_TABLE (case_table)->extras[0];
 181   Lisp_Object canon = XCHAR_TABLE (case_table)->extras[1];
 182
 183   if (NATNUMP (elt))
 184     Fset_char_table_range (canon, range, Faref (case_table, Faref (up, elt)));
 185 }
 186
 187 /* Set elements of char-table TABLE for C to C itself.  C may be a
 188    cons specifying a character range.  In that case, set characters in
 189    that range to themselves.  This is done only when ELT is a
 190    character.  This is called in map_char_table.  */
 191
 192 static void
 193 set_identity (Lisp_Object table, Lisp_Object c, Lisp_Object elt)
 194 {
 195   if (NATNUMP (elt))
 196     {
 197       int from, to;
 198
 199       if (CONSP (c))
 200         {
 201           from = XINT (XCAR (c));
 202           to = XINT (XCDR (c));
 203         }
 204       else
 205         from = to = XINT (c);
 206
 207       to++;
 208       lint_assume (to <= MAX_CHAR + 1);
 209       for (; from < to; from++)
 210         CHAR_TABLE_SET (table, from, make_number (from));
 211     }
 212 }
 213
 214 /* Permute the elements of TABLE (which is initially an identity
 215    mapping) so that it has one cycle for each equivalence class
 216    induced by the translation table on which map_char_table is
 217    operated.  */
 218
 219 static void
 220 shuffle (Lisp_Object table, Lisp_Object c, Lisp_Object elt)
 221 {
 222   if (NATNUMP (elt))
 223     {
 224       int from, to;
 225
 226       if (CONSP (c))
 227         {
 228           from = XINT (XCAR (c));
 229           to = XINT (XCDR (c));
 230         }
 231       else
 232         from = to = XINT (c);
 233
 234       to++;
 235       lint_assume (to <= MAX_CHAR + 1);
 236       for (; from < to; from++)
 237         {
 238           Lisp_Object tem = Faref (table, elt);
 239           Faset (table, elt, make_number (from));
 240           Faset (table, make_number (from), tem);
 241         }
 242     }
 243 }
 244 \f
 245 void
 246 init_casetab_once (void)
 247 {
 248   register int i;
 249   Lisp_Object down, up;
 250   DEFSYM (Qcase_table, "case-table");
 251
 252   /* Intern this now in case it isn't already done.
 253      Setting this variable twice is harmless.
 254      But don't staticpro it here--that is done in alloc.c.  */
 255   Qchar_table_extra_slots = intern_c_string ("char-table-extra-slots");
 256
 257   /* Now we are ready to set up this property, so we can
 258      create char tables.  */
 259   Fput (Qcase_table, Qchar_table_extra_slots, make_number (3));
 260
 261   down = Fmake_char_table (Qcase_table, Qnil);
 262   Vascii_downcase_table = down;
 263   set_char_table_purpose (down, Qcase_table);
 264
 265   for (i = 0; i < 128; i++)
 266     {
 267       int c = (i >= 'A' && i <= 'Z') ? i + ('a' - 'A') : i;
 268       CHAR_TABLE_SET (down, i, make_number (c));
 269     }
 270
 271   set_char_table_extras (down, 1, Fcopy_sequence (down));
 272
 273   up = Fmake_char_table (Qcase_table, Qnil);
 274   set_char_table_extras (down, 0, up);
 275
 276   for (i = 0; i < 128; i++)
 277     {
 278       int c = ((i >= 'A' && i <= 'Z') ? i + ('a' - 'A')
 279                : ((i >= 'a' && i <= 'z') ? i + ('A' - 'a')
 280                   : i));
 281       CHAR_TABLE_SET (up, i, make_number (c));
 282     }
 283
 284   set_char_table_extras (down, 2, Fcopy_sequence (up));
 285
 286   /* Fill in what isn't filled in.  */
 287   set_case_table (down, 1);
 288 }
 289
 290 void
 291 syms_of_casetab (void)
 292 {
 293   DEFSYM (Qcase_table_p, "case-table-p");
 294
 295   staticpro (&Vascii_canon_table);
 296   staticpro (&Vascii_downcase_table);
 297   staticpro (&Vascii_eqv_table);
 298   staticpro (&Vascii_upcase_table);
 299
 300   defsubr (&Scase_table_p);
 301   defsubr (&Scurrent_case_table);
 302   defsubr (&Sstandard_case_table);
 303   defsubr (&Sset_case_table);
 304   defsubr (&Sset_standard_case_table);
 305 }