src/casetab.c

   1 /* GNU Emacs routines to deal with case tables.
   2    Copyright (C) 1993-1994, 2001-2011  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 #include <setjmp.h>
  23 #include "lisp.h"
  24 #include "buffer.h"
  25 #include "character.h"
  26
  27 Lisp_Object Qcase_table_p, Qcase_table;
  28 Lisp_Object Vascii_downcase_table, Vascii_upcase_table;
  29 Lisp_Object Vascii_canon_table, Vascii_eqv_table;
  30
  31 static void set_canon (Lisp_Object case_table, Lisp_Object range, Lisp_Object elt);
  32 static void set_identity (Lisp_Object table, Lisp_Object c, Lisp_Object elt);
  33 static void shuffle (Lisp_Object table, Lisp_Object c, Lisp_Object elt);
  34
  35 DEFUN ("case-table-p", Fcase_table_p, Scase_table_p, 1, 1, 0,
  36        doc: /* Return t if OBJECT is a case table.
  37 See `set-case-table' for more information on these data structures.  */)
  38   (Lisp_Object object)
  39 {
  40   Lisp_Object up, canon, eqv;
  41
  42   if (! CHAR_TABLE_P (object))
  43     return Qnil;
  44   if (! EQ (XCHAR_TABLE (object)->purpose, Qcase_table))
  45     return Qnil;
  46
  47   up = XCHAR_TABLE (object)->extras[0];
  48   canon = XCHAR_TABLE (object)->extras[1];
  49   eqv = XCHAR_TABLE (object)->extras[2];
  50
  51   return ((NILP (up) || CHAR_TABLE_P (up))
  52           && ((NILP (canon) && NILP (eqv))
  53               || (CHAR_TABLE_P (canon)
  54                   && (NILP (eqv) || CHAR_TABLE_P (eqv))))
  55           ? Qt : Qnil);
  56 }
  57
  58 static Lisp_Object
  59 check_case_table (Lisp_Object obj)
  60 {
  61   CHECK_TYPE (!NILP (Fcase_table_p (obj)), Qcase_table_p, obj);
  62   return (obj);
  63 }
  64
  65 DEFUN ("current-case-table", Fcurrent_case_table, Scurrent_case_table, 0, 0, 0,
  66        doc: /* Return the case table of the current buffer.  */)
  67   (void)
  68 {
  69   return BVAR (current_buffer, downcase_table);
  70 }
  71
  72 DEFUN ("standard-case-table", Fstandard_case_table, Sstandard_case_table, 0, 0, 0,
  73        doc: /* Return the standard case table.
  74 This is the one used for new buffers.  */)
  75   (void)
  76 {
  77   return Vascii_downcase_table;
  78 }
  79
  80 static Lisp_Object set_case_table (Lisp_Object table, int standard);
  81
  82 DEFUN ("set-case-table", Fset_case_table, Sset_case_table, 1, 1, 0,
  83        doc: /* Select a new case table for the current buffer.
  84 A case table is a char-table which maps characters
  85 to their lower-case equivalents.  It also has three \"extra\" slots
  86 which may be additional char-tables or nil.
  87 These slots are called UPCASE, CANONICALIZE and EQUIVALENCES.
  88 UPCASE maps each non-upper-case character to its upper-case equivalent.
  89  (The value in UPCASE for an upper-case character is never used.)
  90  If lower and upper case characters are in 1-1 correspondence,
  91  you may use nil and the upcase table will be deduced from DOWNCASE.
  92 CANONICALIZE maps each character to a canonical equivalent;
  93  any two characters that are related by case-conversion have the same
  94  canonical equivalent character; it may be nil, in which case it is
  95  deduced from DOWNCASE and UPCASE.
  96 EQUIVALENCES is a map that cyclicly permutes each equivalence class
  97  (of characters with the same canonical equivalent); it may be nil,
  98  in which case it is deduced from CANONICALIZE.  */)
  99   (Lisp_Object table)
 100 {
 101   return set_case_table (table, 0);
 102 }
 103
 104 DEFUN ("set-standard-case-table", Fset_standard_case_table, Sset_standard_case_table, 1, 1, 0,
 105        doc: /* Select a new standard case table for new buffers.
