gcc/cppcharset.c

   1 /* CPP Library - charsets
   2    Copyright (C) 1998, 1999, 2000, 2001, 2002, 2003
   3    Free Software Foundation, Inc.
   4
   5    Broken out of c-lex.c Apr 2003, adding valid C99 UCN ranges.
   6
   7 This program is free software; you can redistribute it and/or modify it
   8 under the terms of the GNU General Public License as published by the
   9 Free Software Foundation; either version 2, or (at your option) any
  10 later version.
  11
  12 This program is distributed in the hope that it will be useful,
  13 but WITHOUT ANY WARRANTY; without even the implied warranty of
  14 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  15 GNU General Public License for more details.
  16
  17 You should have received a copy of the GNU General Public License
  18 along with this program; if not, write to the Free Software
  19 Foundation, 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.  */
  20
  21 #include "config.h"
  22 #include "system.h"
  23 #include "coretypes.h"
  24 #include "tm.h"
  25 #include "cpplib.h"
  26 #include "cpphash.h"
  27
  28 static int ucn_valid_in_identifier PARAMS ((cpp_reader *, cppchar_t));
  29
  30 /* [lex.charset]: The character designated by the universal character
  31    name \UNNNNNNNN is that character whose character short name in
  32    ISO/IEC 10646 is NNNNNNNN; the character designated by the
  33    universal character name \uNNNN is that character whose character
  34    short name in ISO/IEC 10646 is 0000NNNN.  If the hexadecimal value
  35    for a universal character name is less than 0x20 or in the range
  36    0x7F-0x9F (inclusive), or if the universal character name
  37    designates a character in the basic source character set, then the
  38    program is ill-formed.
  39
  40    *PSTR must be preceded by "\u" or "\U"; it is assumed that the
  41    buffer end is delimited by a non-hex digit.  Returns zero if UCNs
  42    are not part of the relevant standard, or if the string beginning
  43    at *PSTR doesn't syntactically match the form 'NNNN' or 'NNNNNNNN'.
  44
  45    Otherwise the nonzero value of the UCN, whether valid or invalid,
  46    is returned.  Diagnostics are emitted for invalid values.  PSTR
  47    is updated to point one beyond the UCN, or to the syntactically
  48    invalid character.
  49
  50    IDENTIFIER_POS is 0 when not in an identifier, 1 for the start of
  51    an identifier, or 2 otherwise.
  52 */
  53
  54 cppchar_t
  55 _cpp_valid_ucn (pfile, pstr, identifier_pos)
  56      cpp_reader *pfile;
  57      const uchar **pstr;
  58      int identifier_pos;
  59 {
  60   cppchar_t result, c;
  61   unsigned int length;
  62   const uchar *str = *pstr;
  63   const uchar *base = str - 2;
  64
  65   /* Only attempt to interpret a UCS for C++ and C99.  */
  66   if (!CPP_OPTION (pfile, cplusplus) && !CPP_OPTION (pfile, c99))
  67     return 0;
  68
  69   /* We don't accept UCNs for an EBCDIC target.  */
  70   if (CPP_OPTION (pfile, EBCDIC))
  71     return 0;
  72
  73   if (str[-1] == 'u')
  74     length = 4;
  75   else if (str[-1] == 'U')
  76     length = 8;
  77   else
  78     abort();
  79
  80   result = 0;
  81   do
  82     {
  83       c = *str;
  84       if (!ISXDIGIT (c))
  85         break;
  86       str++;
  87       result = (result << 4) + hex_value (c);
  88     }
