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 /* Used as a temporary in DOWNCASE and other macros in lisp.h.  No
  32    need to mark it, since it is used only very temporarily.  */
  33 int case_temp1;
  34 Lisp_Object case_temp2;
  35
  36 static void set_canon (Lisp_Object case_table, Lisp_Object range, Lisp_Object elt);
  37 static void set_identity (Lisp_Object table, Lisp_Object c, Lisp_Object elt);
  38 static void shuffle (Lisp_Object table, Lisp_Object c, Lisp_Object elt);
  39
  40 DEFUN ("case-table-p", Fcase_table_p, Scase_table_p, 1, 1, 0,
  41        doc: /* Return t if OBJECT is a case table.
  42 See `set-case-table' for more information on these data structures.  */)
  43   (Lisp_Object object)
  44 {
  45   Lisp_Object up, canon, eqv;
  46
  47   if (! CHAR_TABLE_P (object))
  48     return Qnil;
  49   if (! EQ (XCHAR_TABLE (object)->purpose, Qcase_table))
  50     return Qnil;
  51
  52   up = XCHAR_TABLE (object)->extras[0];
  53   canon = XCHAR_TABLE (object)->extras[1];
  54   eqv = XCHAR_TABLE (object)->extras[2];
  55
  56   return ((NILP (up) || CHAR_TABLE_P (up))
  57           && ((NILP (canon) && NILP (eqv))
  58               || (CHAR_TABLE_P (canon)
  59                   && (NILP (eqv) || CHAR_TABLE_P (eqv))))
  60           ? Qt : Qnil);
  61 }
  62
  63 static Lisp_Object
  64 check_case_table (Lisp_Object obj)
  65 {
  66   CHECK_TYPE (!NILP (Fcase_table_p (obj)), Qcase_table_p, obj);
  67   return (obj);
  68 }
  69
  70 DEFUN ("current-case-table", Fcurrent_case_table, Scurrent_case_table, 0, 0, 0,
  71        doc: /* Return the case table of the current buffer.  */)
  72   (void)
  73 {
  74   return BVAR (current_buffer, downcase_table);
  75 }
  76
  77 DEFUN ("standard-case-table", Fstandard_case_table, Sstandard_case_table, 0, 0, 0,
  78        doc: /* Return the standard case table.
  79 This is the one used for new buffers.  */)
  80   (void)
  81 {
  82   return Vascii_downcase_table;
  83 }
  84
  85 static Lisp_Object set_case_table (Lisp_Object table, int standard);
  86
  87 DEFUN ("set-case-table", Fset_case_table, Sset_case_table, 1, 1, 0,
  88        doc: /* Select a new case table for the current buffer.
  89 A case table is a char-table which maps characters
  90 to their lower-case equivalents.  It also has three \"extra\" slots
  91 which may be additional char-tables or nil.
  92 These slots are called UPCASE, CANONICALIZE and EQUIVALENCES.
  93 UPCASE maps each non-upper-case character to its upper-case equivalent.
  94  (The value in UPCASE for an upper-case character is never used.)
  95  If lower and upper case characters are in 1-1 correspondence,
  96  you may use nil and the upcase table will be deduced from DOWNCASE.
  97 CANONICALIZE maps each character to a canonical equivalent;
  98  any two characters that are related by case-conversion have the same
  99  canonical equivalent character; it may be nil, in which case it is
 100  deduced from DOWNCASE and UPCASE.
 101 EQUIVALENCES is a map that cyclicly permutes each equivalence class
 102  (of characters with the same canonical equivalent); it may be nil,
 103  in which case it is deduced from CANONICALIZE.  */)
 104   (Lisp_Object table)
 105 {
 106   return set_case_table (table, 0);
 107 }
 108
 109 DEFUN ("set-standard-case-table", Fset_standard_case_table, Sset_standard_case_table, 1, 1, 0,
 110        doc: /* Select a new standard case table for new buffers.
