src/encoding.c

   1 /* Copyright (c) 1993-2003
   2  *      Juergen Weigert (jnweiger@immd4.informatik.uni-erlangen.de)
   3  *      Michael Schroeder (mlschroe@immd4.informatik.uni-erlangen.de)
   4  * Copyright (c) 1987 Oliver Laumann
   5  *
   6  * This program is free software; you can redistribute it and/or modify
   7  * it under the terms of the GNU General Public License as published by
   8  * the Free Software Foundation; either version 2, or (at your option)
   9  * any later version.
  10  *
  11  * This program is distributed in the hope that it will be useful,
  12  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  13  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  14  * GNU General Public License for more details.
  15  *
  16  * You should have received a copy of the GNU General Public License
  17  * along with this program (see the file COPYING); if not, write to the
  18  * Free Software Foundation, Inc.,
  19  * 59 Temple Place - Suite 330, Boston, MA  02111-1307, USA
  20  *
  21  ****************************************************************
  22  */
  23
  24 #include <sys/types.h>
  25
  26 #include "config.h"
  27 #include "screen.h"
  28 #include "extern.h"
  29
  30 #ifdef ENCODINGS
  31
  32 extern unsigned char *null;
  33 extern struct display *display, *displays;
  34 extern struct layer *flayer;
  35
  36 extern char *screenencodings;
  37
  38 #ifdef DW_CHARS
  39 extern int cjkwidth;
  40 #endif
  41
  42 static int  encmatch __P((char *, char *));
  43 # ifdef UTF8
  44 static int   recode_char __P((int, int, int));
  45 static int   recode_char_to_encoding __P((int, int));
  46 static void  comb_tofront __P((int, int));
  47 #  ifdef DW_CHARS
  48 static int   recode_char_dw __P((int, int *, int, int));
  49 static int   recode_char_dw_to_encoding __P((int, int *, int));
  50 #  endif
  51 # endif
  52
  53 struct encoding {
  54   char *name;
  55   char *charsets;
  56   int  deffont;
  57   int  usegr;
  58   int  noc1;
  59   char *fontlist;
  60 };
  61
  62 /* big5 font:   ^X */
  63 /* KOI8-R font: 96 ! */
  64 /* CP1251 font: 96 ? */
  65
  66 struct encoding encodings[] = {
  67   { "C",                0,              0,              0, 0, 0 },
  68   { "eucJP",            "B\002I\00401", 0,              1, 0, "\002\004I" },
  69   { "SJIS",             "BIBB01",       0,              1, 1, "\002I" },
  70   { "eucKR",            "B\003BB01",    0,              1, 0, "\003" },
  71   { "eucCN",            "B\001BB01",    0,              1, 0, "\001" },
  72   { "Big5",             "B\030BB01",    0,              1, 0, "\030" },
  73   { "KOI8-R",           0,              0x80|'!',       0, 1, 0 },
  74   { "CP1251",           0,              0x80|'?',       0, 1, 0 },
  75   { "UTF-8",            0,              -1,             0, 0, 0 },
  76   { "ISO8859-2",        0,              0x80|'B',       0, 0, 0 },
  77   { "ISO8859-3",        0,              0x80|'C',       0, 0, 0 },
  78   { "ISO8859-4",        0,              0x80|'D',       0, 0, 0 },
  79   { "ISO8859-5",        0,              0x80|'L',       0, 0, 0 },
  80   { "ISO8859-6",        0,              0x80|'G',       0, 0, 0 },
  81   { "ISO8859-7",        0,              0x80|'F',       0, 0, 0 },
  82   { "ISO8859-8",        0,              0x80|'H',       0, 0, 0 },
  83   { "ISO8859-9",        0,              0x80|'M',       0, 0, 0 },
  84   { "ISO8859-10",       0,              0x80|'V',       0, 0, 0 },
  85   { "ISO8859-15",       0,              0x80|'b',       0, 0, 0 },
  86   { "jis",              0,              0,              0, 0, "\002\004I" },
  87   { "GBK",              "B\031BB01",    0x80|'b',       1, 1, "\031" }
  88 };
  89
  90 #ifdef UTF8
  91
  92 static unsigned short builtin_tabs[][2] = {
  93   { 0x30, 0 },          /* 0: special graphics (line drawing) */
  94   { 0x005f, 0x25AE },
  95   { 0x0060, 0x25C6 },
  96   { 0x0061, 0x2592 },
  97   { 0x0062, 0x2409 },
  98   { 0x0063, 0x240C },
  99   { 0x0064, 0x240D },
 100   { 0x0065, 0x240A },
 101   { 0x0066, 0x00B0 },
 102   { 0x0067, 0x00B1 },
 103   { 0x0068, 0x2424 },
 104   { 0x0069, 0x240B },
 105   { 0x006a, 0x2518 },
 106   { 0x006b, 0x2510 },
 107   { 0x006c, 0x250C },
 108   { 0x006d, 0x2514 },
 109   { 0x006e, 0x253C },
 110   { 0x006f, 0x23BA },
 111   { 0x0070, 0x23BB },
 112   { 0x0071, 0x2500 },
 113   { 0x0072, 0x23BC },
 114   { 0x0073, 0x23BD },
 115   { 0x0074, 0x251C },
 116   { 0x0075, 0x2524 },
 117   { 0x0076, 0x2534 },
 118   { 0x0077, 0x252C },
 119   { 0x0078, 0x2502 },
 120   { 0x0079, 0x2264 },
 121   { 0x007a, 0x2265 },
 122   { 0x007b, 0x03C0 },
 123   { 0x007c, 0x2260 },
 124   { 0x007d, 0x00A3 },
 125   { 0x007e, 0x00B7 },
 126   { 0, 0},
 127
 128   { 0x34, 0 },          /* 4: Dutch */
 129   { 0x0023, 0x00a3 },
 130   { 0x0040, 0x00be },
 131   { 0x005b, 0x00ff },
 132   { 0x005c, 0x00bd },
 133   { 0x005d, 0x007c },
 134   { 0x007b, 0x00a8 },
