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 static 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 DEFUE ("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 DEFUE ("set-standard-case-table", Fset_standard_case_table,
 105        Sset_standard_case_table, 1, 1, 0,
 106        doc: /* Select a new standard case table for new buffers.
 107 See `set-case-table' for more info on case tables.  */)
 108   (Lisp_Object table)
 109 {
 110   return set_case_table (table, 1);
 111 }
 112
 113 static Lisp_Object
 114 set_case_table (Lisp_Object table, int standard)
 115 {
 116   Lisp_Object up, canon, eqv;
 117
 118   check_case_table (table);
 119
 120   up = XCHAR_TABLE (table)->extras[0];
 121   canon = XCHAR_TABLE (table)->extras[1];
 122   eqv = XCHAR_TABLE (table)->extras[2];
 123
 124   if (NILP (up))
 125     {
 126       up = Fmake_char_table (Qcase_table, Qnil);
 127       map_char_table (set_identity, Qnil, table, up);
 128       map_char_table (shuffle, Qnil, table, up);
 129       XCHAR_TABLE (table)->extras[0] = up;
 130     }
 131
 132   if (NILP (canon))
 133     {
 134       canon = Fmake_char_table (Qcase_table, Qnil);
 135       XCHAR_TABLE (table)->extras[1] = canon;
 136       map_char_table (set_canon, Qnil, table, table);
 137     }
 138
 139   if (NILP (eqv))
 140     {
 141       eqv = Fmake_char_table (Qcase_table, Qnil);
 142       map_char_table (set_identity, Qnil, canon, eqv);
 143       map_char_table (shuffle, Qnil, canon, eqv);
 144       XCHAR_TABLE (table)->extras[2] = eqv;
 145     }
 146
 147   /* This is so set_image_of_range_1 in regex.c can find the EQV table.  */
 148   XCHAR_TABLE (canon)->extras[2] = eqv;
 149
 150   if (standard)
 151     {
 152       Vascii_downcase_table = table;
 153       Vascii_upcase_table = up;
 154       Vascii_canon_table = canon;
 155       Vascii_eqv_table = eqv;
 156     }
 157   else
 158     {
 159       BVAR (current_buffer, downcase_table) = table;
 160       BVAR (current_buffer, upcase_table) = up;
 161       BVAR (current_buffer, case_canon_table) = canon;
 162       BVAR (current_buffer, case_eqv_table) = eqv;
 163     }
 164
 165   return table;
 166 }
 167 \f
 168 /* The following functions are called in map_char_table.  */
 169
 170 /* Set CANON char-table element for characters in RANGE to a
 171    translated ELT by UP and DOWN char-tables.  This is done only when
 172    ELT is a character.  The char-tables CANON, UP, and DOWN are in
 173    CASE_TABLE.  */
 174
 175 static void
 176 set_canon (Lisp_Object case_table, Lisp_Object range, Lisp_Object elt)
 177 {
 178   Lisp_Object up = XCHAR_TABLE (case_table)->extras[0];
 179   Lisp_Object canon = XCHAR_TABLE (case_table)->extras[1];
 180
 181   if (NATNUMP (elt))
 182     Fset_char_table_range (canon, range, Faref (case_table, Faref (up, elt)));
 183 }
 184
 185 /* Set elements of char-table TABLE for C to C itself.  C may be a
 186    cons specifying a character range.  In that case, set characters in
 187    that range to themselves.  This is done only when ELT is a
 188    character.  This is called in map_char_table.  */
 189
 190 static void
 191 set_identity (Lisp_Object table, Lisp_Object c, Lisp_Object elt)
 192 {
 193   if (NATNUMP (elt))
 194     {
 195       int from;
 196       unsigned to;
 197
 198       if (CONSP (c))
 199         {
 200           from = XINT (XCAR (c));
 201           to = XINT (XCDR (c));
 202         }
 203       else
 204         from = to = XINT (c);
 205       for (to++; from < to; from++)
 206         CHAR_TABLE_SET (table, from, make_number (from));
 207     }
 208 }
 209
 210 /* Permute the elements of TABLE (which is initially an identity
 211    mapping) so that it has one cycle for each equivalence class
 212    induced by the translation table on which map_char_table is
 213    operated.  */
 214
 215 static void
 216 shuffle (Lisp_Object table, Lisp_Object c, Lisp_Object elt)
 217 {
 218   if (NATNUMP (elt))
 219     {
 220       int from;
 221       unsigned to;
 222
 223       if (CONSP (c))
 224         {
 225           from = XINT (XCAR (c));
 226           to = XINT (XCDR (c));
 227         }
 228       else
 229         from = to = XINT (c);
 230
 231       for (to++; from < to; from++)
 232         {
 233           Lisp_Object tem = Faref (table, elt);
 234           Faset (table, elt, make_number (from));
 235           Faset (table, make_number (from), tem);
 236         }
 237     }
 238 }
 239 \f
 240 void
 241 init_casetab_once (void)
 242 {
 243   register int i;
 244   Lisp_Object down, up;
 245   Qcase_table = intern_c_string ("case-table");
 246   staticpro (&Qcase_table);
 247
 248   /* Intern this now in case it isn't already done.
 249      Setting this variable twice is harmless.
 250      But don't staticpro it here--that is done in alloc.c.  */
 251   Qchar_table_extra_slots = intern_c_string ("char-table-extra-slots");
 252
 253   /* Now we are ready to set up this property, so we can
 254      create char tables.  */
 255   Fput (Qcase_table, Qchar_table_extra_slots, make_number (3));
 256
 257   down = Fmake_char_table (Qcase_table, Qnil);
 258   Vascii_downcase_table = down;
 259   XCHAR_TABLE (down)->purpose = Qcase_table;
 260
 261   for (i = 0; i < 128; i++)
 262     {
 263       int c = (i >= 'A' && i <= 'Z') ? i + ('a' - 'A') : i;
 264       CHAR_TABLE_SET (down, i, make_number (c));
 265     }
 266
 267   XCHAR_TABLE (down)->extras[1] = Fcopy_sequence (down);
 268
 269   up = Fmake_char_table (Qcase_table, Qnil);
 270   XCHAR_TABLE (down)->extras[0] = up;
 271
 272   for (i = 0; i < 128; i++)
 273     {
 274       int c = ((i >= 'A' && i <= 'Z') ? i + ('a' - 'A')
 275                : ((i >= 'a' && i <= 'z') ? i + ('A' - 'a')
 276                   : i));
 277       CHAR_TABLE_SET (up, i, make_number (c));
 278     }
 279
 280   XCHAR_TABLE (down)->extras[2] = Fcopy_sequence (up);
 281
 282   /* Fill in what isn't filled in.  */
 283   set_case_table (down, 1);
 284 }
 285
 286 void
 287 syms_of_casetab (void)
 288 {
 289   Qcase_table_p = intern_c_string ("case-table-p");
 290   staticpro (&Qcase_table_p);
 291
 292   staticpro (&Vascii_canon_table);
 293   staticpro (&Vascii_downcase_table);
 294   staticpro (&Vascii_eqv_table);
 295   staticpro (&Vascii_upcase_table);
 296
 297   defsubr (&Scase_table_p);
 298   defsubr (&Scurrent_case_table);
 299   defsubr (&Sstandard_case_table);
 300   defsubr (&Sset_case_table);
 301   defsubr (&Sset_standard_case_table);
 302 }