 111 See `set-case-table' for more info on case tables.  */)
 112   (Lisp_Object table)
 113 {
 114   return set_case_table (table, 1);
 115 }
 116
 117 static Lisp_Object
 118 set_case_table (Lisp_Object table, int standard)
 119 {
 120   Lisp_Object up, canon, eqv;
 121
 122   check_case_table (table);
 123
 124   up = XCHAR_TABLE (table)->extras[0];
 125   canon = XCHAR_TABLE (table)->extras[1];
 126   eqv = XCHAR_TABLE (table)->extras[2];
 127
 128   if (NILP (up))
 129     {
 130       up = Fmake_char_table (Qcase_table, Qnil);
 131       map_char_table (set_identity, Qnil, table, up);
 132       map_char_table (shuffle, Qnil, table, up);
 133       XCHAR_TABLE (table)->extras[0] = up;
 134     }
 135
 136   if (NILP (canon))
 137     {
 138       canon = Fmake_char_table (Qcase_table, Qnil);
 139       XCHAR_TABLE (table)->extras[1] = canon;
 140       map_char_table (set_canon, Qnil, table, table);
 141     }
 142
 143   if (NILP (eqv))
 144     {
 145       eqv = Fmake_char_table (Qcase_table, Qnil);
 146       map_char_table (set_identity, Qnil, canon, eqv);
 147       map_char_table (shuffle, Qnil, canon, eqv);
 148       XCHAR_TABLE (table)->extras[2] = eqv;
 149     }
 150
 151   /* This is so set_image_of_range_1 in regex.c can find the EQV table.  */
 152   XCHAR_TABLE (canon)->extras[2] = eqv;
 153
 154   if (standard)
 155     {
 156       Vascii_downcase_table = table;
 157       Vascii_upcase_table = up;
 158       Vascii_canon_table = canon;
 159       Vascii_eqv_table = eqv;
 160     }
 161   else
 162     {
 163       BVAR (current_buffer, downcase_table) = table;
 164       BVAR (current_buffer, upcase_table) = up;
 165       BVAR (current_buffer, case_canon_table) = canon;
 166       BVAR (current_buffer, case_eqv_table) = eqv;
 167     }
 168
 169   return table;
 170 }
 171 \f
 172 /* The following functions are called in map_char_table.  */
 173
 174 /* Set CANON char-table element for characters in RANGE to a
 175    translated ELT by UP and DOWN char-tables.  This is done only when
 176    ELT is a character.  The char-tables CANON, UP, and DOWN are in
 177    CASE_TABLE.  */
 178
 179 static void
 180 set_canon (Lisp_Object case_table, Lisp_Object range, Lisp_Object elt)
 181 {
 182   Lisp_Object up = XCHAR_TABLE (case_table)->extras[0];
 183   Lisp_Object canon = XCHAR_TABLE (case_table)->extras[1];
 184
 185   if (NATNUMP (elt))
 186     Fset_char_table_range (canon, range, Faref (case_table, Faref (up, elt)));
 187 }
 188
 189 /* Set elements of char-table TABLE for C to C itself.  C may be a
 190    cons specifying a character range.  In that case, set characters in
 191    that range to themselves.  This is done only when ELT is a
 192    character.  This is called in map_char_table.  */
 193
 194 static void
 195 set_identity (Lisp_Object table, Lisp_Object c, Lisp_Object elt)
 196 {
 197   if (NATNUMP (elt))
 198     {
 199       int from, to;
 200
 201       if (CONSP (c))
 202         {
 203           from = XINT (XCAR (c));
 204           to = XINT (XCDR (c));
 205         }
 206       else
 207         from = to = XINT (c);
 208       for (; from <= to; from++)
 209         CHAR_TABLE_SET (table, from, make_number (from));
 210     }
 211 }
 212
 213 /* Permute the elements of TABLE (which is initially an identity
 214    mapping) so that it has one cycle for each equivalence class
 215    induced by the translation table on which map_char_table is
 216    operated.  */
 217
 218 static void
 219 shuffle (Lisp_Object table, Lisp_Object c, Lisp_Object elt)
 220 {
 221   if (NATNUMP (elt))
 222     {
 223       int from, to;
 224
 225       if (CONSP (c))
 226         {
 227           from = XINT (XCAR (c));
 228           to = XINT (XCDR (c));
 229         }
 230       else
 231         from = to = XINT (c);
 232
 233       for (; from <= to; from++)
 234         {
 235           Lisp_Object tem = Faref (table, elt);
 236           Faset (table, elt, make_number (from));
 237           Faset (table, make_number (from), tem);
 238         }
 239     }
 240 }
 241 \f
 242 void
 243 init_casetab_once (void)
 244 {
 245   register int i;
 246   Lisp_Object down, up;
 247   Qcase_table = intern_c_string ("case-table");
 248   staticpro (&Qcase_table);
 249
 250   /* Intern this now in case it isn't already done.
 251      Setting this variable twice is harmless.
 252      But don't staticpro it here--that is done in alloc.c.  */
 253   Qchar_table_extra_slots = intern_c_string ("char-table-extra-slots");
 254
 255   /* Now we are ready to set up this property, so we can
 256      create char tables.  */
 257   Fput (Qcase_table, Qchar_table_extra_slots, make_number (3));
 258
 259   down = Fmake_char_table (Qcase_table, Qnil);
 260   Vascii_downcase_table = down;
 261   XCHAR_TABLE (down)->purpose = Qcase_table;
 262
 263   for (i = 0; i < 128; i++)
 264     {
 265       int c = (i >= 'A' && i <= 'Z') ? i + ('a' - 'A') : i;
 266       CHAR_TABLE_SET (down, i, make_number (c));
 267     }
 268
 269   XCHAR_TABLE (down)->extras[1] = Fcopy_sequence (down);
 270
 271   up = Fmake_char_table (Qcase_table, Qnil);
 272   XCHAR_TABLE (down)->extras[0] = up;
 273
 274   for (i = 0; i < 128; i++)
 275     {
 276       int c = ((i >= 'A' && i <= 'Z') ? i + ('a' - 'A')
 277                : ((i >= 'a' && i <= 'z') ? i + ('A' - 'a')
 278                   : i));
 279       CHAR_TABLE_SET (up, i, make_number (c));
 280     }
 281
 282   XCHAR_TABLE (down)->extras[2] = Fcopy_sequence (up);
 283
 284   /* Fill in what isn't filled in.  */
 285   set_case_table (down, 1);
 286 }
 287
 288 void
 289 syms_of_casetab (void)
 290 {
 291   Qcase_table_p = intern_c_string ("case-table-p");
 292   staticpro (&Qcase_table_p);
 293
 294   staticpro (&Vascii_canon_table);
 295   staticpro (&Vascii_downcase_table);
 296   staticpro (&Vascii_eqv_table);
 297   staticpro (&Vascii_upcase_table);
 298
 299   defsubr (&Scase_table_p);
 300   defsubr (&Scurrent_case_table);
 301   defsubr (&Sstandard_case_table);
 302   defsubr (&Sset_case_table);
 303   defsubr (&Sset_standard_case_table);
 304 }
 305