 135   { 0x007c, 0x0066 },
 136   { 0x007d, 0x00bc },
 137   { 0x007e, 0x00b4 },
 138   { 0, 0},
 139
 140   { 0x35, 0 },          /* 5: Finnish */
 141   { 0x005b, 0x00c4 },
 142   { 0x005c, 0x00d6 },
 143   { 0x005d, 0x00c5 },
 144   { 0x005e, 0x00dc },
 145   { 0x0060, 0x00e9 },
 146   { 0x007b, 0x00e4 },
 147   { 0x007c, 0x00f6 },
 148   { 0x007d, 0x00e5 },
 149   { 0x007e, 0x00fc },
 150   { 0, 0},
 151
 152   { 0x36, 0 },          /* 6: Norwegian/Danish */
 153   { 0x0040, 0x00c4 },
 154   { 0x005b, 0x00c6 },
 155   { 0x005c, 0x00d8 },
 156   { 0x005d, 0x00c5 },
 157   { 0x005e, 0x00dc },
 158   { 0x0060, 0x00e4 },
 159   { 0x007b, 0x00e6 },
 160   { 0x007c, 0x00f8 },
 161   { 0x007d, 0x00e5 },
 162   { 0x007e, 0x00fc },
 163   { 0, 0},
 164
 165   { 0x37, 0 },          /* 7: Swedish */
 166   { 0x0040, 0x00c9 },
 167   { 0x005b, 0x00c4 },
 168   { 0x005c, 0x00d6 },
 169   { 0x005d, 0x00c5 },
 170   { 0x005e, 0x00dc },
 171   { 0x0060, 0x00e9 },
 172   { 0x007b, 0x00e4 },
 173   { 0x007c, 0x00f6 },
 174   { 0x007d, 0x00e5 },
 175   { 0x007e, 0x00fc },
 176   { 0, 0},
 177
 178   { 0x3d, 0},           /* =: Swiss */
 179   { 0x0023, 0x00f9 },
 180   { 0x0040, 0x00e0 },
 181   { 0x005b, 0x00e9 },
 182   { 0x005c, 0x00e7 },
 183   { 0x005d, 0x00ea },
 184   { 0x005e, 0x00ee },
 185   { 0x005f, 0x00e8 },
 186   { 0x0060, 0x00f4 },
 187   { 0x007b, 0x00e4 },
 188   { 0x007c, 0x00f6 },
 189   { 0x007d, 0x00fc },
 190   { 0x007e, 0x00fb },
 191   { 0, 0},
 192
 193   { 0x41, 0},           /* A: UK */
 194   { 0x0023, 0x00a3 },
 195   { 0, 0},
 196
 197   { 0x4b, 0},           /* K: German */
 198   { 0x0040, 0x00a7 },
 199   { 0x005b, 0x00c4 },
 200   { 0x005c, 0x00d6 },
 201   { 0x005d, 0x00dc },
 202   { 0x007b, 0x00e4 },
 203   { 0x007c, 0x00f6 },
 204   { 0x007d, 0x00fc },
 205   { 0x007e, 0x00df },
 206   { 0, 0},
 207
 208   { 0x51, 0},           /* Q: French Canadian */
 209   { 0x0040, 0x00e0 },
 210   { 0x005b, 0x00e2 },
 211   { 0x005c, 0x00e7 },
 212   { 0x005d, 0x00ea },
 213   { 0x005e, 0x00ee },
 214   { 0x0060, 0x00f4 },
 215   { 0x007b, 0x00e9 },
 216   { 0x007c, 0x00f9 },
 217   { 0x007d, 0x00e8 },
 218   { 0x007e, 0x00fb },
 219   { 0, 0},
 220
 221   { 0x52, 0},           /* R: French */
 222   { 0x0023, 0x00a3 },
 223   { 0x0040, 0x00e0 },
 224   { 0x005b, 0x00b0 },
 225   { 0x005c, 0x00e7 },
 226   { 0x005d, 0x00a7 },
 227   { 0x007b, 0x00e9 },
 228   { 0x007c, 0x00f9 },
 229   { 0x007d, 0x00e8 },
 230   { 0x007e, 0x00a8 },
 231   { 0, 0},
 232
 233   { 0x59, 0},           /* Y: Italian */
 234   { 0x0023, 0x00a3 },
 235   { 0x0040, 0x00a7 },
 236   { 0x005b, 0x00b0 },
 237   { 0x005c, 0x00e7 },
 238   { 0x005d, 0x00e9 },
 239   { 0x0060, 0x00f9 },
 240   { 0x007b, 0x00e0 },
 241   { 0x007c, 0x00f2 },
 242   { 0x007d, 0x00e8 },
 243   { 0x007e, 0x00ec },
 244   { 0, 0},
 245
 246   { 0x5a, 0},           /* Z: Spanish */
 247   { 0x0023, 0x00a3 },
 248   { 0x0040, 0x00a7 },
 249   { 0x005b, 0x00a1 },
 250   { 0x005c, 0x00d1 },
 251   { 0x005d, 0x00bf },
 252   { 0x007b, 0x00b0 },
 253   { 0x007c, 0x00f1 },
 254   { 0x007d, 0x00e7 },
 255   { 0, 0},
 256
 257   { 0xe2, 0},           /* 96-b: ISO-8859-15 */
 258   { 0x00a4, 0x20ac },
 259   { 0x00a6, 0x0160 },
 260   { 0x00a8, 0x0161 },
 261   { 0x00b4, 0x017D },
 262   { 0x00b8, 0x017E },
 263   { 0x00bc, 0x0152 },
 264   { 0x00bd, 0x0153 },
 265   { 0x00be, 0x0178 },
 266   { 0, 0},
 267
 268   { 0x4a, 0},           /* J: JIS 0201 Roman */
 269   { 0x005c, 0x00a5 },
 270   { 0x007e, 0x203e },
 271   { 0, 0},
 272
 273   { 0x49, 0},           /* I: halfwidth katakana */
 274   { 0x0021, 0xff61 },
 275   { 0x005f|0x8000, 0xff9f },
 276   { 0, 0},
 277
 278   { 0, 0}
 279 };
 280
 281 struct recodetab
 282 {
 283   unsigned short (*tab)[2];
 284   int flags;
 285 };
 286
 287 #define RECODETAB_ALLOCED       1
 288 #define RECODETAB_BUILTIN       2
 289 #define RECODETAB_TRIED         4
 290
 291 static struct recodetab recodetabs[256];
 292
 293 void
 294 InitBuiltinTabs()
 295 {
 296   unsigned short (*p)[2];
 297   for (p = builtin_tabs; (*p)[0]; p++)
 298     {
 299       recodetabs[(*p)[0]].flags = RECODETAB_BUILTIN;
 300       recodetabs[(*p)[0]].tab = p + 1;
 301       p++;
 302       while((*p)[0])
 303         p++;
 304     }
 305 }
 306
 307 static int
 308 recode_char(c, to_utf, font)
 309 int c, to_utf, font;
 310 {
 311   int f;
 312   unsigned short (*p)[2];
 313
 314   if (to_utf)
 315     {
 316       if (c < 256)
 317         return c;
 318       f = (c >> 8) & 0xff;
 319       c &= 0xff;
 320       /* map aliases to keep the table small */
 321       switch (f)
 322         {
 323           case 'C':
 324             f ^= ('C' ^ '5');
 325             break;
 326           case 'E':
 327             f ^= ('E' ^ '6');
 328             break;
 329           case 'H':
 330             f ^= ('H' ^ '7');
 331             break;
 332           default:
 333             break;
 334         }
 335       p = recodetabs[f].tab;