 106 See `set-case-table' for more info on case tables.  */)
 107   (Lisp_Object table)
 108 {
 109   return set_case_table (table, 1);
 110 }
 111
 112 static Lisp_Object
 113 set_case_table (Lisp_Object table, int standard)
 114 {
 115   Lisp_Object up, canon, eqv;
 116
 117   check_case_table (table);
 118
 119   up = XCHAR_TABLE (table)->extras[0];
 120   canon = XCHAR_TABLE (table)->extras[1];
 121   eqv = XCHAR_TABLE (table)->extras[2];
 122
 123   if (NILP (up))
 124     {
 125       up = Fmake_char_table (Qcase_table, Qnil);
 126       map_char_table (set_identity, Qnil, table, up);
 127       map_char_table (shuffle, Qnil, table, up);
 128       XCHAR_TABLE (table)->extras[0] = up;
 129     }
 130
 131   if (NILP (canon))
 132     {
 133       canon = Fmake_char_table (Qcase_table, Qnil);
 134       XCHAR_TABLE (table)->extras[1] = canon;
 135       map_char_table (set_canon, Qnil, table, table);
 136     }
 137
 138   if (NILP (eqv))
 139     {
 140       eqv = Fmake_char_table (Qcase_table, Qnil);
 141       map_char_table (set_identity, Qnil, canon, eqv);
 142       map_char_table (shuffle, Qnil, canon, eqv);
 143       XCHAR_TABLE (table)->extras[2] = eqv;
 144     }
 145
 146   /* This is so set_image_of_range_1 in regex.c can find the EQV table.  */
 147   XCHAR_TABLE (canon)->extras[2] = eqv;
 148
 149   if (standard)
 150     {
 151       Vascii_downcase_table = table;
 152       Vascii_upcase_table = up;
 153       Vascii_canon_table = canon;
 154       Vascii_eqv_table = eqv;
 155     }
 156   else
 157     {
 158       BVAR (current_buffer, downcase_table) = table;
 159       BVAR (current_buffer, upcase_table) = up;
 160       BVAR (current_buffer, case_canon_table) = canon;
 161       BVAR (current_buffer, case_eqv_table) = eqv;
 162     }
 163
 164   return table;
 165 }
 166 \f
 167 /* The following functions are called in map_char_table.  */
 168
 169 /* Set CANON char-table element for characters in RANGE to a
 170    translated ELT by UP and DOWN char-tables.  This is done only when
 171    ELT is a character.  The char-tables CANON, UP, and DOWN are in
 172    CASE_TABLE.  */
 173
 174 static void
 175 set_canon (Lisp_Object case_table, Lisp_Object range, Lisp_Object elt)
 176 {
 177   Lisp_Object up = XCHAR_TABLE (case_table)->extras[0];
 178   Lisp_Object canon = XCHAR_TABLE (case_table)->extras[1];
 179
 180   if (NATNUMP (elt))
 181     Fset_char_table_range (canon, range, Faref (case_table, Faref (up, elt)));
 182 }
 183
 184 /* Set elements of char-table TABLE for C to C itself.  C may be a
 185    cons specifying a character range.  In that case, set characters in
 186    that range to themselves.  This is done only when ELT is a
 187    character.  This is called in map_char_table.  */
 188
 189 static void
 190 set_identity (Lisp_Object table, Lisp_Object c, Lisp_Object elt)
 191 {
 192   if (NATNUMP (elt))
 193     {
 194       int from, to;
 195
 196       if (CONSP (c))
 197         {
 198           from = XINT (XCAR (c));
 199           to = XINT (XCDR (c));
 200         }
 201       else
 202         from = to = XINT (c);
 203       for (; from <= to; from++)
 204         CHAR_TABLE_SET (table, from, make_number (from));
 205     }
 206 }
 207
 208 /* Permute the elements of TABLE (which is initially an identity
 209    mapping) so that it has one cycle for each equivalence class
 210    induced by the translation table on which map_char_table is
 211    operated.  */
 212
 213 static void
 214 shuffle (Lisp_Object table, Lisp_Object c, Lisp_Object elt)
 215 {
 216   if (NATNUMP (elt))
 217     {
 218       int from, to;
 219
 220       if (CONSP (c))
 221         {
 222           from = XINT (XCAR (c));
 223           to = XINT (XCDR (c));
 224         }
 225       else
 226         from = to = XINT (c);
 227
 228       for (; from <= to; from++)
 229         {
 230           Lisp_Object tem = Faref (table, elt);
 231           Faset (table, elt, make_number (from));
 232           Faset (table, make_number (from), tem);
 233         }
 234     }
 235 }
 236 \f
 237 void
 238 init_casetab_once (void)
 239 {
 240   register int i;
 241   Lisp_Object down, up;
 242   Qcase_table = intern_c_string ("case-table");
 243   staticpro (&Qcase_table);
 244
 245   /* Intern this now in case it isn't already done.
 246      Setting this variable twice is harmless.
 247      But don't staticpro it here--that is done in alloc.c.  */
 248   Qchar_table_extra_slots = intern_c_string ("char-table-extra-slots");
 249
 250   /* Now we are ready to set up this property, so we can
 251      create char tables.  */
 252   Fput (Qcase_table, Qchar_table_extra_slots, make_number (3));
 253
 254   down = Fmake_char_table (Qcase_table, Qnil);
 255   Vascii_downcase_table = down;
 256   XCHAR_TABLE (down)->purpose = Qcase_table;
 257
 258   for (i = 0; i < 128; i++)
 259     {
 260       int c = (i >= 'A' && i <= 'Z') ? i + ('a' - 'A') : i;
 261       CHAR_TABLE_SET (down, i, make_number (c));
 262     }
 263
 264   XCHAR_TABLE (down)->extras[1] = Fcopy_sequence (down);
 265
 266   up = Fmake_char_table (Qcase_table, Qnil);
 267   XCHAR_TABLE (down)->extras[0] = up;
 268
 269   for (i = 0; i < 128; i++)
 270     {
 271       int c = ((i >= 'A' && i <= 'Z') ? i + ('a' - 'A')
 272                : ((i >= 'a' && i <= 'z') ? i + ('A' - 'a')
 273                   : i));
 274       CHAR_TABLE_SET (up, i, make_number (c));
 275     }
 276
 277   XCHAR_TABLE (down)->extras[2] = Fcopy_sequence (up);
 278
 279   /* Fill in what isn't filled in.  */
 280   set_case_table (down, 1);
 281 }
 282
 283 void
 284 syms_of_casetab (void)
 285 {
 286   Qcase_table_p = intern_c_string ("case-table-p");
 287   staticpro (&Qcase_table_p);
 288
 289   staticpro (&Vascii_canon_table);
 290   staticpro (&Vascii_downcase_table);
 291   staticpro (&Vascii_eqv_table);
 292   staticpro (&Vascii_upcase_table);
 293
 294   defsubr (&Scase_table_p);
 295   defsubr (&Scurrent_case_table);
 296   defsubr (&Sstandard_case_table);
 297   defsubr (&Sset_case_table);
 298   defsubr (&Sset_standard_case_table);
 299 }