  89   while (--length);
  90
  91   *pstr = str;
  92   if (length)
  93     /* We'll error when we try it out as the start of an identifier.  */
  94     cpp_error (pfile, DL_ERROR, "incomplete universal character name %.*s",
  95                (int) (str - base), base);
  96   /* The standard permits $, @ and ` to be specified as UCNs.  We use
  97      hex escapes so that this also works with EBCDIC hosts.  */
  98   else if ((result < 0xa0
  99             && (result != 0x24 && result != 0x40 && result != 0x60))
 100            || (result & 0x80000000)
 101            || (result >= 0xD800 && result <= 0xDFFF))
 102     {
 103       cpp_error (pfile, DL_ERROR, "%.*s is not a valid universal character",
 104                  (int) (str - base), base);
 105     }
 106   else if (identifier_pos)
 107     {
 108       int validity = ucn_valid_in_identifier (pfile, result);
 109
 110       if (validity == 0)
 111         cpp_error (pfile, DL_ERROR,
 112                    "universal character %.*s is not valid in an identifier",
 113                    (int) (str - base), base);
 114       else if (validity == 2 && identifier_pos == 1)
 115         cpp_error (pfile, DL_ERROR,
 116    "universal character %.*s is not valid at the start of an identifier",
 117                    (int) (str - base), base);
 118     }
 119
 120   if (result == 0)
 121     result = 1;
 122
 123   return result;
 124 }
 125
 126 /* Returns 1 if C is valid in an identifier, 2 if C is valid except at
 127    the start of an identifier, and 0 if C is not valid in an
 128    identifier.  We assume C has already gone through the checks of
 129    _cpp_valid_ucn.  */
 130 static int
 131 ucn_valid_in_identifier (pfile, c)
 132      cpp_reader *pfile;
 133      cppchar_t c;
 134 {
 135   /* None of the valid chars are outside the Basic Multilingual Plane (the
 136      low 16 bits).  */
 137   if (c > 0xffff)
 138     return 0;
 139
 140   if (CPP_OPTION (pfile, c99) || !CPP_PEDANTIC (pfile))
 141     {
 142       /* Latin.  */
 143       if (c == 0x0aa || c == 0x00ba || c == 0x207f || c == 0x1e9b)
 144         return 1;
 145
 146       /* Greek.  */
 147       if (c == 0x0386)
 148         return 1;
 149
 150       /* Cyrillic.  */
 151       if (c == 0x040c)
 152         return 1;
 153
 154       /* Hebrew.  */
 155       if ((c >= 0x05b0 && c <= 0x05b9)
 156           || (c >= 0x05bb && c <= 0x005bd)
 157           || c == 0x05bf
 158           || (c >= 0x05c1 && c <= 0x05c2))
 159         return 1;
 160
 161       /* Arabic.  */
 162       if ((c >= 0x06d0 && c <= 0x06dc)
 163           || c == 0x06e8
 164           || (c >= 0x06ea && c <= 0x06ed))
 165         return 1;
 166
 167       /* Devanagari */
 168       if ((c >= 0x0901 && c <= 0x0903)
 169           || (c >= 0x093e && c <= 0x094d)
 170           || (c >= 0x0950 && c <= 0x0952)
 171           || c == 0x0963)
 172         return 1;
 173
 174       /* Bengali */
 175       if ((c >= 0x0981 && c <= 0x0983)
 176           || (c >= 0x09be && c <= 0x09c4)
 177           || (c >= 0x09c7 && c <= 0x09c8)
 178           || (c >= 0x09cb && c <= 0x09cd)
 179           || (c >= 0x09e2 && c <= 0x09e3))
 180         return 1;
 181
 182       /* Gurmukhi */
 183       if (c == 0x0a02
 184           || (c >= 0x0a3e && c <= 0x0a42)
 185           || (c >= 0x0a47 && c <= 0x0a48)
 186           || (c >= 0x0a4b && c <= 0x0a4d)