 336       if (p == 0 && recodetabs[f].flags == 0)
 337         {
 338           LoadFontTranslation(f, 0);
 339           p = recodetabs[f].tab;
 340         }
 341       if (p)
 342         for (; (*p)[0]; p++)
 343           {
 344             if ((p[0][0] & 0x8000) && (c <= (p[0][0] & 0x7fff)) && c >= p[-1][0])
 345               return c - p[-1][0] + p[-1][1];
 346             if ((*p)[0] == c)
 347               return (*p)[1];
 348           }
 349       return c & 0xff;  /* map to latin1 */
 350     }
 351   if (font == -1)
 352     {
 353       if (c < 256)
 354         return c;       /* latin1 */
 355       for (font = 32; font < 128; font++)
 356         {
 357           p = recodetabs[font].tab;
 358           if (p)
 359             for (; (*p)[1]; p++)
 360               {
 361                 if ((p[0][0] & 0x8000) && c <= p[0][1] && c >= p[-1][1])
 362                   return (c - p[-1][1] + p[-1][0]) | (font << 8);
 363                 if ((*p)[1] == c)
 364                   return (*p)[0] | (font << 8);
 365               }
 366         }
 367       return '?';
 368     }
 369   if (c < 128 && (font & 128) != 0)
 370     return c;
 371   if (font >= 32)
 372     {
 373       p = recodetabs[font].tab;
 374       if (p == 0 && recodetabs[font].flags == 0)
 375         {
 376           LoadFontTranslation(font, 0);
 377           p = recodetabs[font].tab;
 378         }
 379       if (p)
 380         for (; (*p)[1]; p++)
 381           {
 382             if ((p[0][0] & 0x8000) && c <= p[0][1] && c >= p[-1][1])
 383               return (c - p[-1][1] + p[-1][0]) | (font & 128 ? 0 : font << 8);
 384             if ((*p)[1] == c)
 385               return (*p)[0] | (font & 128 ? 0 : font << 8);
 386           }
 387     }
 388   return -1;
 389 }
 390
 391
 392 #ifdef DW_CHARS
 393 static int
 394 recode_char_dw(c, c2p, to_utf, font)
 395 int c, *c2p, to_utf, font;
 396 {
 397   int f;
 398   unsigned short (*p)[2];
 399
 400   if (to_utf)
 401     {
 402       f = (c >> 8) & 0xff;
 403       c = (c & 255) << 8 | (*c2p & 255);
 404       *c2p = 0xffff;
 405       p = recodetabs[f].tab;
 406       if (p == 0 && recodetabs[f].flags == 0)
 407         {
 408           LoadFontTranslation(f, 0);
 409           p = recodetabs[f].tab;
 410         }
 411       if (p)
 412         for (; (*p)[0]; p++)
 413           if ((*p)[0] == c)
 414             {
 415 #ifdef DW_CHARS
 416               if (!utf8_isdouble((*p)[1]))
 417                 *c2p = ' ';
 418 #endif
 419               return (*p)[1];
 420             }
 421       return UCS_REPL_DW;
 422     }
 423   if (font == -1)
 424     {
 425       for (font = 0; font < 030; font++)
 426         {
 427           p = recodetabs[font].tab;
 428           if (p)
 429             for (; (*p)[1]; p++)
 430               if ((*p)[1] == c)
 431                 {
 432                   *c2p = ((*p)[0] & 255) | font << 8 | 0x8000;
 433                   return ((*p)[0] >> 8) | font << 8;
 434                 }
 435         }
 436       *c2p = '?';
 437       return '?';
 438     }
 439   if (font < 32)
 440     {
 441       p = recodetabs[font].tab;
 442       if (p == 0 && recodetabs[font].flags == 0)
 443         {
 444           LoadFontTranslation(font, 0);
 445           p = recodetabs[font].tab;
 446         }
 447       if (p)
 448         for (; (*p)[1]; p++)
 449           if ((*p)[1] == c)
 450             {
 451               *c2p = ((*p)[0] & 255) | font << 8 | 0x8000;
 452               return ((*p)[0] >> 8) | font << 8;
 453             }
 454     }
 455   return -1;
 456 }
 457 #endif
 458
 459 static int
 460 recode_char_to_encoding(c, encoding)
 461 int c, encoding;
 462 {
 463   char *fp;
 464   int x;
 465
 466   if (encoding == UTF8)
 467     return recode_char(c, 1, -1);
 468   if ((fp = encodings[encoding].fontlist) != 0)
 469     while(*fp)
 470       if ((x = recode_char(c, 0, (unsigned char)*fp++)) != -1)
 471         return x;
 472   if (encodings[encoding].deffont)
 473     if ((x = recode_char(c, 0, encodings[encoding].deffont)) != -1)
 474       return x;
 475   return recode_char(c, 0, -1);
 476 }
 477
 478 #ifdef DW_CHARS
 479 static int
 480 recode_char_dw_to_encoding(c, c2p, encoding)
 481 int c, *c2p, encoding;
 482 {
 483   char *fp;
 484   int x;
 485
 486   if (encoding == UTF8)
 487     return recode_char_dw(c, c2p, 1, -1);
 488   if ((fp = encodings[encoding].fontlist) != 0)
 489     while(*fp)
 490       if ((x = recode_char_dw(c, c2p, 0, (unsigned char)*fp++)) != -1)
 491         return x;
 492   if (encodings[encoding].deffont)
 493     if ((x = recode_char_dw(c, c2p, 0, encodings[encoding].deffont)) != -1)
 494       return x;
 495   return recode_char_dw(c, c2p, 0, -1);
 496 }
 497 #endif
 498
 499
 500 struct mchar *
 501 recode_mchar(mc, from, to)
 502 struct mchar *mc;
 503 int from, to;
 504 {
 505   static struct mchar rmc;
 506   int c;
 507
 508   debug3("recode_mchar %02x from %d to %d\n", mc->image, from, to);
 509   if (from == to || (from != UTF8 && to != UTF8))
 510     return mc;
 511   rmc = *mc;
 512   if (rmc.font == 0 && from != UTF8)
 513     rmc.font = encodings[from].deffont;
 514   if (rmc.font == 0)    /* latin1 is the same in unicode */
 515     return mc;
 516   c = rmc.image | (rmc.font << 8);
 517 #ifdef DW_CHARS
 518   if (rmc.mbcs)
 519     {
 520       int c2 = rmc.mbcs;
 521       c = recode_char_dw_to_encoding(c, &c2, to);
 522       rmc.mbcs = c2;
 523     }
 524   else
 525 #endif
 526     c = recode_char_to_encoding(c, to);
 527   rmc.image = c & 255;
 528   rmc.font = c >> 8 & 255;
 529   return &rmc;
 530 }
 531
 532 struct mline *
 533 recode_mline(ml, w, from, to)
 534 struct mline *ml;
 535 int w;
 536 int from, to;
 537 {
 538   static int maxlen;
 539   static int last;
 540   static struct mline rml[2], *rl;
 541   int i, c;
 542
 543   if (from == to || (from != UTF8 && to != UTF8) || w == 0)
 544     return ml;
 545   if (ml->font == null && encodings[from].deffont == 0)
 546     return ml;
 547   if (w > maxlen)
 548     {
 549       for (i = 0; i < 2; i++)
 550         {
 551           if (rml[i].image == 0)
 552             rml[i].image = malloc(w);
 553           else
 554             rml[i].image = realloc(rml[i].image, w);
 555           if (rml[i].font == 0)
 556             rml[i].font = malloc(w);
 557           else
 558             rml[i].font = realloc(rml[i].font, w);
 559           if (rml[i].image == 0 || rml[i].font == 0)
 560             {
 561               maxlen = 0;
 562               return ml;        /* sorry */
 563             }
 564         }
 565       maxlen = w;
 566     }
 567
 568   debug("recode_mline: from\n");
 569   for (i = 0; i < w; i++)
 570     debug1("%c", "0123456789abcdef"[(ml->image[i] >> 4) & 15]);
 571   debug("\n");
 572   for (i = 0; i < w; i++)
 573     debug1("%c", "0123456789abcdef"[(ml->image[i]     ) & 15]);
 574   debug("\n");
 575   for (i = 0; i < w; i++)
 576     debug1("%c", "0123456789abcdef"[(ml->font[i] >> 4) & 15]);
 577   debug("\n");
 578   for (i = 0; i < w; i++)
 579     debug1("%c", "0123456789abcdef"[(ml->font[i]     ) & 15]);
 580   debug("\n");
 581
 582   rl = rml + last;
 583   rl->attr = ml->attr;
 584 #ifdef COLOR
 585   rl->color = ml->color;
 586 # ifdef COLORS256
 587   rl->colorx = ml->colorx;
 588 # endif
 589 #endif
 590   for (i = 0; i < w; i++)
 591     {
 592       c = ml->image[i] | (ml->font[i] << 8);
 593       if (from != UTF8 && c < 256)
 594         c |= encodings[from].deffont << 8;
 595 #ifdef DW_CHARS
 596       if ((from != UTF8 && (c & 0x1f00) != 0 && (c & 0xe000) == 0) || (from == UTF8 && utf8_isdouble(c)))
 597         {
 598           if (i + 1 == w)
 599             c = '?';
 600           else
 601             {
 602               int c2;
 603               i++;
 604               c2 = ml->image[i] | (ml->font[i] << 8);
 605               c = recode_char_dw_to_encoding(c, &c2, to);
 606               rl->font[i - 1]  = c >> 8 & 255;
 607               rl->image[i - 1] = c      & 255;
 608               c = c2;
 609             }
 610         }
 611       else
 612 #endif
 613         c = recode_char_to_encoding(c, to);
 614       rl->image[i] = c & 255;
 615       rl->font[i] = c >> 8 & 255;
 616     }
 617   last ^= 1;
 618   debug("recode_mline: to\n");
 619   for (i = 0; i < w; i++)
 620     debug1("%c", "0123456789abcdef"[(rl->image[i] >> 4) & 15]);
 621   debug("\n");
 622   for (i = 0; i < w; i++)
 623     debug1("%c", "0123456789abcdef"[(rl->image[i]     ) & 15]);
 624   debug("\n");
 625   for (i = 0; i < w; i++)
 626     debug1("%c", "0123456789abcdef"[(rl->font[i] >> 4) & 15]);
 627   debug("\n");
 628   for (i = 0; i < w; i++)
 629     debug1("%c", "0123456789abcdef"[(rl->font[i]     ) & 15]);
 630   debug("\n");
 631   return rl;
 632 }
 633
 634 struct combchar {
 635   unsigned short c1;
 636   unsigned short c2;
 637   unsigned short next;
 638   unsigned short prev;
 639 };
 640 struct combchar **combchars;
 641
 642 void
 643 AddUtf8(c)
 644 int c;
 645 {
 646   ASSERT(D_encoding == UTF8);
 647   if (c >= 0xd800 && c < 0xe000 && combchars && combchars[c - 0xd800])
 648     {
 649       AddUtf8(combchars[c - 0xd800]->c1);
 650       c = combchars[c - 0xd800]->c2;
 651     }
 652   if (c >= 0x800)
 653     {
 654       AddChar((c & 0xf000) >> 12 | 0xe0);
 655       c = (c & 0x0fff) | 0x1000;
 656     }
 657   if (c >= 0x80)
 658     {
 659       AddChar((c & 0x1fc0) >> 6 ^ 0xc0);
 660       c = (c & 0x3f) | 0x80;
 661     }
 662   AddChar(c);
 663 }
 664
 665 int
 666 ToUtf8_comb(p, c)
 667 char *p;
 668 int c;
 669 {
 670   int l;
 671
 672   if (c >= 0xd800 && c < 0xe000 && combchars && combchars[c - 0xd800])
 673     {
 674       l = ToUtf8_comb(p, combchars[c - 0xd800]->c1);
 675       return l + ToUtf8(p ? p + l : 0, combchars[c - 0xd800]->c2);
 676     }
 677   return ToUtf8(p, c);
 678 }
 679
 680 int
 681 ToUtf8(p, c)