 187           || (c == 0x0a74))
 188         return 1;
 189
 190       /* Gujarati */
 191       if ((c >= 0x0a81 && c <= 0x0a83)
 192           || (c >= 0x0abd && c <= 0x0ac5)
 193           || (c >= 0x0ac7 && c <= 0x0ac9)
 194           || (c >= 0x0acb && c <= 0x0acd)
 195           || (c == 0x0ad0))
 196         return 1;
 197
 198       /* Oriya */
 199       if ((c >= 0x0b01 && c <= 0x0b03)
 200           || (c >= 0x0b3e && c <= 0x0b43)
 201           || (c >= 0x0b47 && c <= 0x0b48)
 202           || (c >= 0x0b4b && c <= 0x0b4d))
 203         return 1;
 204
 205       /* Tamil */
 206       if ((c >= 0x0b82 && c <= 0x0b83)
 207           || (c >= 0x0bbe && c <= 0x0bc2)
 208           || (c >= 0x0bc6 && c <= 0x0bc8)
 209           || (c >= 0x0bc8 && c <= 0x0bcd))
 210         return 1;
 211
 212       /* Telugu */
 213       if ((c >= 0x0c01 && c <= 0x0c03)
 214           || (c >= 0x0c3e && c <= 0x0c44)
 215           || (c >= 0x0c46 && c <= 0x0c48)
 216           || (c >= 0x0c4a && c <= 0x0c4d))
 217         return 1;
 218
 219       /* Kannada */
 220       if ((c >= 0x0c82 && c <= 0x0c83)
 221           || (c >= 0x0cbe && c <= 0x0cc4)
 222           || (c >= 0x0cc6 && c <= 0x0cc8)
 223           || (c >= 0x0cca && c <= 0x0ccd)
 224           || c == 0x0cde)
 225         return 1;
 226
 227       /* Malayalam */
 228       if ((c >= 0x0d02 && c <= 0x0d03)
 229           || (c >= 0x0d3e && c <= 0x0d43)
 230           || (c >= 0x0d46 && c <= 0x0d48)
 231           || (c >= 0x0d4a && c <= 0x0d4d))
 232         return 1;
 233
 234       /* Thai */
 235       if ((c >= 0x0e01 && c <= 0x0e3a)
 236           || (c >= 0x0e40 && c <= 0x0e5b))
 237         return 1;
 238
 239       /* Lao */
 240       if ((c >= 0x0ead && c <= 0x0eae)
 241           || (c >= 0x0eb0 && c <= 0x0eb9)
 242           || (c >= 0x0ebb && c <= 0x0ebd)
 243           || (c >= 0x0ec0 && c <= 0x0ec4)
 244           || c == 0x0ec6
 245           || (c >= 0x0ec8 && c <= 0x0ecd)
 246           || (c >= 0x0edc && c <= 0x0ed))
 247         return 1;
 248
 249       /* Tibetan.  */
 250       if (c == 0x0f00
 251           || (c >= 0x0f18 && c <= 0x0f19)
 252           || c == 0x0f35
 253           || c == 0x0f37
 254           || c == 0x0f39
 255           || (c >= 0x0f3e && c <= 0x0f47)
 256           || (c >= 0x0f49 && c <= 0x0f69)
 257           || (c >= 0x0f71 && c <= 0x0f84)
 258           || (c >= 0x0f86 && c <= 0x0f8b)
 259           || (c >= 0x0f90 && c <= 0x0f95)
 260           || c == 0x0f97
 261           || (c >= 0x0f99 && c <= 0x0fad)
 262           || (c >= 0x0fb1 && c <= 0x0fb7)
 263           || c == 0x0fb9)
 264         return 1;
 265
 266       /* Katakana */
 267       if ((c >= 0x30a1 && c <= 0x30f6)
 268           || (c >= 0x30fb && c <= 0x30fc))
 269         return 1;
 270
 271       /* CJK Unified Ideographs.  */
 272       if (c >= 0x4e00 && c <= 0x9fa5)
 273         return 1;
 274
 275       /* Hangul.  */
 276       if (c >= 0xac00 && c <= 0xd7a3)
 277         return 1;
 278
 279       /* Digits.  */
 280       if ((c >= 0x0660 && c <= 0x0669)
 281           || (c >= 0x06f0 && c <= 0x06f9)
 282           || (c >= 0x0966 && c <= 0x096f)
 283           || (c >= 0x09e6 && c <= 0x09ef)
 284           || (c >= 0x0a66 && c <= 0x0a6f)
 285           || (c >= 0x0ae6 && c <= 0x0aef)
 286           || (c >= 0x0b66 && c <= 0x0b6f)