 682 char *p;
 683 int c;
 684 {
 685   int l = 1;
 686   if (c >= 0x800)
 687     {
 688       if (p)
 689         *p++ = (c & 0xf000) >> 12 | 0xe0;
 690       l++;
 691       c = (c & 0x0fff) | 0x1000;
 692     }
 693   if (c >= 0x80)
 694     {
 695       if (p)
 696         *p++ = (c & 0x1fc0) >> 6 ^ 0xc0;
 697       l++;
 698       c = (c & 0x3f) | 0x80;
 699     }
 700   if (p)
 701     *p++ = c;
 702   return l;
 703 }
 704
 705 /*
 706  * returns:
 707  * -1: need more bytes, sequence not finished
 708  * -2: corrupt sequence found, redo last char
 709  * >= 0: decoded character
 710  */
 711 int
 712 FromUtf8(c, utf8charp)
 713 int c, *utf8charp;
 714 {
 715   int utf8char = *utf8charp;
 716   if (utf8char)
 717     {
 718       if ((c & 0xc0) != 0x80)
 719         {
 720           *utf8charp = 0;
 721           return -2; /* corrupt sequence! */
 722         }
 723       else
 724         c = (c & 0x3f) | (utf8char << 6);
 725       if (!(utf8char & 0x40000000))
 726         {
 727           /* check for overlong sequences */
 728           if ((c & 0x820823e0) == 0x80000000)
 729             c = 0xfdffffff;
 730           else if ((c & 0x020821f0) == 0x02000000)
 731             c = 0xfff7ffff;
 732           else if ((c & 0x000820f8) == 0x00080000)
 733             c = 0xffffd000;
 734           else if ((c & 0x0000207c) == 0x00002000)
 735             c = 0xffffff70;
 736         }
 737     }
 738   else
 739     {
 740       /* new sequence */
 741       if (c >= 0xfe)
 742         c = UCS_REPL;
 743       else if (c >= 0xfc)
 744         c = (c & 0x01) | 0xbffffffc;    /* 5 bytes to follow */
 745       else if (c >= 0xf8)
 746         c = (c & 0x03) | 0xbfffff00;    /* 4 */
 747       else if (c >= 0xf0)
 748         c = (c & 0x07) | 0xbfffc000;    /* 3 */
 749       else if (c >= 0xe0)
 750         c = (c & 0x0f) | 0xbff00000;    /* 2 */
 751       else if (c >= 0xc2)
 752         c = (c & 0x1f) | 0xfc000000;    /* 1 */
 753       else if (c >= 0xc0)
 754         c = 0xfdffffff;         /* overlong */
 755       else if (c >= 0x80)
 756         c = UCS_REPL;
 757     }
 758   *utf8charp = utf8char = (c & 0x80000000) ? c : 0;
 759   if (utf8char)
 760     return -1;
 761   if (c & 0xffff0000)
 762     c = UCS_REPL;       /* sorry, only know 16bit Unicode */
 763   if (c >= 0xd800 && (c <= 0xdfff || c == 0xfffe || c == 0xffff))
 764     c = UCS_REPL;       /* illegal code */
 765   return c;
 766 }
 767
 768
 769 void
 770 WinSwitchEncoding(p, encoding)
 771 struct win *p;
 772 int encoding;
 773 {
 774   int i, j, c;
 775   struct mline *ml;
 776   struct display *d;
 777   struct canvas *cv;
 778   struct layer *oldflayer;
 779
 780   if ((p->w_encoding == UTF8) == (encoding == UTF8))
 781     {
 782       p->w_encoding = encoding;
 783       return;
 784     }
 785   oldflayer = flayer;
 786   for (d = displays; d; d = d->d_next)
 787     for (cv = d->d_cvlist; cv; cv = cv->c_next)
 788       if (p == Layer2Window(cv->c_layer))
 789         {
 790           flayer = cv->c_layer;
 791           while(flayer->l_next)
 792             {
 793               if (oldflayer == flayer)
 794                 oldflayer = flayer->l_next;
 795               ExitOverlayPage();
 796             }
 797         }
 798   flayer = oldflayer;
 799   for (j = 0; j < p->w_height + p->w_histheight; j++)
 800     {
 801 #ifdef COPY_PASTE
 802       ml = j < p->w_height ? &p->w_mlines[j] : &p->w_hlines[j - p->w_height];
 803 #else
 804       ml = &p->w_mlines[j];
 805 #endif
 806       if (ml->font == null && encodings[p->w_encoding].deffont == 0)
 807         continue;
 808       for (i = 0; i < p->w_width; i++)
 809         {
 810           c = ml->image[i] | (ml->font[i] << 8);
 811           if (p->w_encoding != UTF8 && c < 256)
 812             c |= encodings[p->w_encoding].deffont << 8;
 813           if (c < 256)
 814             continue;
 815           if (ml->font == null)
 816             {
 817               if ((ml->font = (unsigned char *)malloc(p->w_width + 1)) == 0)
 818                 {
 819                   ml->font = null;
 820                   break;
 821                 }
 822               bzero(ml->font, p->w_width + 1);
 823             }
 824 #ifdef DW_CHARS
 825           if ((p->w_encoding != UTF8 && (c & 0x1f00) != 0 && (c & 0xe000) == 0) || (p->w_encoding == UTF8 && utf8_isdouble(c)))
 826             {
 827               if (i + 1 == p->w_width)
 828                 c = '?';
 829               else
 830                 {
 831                   int c2;
 832                   i++;
 833                   c2 = ml->image[i] | (ml->font[i] << 8);
 834                   c = recode_char_dw_to_encoding(c, &c2, encoding);
 835                   ml->font[i - 1]  = c >> 8 & 255;
 836                   ml->image[i - 1] = c      & 255;
 837                   c = c2;
 838                 }
 839             }
 840           else
 841 #endif
 842             c = recode_char_to_encoding(c, encoding);
 843           ml->image[i] = c & 255;
 844           ml->font[i] = c >> 8 & 255;
 845         }
 846     }
 847   p->w_encoding = encoding;
 848   return;
 849 }
 850
 851 #ifdef DW_CHARS
 852 struct interval {
 853   int first;
 854   int last;
 855 };
 856
 857 /* auxiliary function for binary search in interval table */
 858 static int bisearch(int ucs, const struct interval *table, int max) {
 859   int min = 0;
 860   int mid;
 861
 862   if (ucs < table[0].first || ucs > table[max].last)
 863     return 0;
 864   while (max >= min) {
 865     mid = (min + max) / 2;
 866     if (ucs > table[mid].last)
 867       min = mid + 1;
 868     else if (ucs < table[mid].first)
 869       max = mid - 1;
 870     else
 871       return 1;
 872   }
 873
 874   return 0;
 875 }
 876
 877 int
 878 utf8_isdouble(c)
 879 int c;
 880 {
 881   /* sorted list of non-overlapping intervals of East Asian Ambiguous
 882    * characters, generated by "uniset +WIDTH-A -cat=Me -cat=Mn -cat=Cf c" */
 883   static const struct interval ambiguous[] = {
 884     { 0x00A1, 0x00A1 }, { 0x00A4, 0x00A4 }, { 0x00A7, 0x00A8 },
 885     { 0x00AA, 0x00AA }, { 0x00AE, 0x00AE }, { 0x00B0, 0x00B4 },
 886     { 0x00B6, 0x00BA }, { 0x00BC, 0x00BF }, { 0x00C6, 0x00C6 },
 887     { 0x00D0, 0x00D0 }, { 0x00D7, 0x00D8 }, { 0x00DE, 0x00E1 },
 888     { 0x00E6, 0x00E6 }, { 0x00E8, 0x00EA }, { 0x00EC, 0x00ED },
 889     { 0x00F0, 0x00F0 }, { 0x00F2, 0x00F3 }, { 0x00F7, 0x00FA },
 890     { 0x00FC, 0x00FC }, { 0x00FE, 0x00FE }, { 0x0101, 0x0101 },
 891     { 0x0111, 0x0111 }, { 0x0113, 0x0113 }, { 0x011B, 0x011B },
 892     { 0x0126, 0x0127 }, { 0x012B, 0x012B }, { 0x0131, 0x0133 },
 893     { 0x0138, 0x0138 }, { 0x013F, 0x0142 }, { 0x0144, 0x0144 },
 894     { 0x0148, 0x014B }, { 0x014D, 0x014D }, { 0x0152, 0x0153 },
 895     { 0x0166, 0x0167 }, { 0x016B, 0x016B }, { 0x01CE, 0x01CE },
 896     { 0x01D0, 0x01D0 }, { 0x01D2, 0x01D2 }, { 0x01D4, 0x01D4 },
 897     { 0x01D6, 0x01D6 }, { 0x01D8, 0x01D8 }, { 0x01DA, 0x01DA },
 898     { 0x01DC, 0x01DC }, { 0x0251, 0x0251 }, { 0x0261, 0x0261 },
 899     { 0x02C4, 0x02C4 }, { 0x02C7, 0x02C7 }, { 0x02C9, 0x02CB },
 900     { 0x02CD, 0x02CD }, { 0x02D0, 0x02D0 }, { 0x02D8, 0x02DB },
 901     { 0x02DD, 0x02DD }, { 0x02DF, 0x02DF }, { 0x0391, 0x03A1 },
 902     { 0x03A3, 0x03A9 }, { 0x03B1, 0x03C1 }, { 0x03C3, 0x03C9 },
 903     { 0x0401, 0x0401 }, { 0x0410, 0x044F }, { 0x0451, 0x0451 },
 904     { 0x2010, 0x2010 }, { 0x2013, 0x2016 }, { 0x2018, 0x2019 },
 905     { 0x201C, 0x201D }, { 0x2020, 0x2022 }, { 0x2024, 0x2027 },
 906     { 0x2030, 0x2030 }, { 0x2032, 0x2033 }, { 0x2035, 0x2035 },
 907     { 0x203B, 0x203B }, { 0x203E, 0x203E }, { 0x2074, 0x2074 },
 908     { 0x207F, 0x207F }, { 0x2081, 0x2084 }, { 0x20AC, 0x20AC },
 909     { 0x2103, 0x2103 }, { 0x2105, 0x2105 }, { 0x2109, 0x2109 },
 910     { 0x2113, 0x2113 }, { 0x2116, 0x2116 }, { 0x2121, 0x2122 },
 911     { 0x2126, 0x2126 }, { 0x212B, 0x212B }, { 0x2153, 0x2154 },
 912     { 0x215B, 0x215E }, { 0x2160, 0x216B }, { 0x2170, 0x2179 },
 913     { 0x2190, 0x2199 }, { 0x21B8, 0x21B9 }, { 0x21D2, 0x21D2 },
 914     { 0x21D4, 0x21D4 }, { 0x21E7, 0x21E7 }, { 0x2200, 0x2200 },
 915     { 0x2202, 0x2203 }, { 0x2207, 0x2208 }, { 0x220B, 0x220B },
 916     { 0x220F, 0x220F }, { 0x2211, 0x2211 }, { 0x2215, 0x2215 },
 917     { 0x221A, 0x221A }, { 0x221D, 0x2220 }, { 0x2223, 0x2223 },
 918     { 0x2225, 0x2225 }, { 0x2227, 0x222C }, { 0x222E, 0x222E },
 919     { 0x2234, 0x2237 }, { 0x223C, 0x223D }, { 0x2248, 0x2248 },
 920     { 0x224C, 0x224C }, { 0x2252, 0x2252 }, { 0x2260, 0x2261 },
 921     { 0x2264, 0x2267 }, { 0x226A, 0x226B }, { 0x226E, 0x226F },
 922     { 0x2282, 0x2283 }, { 0x2286, 0x2287 }, { 0x2295, 0x2295 },
 923     { 0x2299, 0x2299 }, { 0x22A5, 0x22A5 }, { 0x22BF, 0x22BF },
 924     { 0x2312, 0x2312 }, { 0x2460, 0x24E9 }, { 0x24EB, 0x254B },
 925     { 0x2550, 0x2573 }, { 0x2580, 0x258F }, { 0x2592, 0x2595 },
 926     { 0x25A0, 0x25A1 }, { 0x25A3, 0x25A9 }, { 0x25B2, 0x25B3 },
 927     { 0x25B6, 0x25B7 }, { 0x25BC, 0x25BD }, { 0x25C0, 0x25C1 },
 928     { 0x25C6, 0x25C8 }, { 0x25CB, 0x25CB }, { 0x25CE, 0x25D1 },
 929     { 0x25E2, 0x25E5 }, { 0x25EF, 0x25EF }, { 0x2605, 0x2606 },
 930     { 0x2609, 0x2609 }, { 0x260E, 0x260F }, { 0x2614, 0x2615 },
 931     { 0x261C, 0x261C }, { 0x261E, 0x261E }, { 0x2640, 0x2640 },
 932     { 0x2642, 0x2642 }, { 0x2660, 0x2661 }, { 0x2663, 0x2665 },
 933     { 0x2667, 0x266A }, { 0x266C, 0x266D }, { 0x266F, 0x266F },
 934     { 0x273D, 0x273D }, { 0x2776, 0x277F }, { 0xE000, 0xF8FF },
 935     { 0xFFFD, 0xFFFD }, { 0xF0000, 0xFFFFD }, { 0x100000, 0x10FFFD }
 936   };
 937
 938   return ((c >= 0x1100 &&
 939            (c <= 0x115f ||                    /* Hangul Jamo init. consonants */
 940             c == 0x2329 || c == 0x232a ||
 941             (c >= 0x2e80 && c <= 0xa4cf &&
 942              c != 0x303f) ||                  /* CJK ... Yi */
 943             (c >= 0xac00 && c <= 0xd7a3) || /* Hangul Syllables */