 287           || (c >= 0x0be7 && c <= 0x0bef)
 288           || (c >= 0x0c66 && c <= 0x0c6f)
 289           || (c >= 0x0ce6 && c <= 0x0cef)
 290           || (c >= 0x0d66 && c <= 0x0d6f)
 291           || (c >= 0x0e50 && c <= 0x0e59)
 292           || (c >= 0x0ed0 && c <= 0x0ed9)
 293           || (c >= 0x0f20 && c <= 0x0f33))
 294         return 2;
 295
 296       /* Special characters.  */
 297       if (c == 0x00b5
 298           || c == 0x00b7
 299           || (c >= 0x02b0 && c <= 0x02b8)
 300           || c == 0x02bb
 301           || (c >= 0x02bd && c <= 0x02c1)
 302           || (c >= 0x02d0 && c <= 0x02d1)
 303           || (c >= 0x02e0 && c <= 0x02e4)
 304           || c == 0x037a
 305           || c == 0x0559
 306           || c == 0x093d
 307           || c == 0x0b3d
 308           || c == 0x1fbe
 309           || (c >= 0x203f && c <= 0x2040)
 310           || c == 0x2102
 311           || c == 0x2107
 312           || (c >= 0x210a && c <= 0x2113)
 313           || c == 0x2115
 314           || (c >= 0x2118 && c <= 0x211d)
 315           || c == 0x2124
 316           || c == 0x2126
 317           || c == 0x2128
 318           || (c >= 0x212a && c <= 0x2131)
 319           || (c >= 0x2133 && c <= 0x2138)
 320           || (c >= 0x2160 && c <= 0x2182)
 321           || (c >= 0x3005 && c <= 0x3007)
 322           || (c >= 0x3021 && c <= 0x3029))
 323         return 1;
 324     }
 325
 326   if (CPP_OPTION (pfile, cplusplus) || !CPP_PEDANTIC (pfile))
 327     {
 328       /* Greek.  */
 329       if (c == 0x0384)
 330         return 1;
 331
 332       /* Cyrillic.  */
 333       if (c == 0x040d)
 334         return 1;
 335
 336       /* Hebrew.  */
 337       if (c >= 0x05f3 && c <= 0x05f4)
 338         return 1;
 339
 340       /* Lao.  */
 341       if ((c >= 0x0ead && c <= 0x0eb0)
 342           || (c == 0x0eb2)
 343           || (c == 0x0eb3)
 344           || (c == 0x0ebd)
 345           || (c >= 0x0ec0 && c <= 0x0ec4)
 346           || (c == 0x0ec6))
 347         return 1;
 348
 349       /* Hiragana */
 350       if (c == 0x3094
 351           || (c >= 0x309d && c <= 0x309e))
 352         return 1;
 353
 354       /* Katakana */
 355       if ((c >= 0x30a1 && c <= 0x30fe))
 356         return 1;
 357
 358       /* Hangul */
 359       if ((c >= 0x1100 && c <= 0x1159)
 360           || (c >= 0x1161 && c <= 0x11a2)
 361           || (c >= 0x11a8 && c <= 0x11f9))
 362         return 1;
 363
 364       /* CJK Unified Ideographs */
 365       if ((c >= 0xf900 && c <= 0xfa2d)
 366           || (c >= 0xfb1f && c <= 0xfb36)
 367           || (c >= 0xfb38 && c <= 0xfb3c)
 368           || (c == 0xfb3e)
 369           || (c >= 0xfb40 && c <= 0xfb41)
 370           || (c >= 0xfb42 && c <= 0xfb44)
 371           || (c >= 0xfb46 && c <= 0xfbb1)
 372           || (c >= 0xfbd3 && c <= 0xfd3f)
 373           || (c >= 0xfd50 && c <= 0xfd8f)
 374           || (c >= 0xfd92 && c <= 0xfdc7)
 375           || (c >= 0xfdf0 && c <= 0xfdfb)
 376           || (c >= 0xfe70 && c <= 0xfe72)
 377           || (c == 0xfe74)
 378           || (c >= 0xfe76 && c <= 0xfefc)
 379           || (c >= 0xff21 && c <= 0xff3a)
 380           || (c >= 0xff41 && c <= 0xff5a)
 381           || (c >= 0xff66 && c <= 0xffbe)
 382           || (c >= 0xffc2 && c <= 0xffc7)
 383           || (c >= 0xffca && c <= 0xffcf)