 944             (c >= 0xf900 && c <= 0xfaff) || /* CJK Compatibility Ideographs */
 945             (c >= 0xfe30 && c <= 0xfe6f) || /* CJK Compatibility Forms */
 946             (c >= 0xff00 && c <= 0xff60) || /* Fullwidth Forms */
 947             (c >= 0xffe0 && c <= 0xffe6) ||
 948             (c >= 0x20000 && c <= 0x2fffd) ||
 949             (c >= 0x30000 && c <= 0x3fffd))) ||
 950           (cjkwidth &&
 951            bisearch(c, ambiguous,
 952                     sizeof(ambiguous) / sizeof(struct interval) - 1)));
 953 }
 954 #endif
 955
 956 int
 957 utf8_iscomb(c)
 958 int c;
 959 {
 960   /* taken from Markus Kuhn's wcwidth */
 961   static const struct interval combining[] = {
 962     { 0x0300, 0x036F }, { 0x0483, 0x0486 }, { 0x0488, 0x0489 },
 963     { 0x0591, 0x05BD }, { 0x05BF, 0x05BF }, { 0x05C1, 0x05C2 },
 964     { 0x05C4, 0x05C5 }, { 0x05C7, 0x05C7 }, { 0x0600, 0x0603 },
 965     { 0x0610, 0x0615 }, { 0x064B, 0x065E }, { 0x0670, 0x0670 },
 966     { 0x06D6, 0x06E4 }, { 0x06E7, 0x06E8 }, { 0x06EA, 0x06ED },
 967     { 0x070F, 0x070F }, { 0x0711, 0x0711 }, { 0x0730, 0x074A },
 968     { 0x07A6, 0x07B0 }, { 0x07EB, 0x07F3 }, { 0x0901, 0x0902 },
 969     { 0x093C, 0x093C }, { 0x0941, 0x0948 }, { 0x094D, 0x094D },
 970     { 0x0951, 0x0954 }, { 0x0962, 0x0963 }, { 0x0981, 0x0981 },
 971     { 0x09BC, 0x09BC }, { 0x09C1, 0x09C4 }, { 0x09CD, 0x09CD },
 972     { 0x09E2, 0x09E3 }, { 0x0A01, 0x0A02 }, { 0x0A3C, 0x0A3C },
 973     { 0x0A41, 0x0A42 }, { 0x0A47, 0x0A48 }, { 0x0A4B, 0x0A4D },
 974     { 0x0A70, 0x0A71 }, { 0x0A81, 0x0A82 }, { 0x0ABC, 0x0ABC },
 975     { 0x0AC1, 0x0AC5 }, { 0x0AC7, 0x0AC8 }, { 0x0ACD, 0x0ACD },
 976     { 0x0AE2, 0x0AE3 }, { 0x0B01, 0x0B01 }, { 0x0B3C, 0x0B3C },
 977     { 0x0B3F, 0x0B3F }, { 0x0B41, 0x0B43 }, { 0x0B4D, 0x0B4D },
 978     { 0x0B56, 0x0B56 }, { 0x0B82, 0x0B82 }, { 0x0BC0, 0x0BC0 },
 979     { 0x0BCD, 0x0BCD }, { 0x0C3E, 0x0C40 }, { 0x0C46, 0x0C48 },
 980     { 0x0C4A, 0x0C4D }, { 0x0C55, 0x0C56 }, { 0x0CBC, 0x0CBC },
 981     { 0x0CBF, 0x0CBF }, { 0x0CC6, 0x0CC6 }, { 0x0CCC, 0x0CCD },
 982     { 0x0CE2, 0x0CE3 }, { 0x0D41, 0x0D43 }, { 0x0D4D, 0x0D4D },
 983     { 0x0DCA, 0x0DCA }, { 0x0DD2, 0x0DD4 }, { 0x0DD6, 0x0DD6 },
 984     { 0x0E31, 0x0E31 }, { 0x0E34, 0x0E3A }, { 0x0E47, 0x0E4E },
 985     { 0x0EB1, 0x0EB1 }, { 0x0EB4, 0x0EB9 }, { 0x0EBB, 0x0EBC },
 986     { 0x0EC8, 0x0ECD }, { 0x0F18, 0x0F19 }, { 0x0F35, 0x0F35 },
 987     { 0x0F37, 0x0F37 }, { 0x0F39, 0x0F39 }, { 0x0F71, 0x0F7E },
 988     { 0x0F80, 0x0F84 }, { 0x0F86, 0x0F87 }, { 0x0F90, 0x0F97 },
 989     { 0x0F99, 0x0FBC }, { 0x0FC6, 0x0FC6 }, { 0x102D, 0x1030 },
 990     { 0x1032, 0x1032 }, { 0x1036, 0x1037 }, { 0x1039, 0x1039 },
 991     { 0x1058, 0x1059 }, { 0x1160, 0x11FF }, { 0x135F, 0x135F },
 992     { 0x1712, 0x1714 }, { 0x1732, 0x1734 }, { 0x1752, 0x1753 },
 993     { 0x1772, 0x1773 }, { 0x17B4, 0x17B5 }, { 0x17B7, 0x17BD },
 994     { 0x17C6, 0x17C6 }, { 0x17C9, 0x17D3 }, { 0x17DD, 0x17DD },
 995     { 0x180B, 0x180D }, { 0x18A9, 0x18A9 }, { 0x1920, 0x1922 },
 996     { 0x1927, 0x1928 }, { 0x1932, 0x1932 }, { 0x1939, 0x193B },
 997     { 0x1A17, 0x1A18 }, { 0x1B00, 0x1B03 }, { 0x1B34, 0x1B34 },
 998     { 0x1B36, 0x1B3A }, { 0x1B3C, 0x1B3C }, { 0x1B42, 0x1B42 },
 999     { 0x1B6B, 0x1B73 }, { 0x1DC0, 0x1DCA }, { 0x1DFE, 0x1DFF },
1000     { 0x200B, 0x200F }, { 0x202A, 0x202E }, { 0x2060, 0x2063 },
1001     { 0x206A, 0x206F }, { 0x20D0, 0x20EF }, { 0x302A, 0x302F },
1002     { 0x3099, 0x309A }, { 0xA806, 0xA806 }, { 0xA80B, 0xA80B },
1003     { 0xA825, 0xA826 }, { 0xFB1E, 0xFB1E }, { 0xFE00, 0xFE0F },
1004     { 0xFE20, 0xFE23 }, { 0xFEFF, 0xFEFF }, { 0xFFF9, 0xFFFB },
1005     { 0x10A01, 0x10A03 }, { 0x10A05, 0x10A06 }, { 0x10A0C, 0x10A0F },
1006     { 0x10A38, 0x10A3A }, { 0x10A3F, 0x10A3F }, { 0x1D167, 0x1D169 },
1007     { 0x1D173, 0x1D182 }, { 0x1D185, 0x1D18B }, { 0x1D1AA, 0x1D1AD },
1008     { 0x1D242, 0x1D244 }, { 0xE0001, 0xE0001 }, { 0xE0020, 0xE007F },
1009     { 0xE0100, 0xE01EF }
1010   };
1011
1012   return bisearch(c, combining, sizeof(combining) / sizeof(struct interval) - 1);
1013 }
1014
1015 static void
1016 comb_tofront(root, i)
1017 int root, i;
1018 {
1019   for (;;)
1020     {
1021       debug1("bring to front: %x\n", i);
1022       combchars[combchars[i]->prev]->next = combchars[i]->next;
1023       combchars[combchars[i]->next]->prev = combchars[i]->prev;
1024       combchars[i]->next = combchars[root]->next;
1025       combchars[i]->prev = root;
1026       combchars[combchars[root]->next]->prev = i;
1027       combchars[root]->next = i;
1028       i = combchars[i]->c1;
1029       if (i < 0xd800 || i >= 0xe000)
1030         return;
1031       i -= 0xd800;
1032     }
1033 }
1034
1035 void
1036 utf8_handle_comb(c, mc)
1037 int c;
1038 struct mchar *mc;
1039 {
1040   int root, i, c1;
1041   int isdouble;
1042
1043   c1 = mc->image | (mc->font << 8);
1044   isdouble = c1 >= 0x1100 && utf8_isdouble(c1);