 384           || (c >= 0xffd2 && c <= 0xffd7)
 385           || (c >= 0xffda && c <= 0xffdc)
 386           || (c >= 0x4e00 && c <= 0x9fa5))
 387         return 1;
 388     }
 389
 390   /* Latin */
 391   if ((c >= 0x00c0 && c <= 0x00d6)
 392       || (c >= 0x00d8 && c <= 0x00f6)
 393       || (c >= 0x00f8 && c <= 0x01f5)
 394       || (c >= 0x01fa && c <= 0x0217)
 395       || (c >= 0x0250 && c <= 0x02a8)
 396       || (c >= 0x1e00 && c <= 0x1e9a)
 397       || (c >= 0x1ea0 && c <= 0x1ef9))
 398     return 1;
 399
 400   /* Greek */
 401   if ((c >= 0x0388 && c <= 0x038a)
 402       || (c == 0x038c)
 403       || (c >= 0x038e && c <= 0x03a1)
 404       || (c >= 0x03a3 && c <= 0x03ce)
 405       || (c >= 0x03d0 && c <= 0x03d6)
 406       || (c == 0x03da)
 407       || (c == 0x03dc)
 408       || (c == 0x03de)
 409       || (c == 0x03e0)
 410       || (c >= 0x03e2 && c <= 0x03f3)
 411       || (c >= 0x1f00 && c <= 0x1f15)
 412       || (c >= 0x1f18 && c <= 0x1f1d)
 413       || (c >= 0x1f20 && c <= 0x1f45)
 414       || (c >= 0x1f48 && c <= 0x1f4d)
 415       || (c >= 0x1f50 && c <= 0x1f57)
 416       || (c == 0x1f59)
 417       || (c == 0x1f5b)
 418       || (c == 0x1f5d)
 419       || (c >= 0x1f5f && c <= 0x1f7d)
 420       || (c >= 0x1f80 && c <= 0x1fb4)
 421       || (c >= 0x1fb6 && c <= 0x1fbc)
 422       || (c >= 0x1fc2 && c <= 0x1fc4)
 423       || (c >= 0x1fc6 && c <= 0x1fcc)
 424       || (c >= 0x1fd0 && c <= 0x1fd3)
 425       || (c >= 0x1fd6 && c <= 0x1fdb)
 426       || (c >= 0x1fe0 && c <= 0x1fec)
 427       || (c >= 0x1ff2 && c <= 0x1ff4)
 428       || (c >= 0x1ff6 && c <= 0x1ffc))
 429     return 1;
 430
 431   /* Cyrillic */
 432   if ((c >= 0x0401 && c <= 0x040c)
 433       || (c >= 0x040f && c <= 0x044f)
 434       || (c >= 0x0451 && c <= 0x045c)
 435       || (c >= 0x045e && c <= 0x0481)
 436       || (c >= 0x0490 && c <= 0x04c4)
 437       || (c >= 0x04c7 && c <= 0x04c8)
 438       || (c >= 0x04cb && c <= 0x04cc)
 439       || (c >= 0x04d0 && c <= 0x04eb)
 440       || (c >= 0x04ee && c <= 0x04f5)
 441       || (c >= 0x04f8 && c <= 0x04f9))
 442     return 1;
 443
 444   /* Armenian */
 445   if ((c >= 0x0531 && c <= 0x0556)
 446       || (c >= 0x0561 && c <= 0x0587))
 447     return 1;
 448
 449   /* Hebrew */
 450   if ((c >= 0x05d0 && c <= 0x05ea)
 451       || (c >= 0x05f0 && c <= 0x05f2))
 452     return 1;
 453
 454   /* Arabic */
 455   if ((c >= 0x0621 && c <= 0x063a)
 456       || (c >= 0x0640 && c <= 0x0652)
 457       || (c >= 0x0670 && c <= 0x06b7)
 458       || (c >= 0x06ba && c <= 0x06be)
 459       || (c >= 0x06c0 && c <= 0x06ce)
 460       || (c >= 0x06e5 && c <= 0x06e7))
 461     return 1;
 462
 463   /* Devanagari */
 464   if ((c >= 0x0905 && c <= 0x0939)
 465       || (c >= 0x0958 && c <= 0x0962))
 466     return 1;
 467
 468   /* Bengali */
 469   if ((c >= 0x0985 && c <= 0x098c)
 470       || (c >= 0x098f && c <= 0x0990)
 471       || (c >= 0x0993 && c <= 0x09a8)
 472       || (c >= 0x09aa && c <= 0x09b0)
 473       || (c == 0x09b2)
 474       || (c >= 0x09b6 && c <= 0x09b9)
 475       || (c >= 0x09dc && c <= 0x09dd)
 476       || (c >= 0x09df && c <= 0x09e1)
 477       || (c >= 0x09f0 && c <= 0x09f1))
 478     return 1;
 479
 480   /* Gurmukhi */
 481   if ((c >= 0x0a05 && c <= 0x0a0a)
 482       || (c >= 0x0a0f && c <= 0x0a10)
 483       || (c >= 0x0a13 && c <= 0x0a28)
 484       || (c >= 0x0a2a && c <= 0x0a30)
 485       || (c >= 0x0a32 && c <= 0x0a33)
 486       || (c >= 0x0a35 && c <= 0x0a36)
 487       || (c >= 0x0a38 && c <= 0x0a39)
 488       || (c >= 0x0a59 && c <= 0x0a5c)
 489       || (c == 0x0a5e))
 490     return 1;
 491
 492   /* Gujarati */
 493   if ((c >= 0x0a85 && c <= 0x0a8b)
 494       || (c == 0x0a8d)
 495       || (c >= 0x0a8f && c <= 0x0a91)
 496       || (c >= 0x0a93 && c <= 0x0aa8)
 497       || (c >= 0x0aaa && c <= 0x0ab0)
 498       || (c >= 0x0ab2 && c <= 0x0ab3)
 499       || (c >= 0x0ab5 && c <= 0x0ab9)
 500       || (c == 0x0ae0))
 501     return 1;
 502
 503   /* Oriya */
 504   if ((c >= 0x0b05 && c <= 0x0b0c)
 505       || (c >= 0x0b0f && c <= 0x0b10)
 506       || (c >= 0x0b13 && c <= 0x0b28)
 507       || (c >= 0x0b2a && c <= 0x0b30)
 508       || (c >= 0x0b32 && c <= 0x0b33)
 509       || (c >= 0x0b36 && c <= 0x0b39)
 510       || (c >= 0x0b5c && c <= 0x0b5d)
 511       || (c >= 0x0b5f && c <= 0x0b61))
 512     return 1;
 513
 514   /* Tamil */
 515   if ((c >= 0x0b85 && c <= 0x0b8a)
 516       || (c >= 0x0b8e && c <= 0x0b90)
 517       || (c >= 0x0b92 && c <= 0x0b95)
 518       || (c >= 0x0b99 && c <= 0x0b9a)
 519       || (c == 0x0b9c)
 520       || (c >= 0x0b9e && c <= 0x0b9f)
 521       || (c >= 0x0ba3 && c <= 0x0ba4)
 522       || (c >= 0x0ba8 && c <= 0x0baa)
 523       || (c >= 0x0bae && c <= 0x0bb5)
 524       || (c >= 0x0bb7 && c <= 0x0bb9))
 525     return 1;
 526
 527   /* Telugu */
 528   if ((c >= 0x0c05 && c <= 0x0c0c)
 529       || (c >= 0x0c0e && c <= 0x0c10)
 530       || (c >= 0x0c12 && c <= 0x0c28)
 531       || (c >= 0x0c2a && c <= 0x0c33)
 532       || (c >= 0x0c35 && c <= 0x0c39)
 533       || (c >= 0x0c60 && c <= 0x0c61))
 534     return 1;
 535
 536   /* Kannada */
 537   if ((c >= 0x0c85 && c <= 0x0c8c)
 538       || (c >= 0x0c8e && c <= 0x0c90)
 539       || (c >= 0x0c92 && c <= 0x0ca8)
 540       || (c >= 0x0caa && c <= 0x0cb3)
 541       || (c >= 0x0cb5 && c <= 0x0cb9)
 542       || (c >= 0x0ce0 && c <= 0x0ce1))
 543     return 1;
 544
 545   /* Malayalam */
 546   if ((c >= 0x0d05 && c <= 0x0d0c)
 547       || (c >= 0x0d0e && c <= 0x0d10)
 548       || (c >= 0x0d12 && c <= 0x0d28)
 549       || (c >= 0x0d2a && c <= 0x0d39)
 550       || (c >= 0x0d60 && c <= 0x0d61))
 551     return 1;
 552
 553   /* Thai */
 554   if ((c >= 0x0e01 && c <= 0x0e30)
 555       || (c >= 0x0e32 && c <= 0x0e33)
 556       || (c >= 0x0e40 && c <= 0x0e46)
 557       || (c >= 0x0e4f && c <= 0x0e5b))
 558     return 1;
 559
 560   /* Lao */
 561   if ((c >= 0x0e81 && c <= 0x0e82)
 562       || (c == 0x0e84)
 563       || (c == 0x0e87)
 564       || (c == 0x0e88)
 565       || (c == 0x0e8a)
 566       || (c == 0x0e8d)
 567       || (c >= 0x0e94 && c <= 0x0e97)
 568       || (c >= 0x0e99 && c <= 0x0e9f)
 569       || (c >= 0x0ea1 && c <= 0x0ea3)
 570       || (c == 0x0ea5)
 571       || (c == 0x0ea7)
 572       || (c == 0x0eaa)
 573       || (c == 0x0eab))
 574     return 1;
 575
 576   /* Georgian */
 577   if ((c >= 0x10a0 && c <= 0x10c5)
 578       || (c >= 0x10d0 && c <= 0x10f6))
 579     return 1;
 580
 581   /* Hiragana */
 582   if ((c >= 0x3041 && c <= 0x3093)
 583       || (c >= 0x309b && c <= 0x309c))
 584     return 1;
 585
 586   /* Bopmofo */
 587   if ((c >= 0x3105 && c <= 0x312c))
 588     return 1;
 589
 590   return 0;
 591 }