1045   if (!combchars)
1046     {
1047       combchars = (struct combchar **)malloc(sizeof(struct combchar *) * 0x802);
1048       if (!combchars)
1049         return;
1050       bzero((char *)combchars, sizeof(struct combchar *) * 0x802);
1051       combchars[0x800] = (struct combchar *)malloc(sizeof(struct combchar));
1052       combchars[0x801] = (struct combchar *)malloc(sizeof(struct combchar));
1053       if (!combchars[0x800] || !combchars[0x801])
1054         {
1055           if (combchars[0x800])
1056             free(combchars[0x800]);
1057           if (combchars[0x801])
1058             free(combchars[0x801]);
1059           free(combchars);
1060           return;
1061         }
1062       combchars[0x800]->c1 = 0x000;
1063       combchars[0x800]->c2 = 0x700;
1064       combchars[0x800]->next = 0x800;
1065       combchars[0x800]->prev = 0x800;
1066       combchars[0x801]->c1 = 0x700;
1067       combchars[0x801]->c2 = 0x800;
1068       combchars[0x801]->next = 0x801;
1069       combchars[0x801]->prev = 0x801;
1070     }
1071   root = isdouble ? 0x801 : 0x800;
1072   for (i = combchars[root]->c1; i < combchars[root]->c2; i++)
1073     {
1074       if (!combchars[i])
1075         break;
1076       if (combchars[i]->c1 == c1 && combchars[i]->c2 == c)
1077         break;
1078     }
1079   if (i == combchars[root]->c2)
1080     {
1081       /* full, recycle old entry */
1082       if (c1 >= 0xd800 && c1 < 0xe000)
1083         comb_tofront(root, c1 - 0xd800);
1084       i = combchars[root]->prev;
1085       if (c1 == i + 0xd800)
1086         {
1087           /* completely full, can't recycle */
1088           debug("utf8_handle_comp: completely full!\n");
1089           mc->image = '?';
1090           mc->font  = 0;
1091           return;
1092         }
1093       /* FIXME: delete old char from all buffers */
1094     }
1095   else if (!combchars[i])
1096     {
1097       combchars[i] = (struct combchar *)malloc(sizeof(struct combchar));
1098       if (!combchars[i])
1099         return;
1100       combchars[i]->prev = i;
1101       combchars[i]->next = i;
1102     }
1103   combchars[i]->c1 = c1;
1104   combchars[i]->c2 = c;
1105   mc->image = i & 0xff;
1106   mc->font  = (i >> 8) + 0xd8;
1107   debug3("combinig char %x %x -> %x\n", c1, c, i + 0xd800);
1108   comb_tofront(root, i);
1109 }
1110
1111 #else /* !UTF8 */
1112
1113 void
1114 WinSwitchEncoding(p, encoding)
1115 struct win *p;
1116 int encoding;
1117 {
1118   p->w_encoding = encoding;
1119   return;
1120 }
1121
1122 #endif /* UTF8 */
1123
1124 static int
1125 encmatch(s1, s2)
1126 char *s1;
1127 char *s2;
1128 {
1129   int c1, c2;
1130   do
1131     {
1132       c1 = (unsigned char)*s1;
1133       if (c1 >= 'A' && c1 <= 'Z')
1134         c1 += 'a' - 'A';
1135       if (!(c1 >= 'a' && c1 <= 'z') && !(c1 >= '0' && c1 <= '9'))
1136         {
1137           s1++;
1138           continue;
1139         }
1140       c2 = (unsigned char)*s2;
1141       if (c2 >= 'A' && c2 <= 'Z')
1142         c2 += 'a' - 'A';
1143       if (!(c2 >= 'a' && c2 <= 'z') && !(c2 >= '0' && c2 <= '9'))
1144         {
1145           s2++;
1146           continue;
1147         }
1148       if (c1 != c2)
1149         return 0;
1150       s1++;
1151       s2++;
1152     }
1153   while(c1);
1154   return 1;
1155 }
1156
1157 int
1158 FindEncoding(name)
1159 char *name;
1160 {
1161   int encoding;
1162
1163   debug1("FindEncoding %s\n", name);
1164   if (name == 0 || *name == 0)
1165     return 0;
1166   if (encmatch(name, "euc"))
1167     name = "eucJP";
1168   if (encmatch(name, "off") || encmatch(name, "iso8859-1"))
1169     return 0;
1170 #ifndef UTF8
1171   if (encmatch(name, "UTF-8"))
1172     return -1;
1173 #endif
1174   for (encoding = 0; encoding < (int)(sizeof(encodings)/sizeof(*encodings)); encoding++)
1175     if (encmatch(name, encodings[encoding].name))
1176       {
1177 #ifdef UTF8
1178         LoadFontTranslationsForEncoding(encoding);
1179 #endif
1180         return encoding;
1181       }
1182   return -1;
1183 }
1184
1185 char *
1186 EncodingName(encoding)
1187 int encoding;
1188 {
1189   if (encoding >= (int)(sizeof(encodings)/sizeof(*encodings)))
1190     return 0;
1191   return encodings[encoding].name;
1192 }
1193
1194 int
1195 EncodingDefFont(encoding)
1196 int encoding;
1197 {
1198   return encodings[encoding].deffont;
1199 }
1200
1201 void
1202 ResetEncoding(p)
1203 struct win *p;
1204 {
1205   char *c;
1206   int encoding = p->w_encoding;
1207
1208   c = encodings[encoding].charsets;
1209   if (c)
1210     SetCharsets(p, c);
1211 #ifdef UTF8
1212   LoadFontTranslationsForEncoding(encoding);
1213 #endif
1214   if (encodings[encoding].usegr)
1215     {
1216       p->w_gr = 2;
1217       p->w_FontE = encodings[encoding].charsets[1];
1218     }
1219   else
1220     p->w_FontE = 0;
1221   if (encodings[encoding].noc1)
1222     p->w_c1 = 0;
1223 }
1224
1225 int
1226 DecodeChar(c, encoding, statep)
1227 int c;
1228 int encoding;
1229 int *statep;
1230 {
1231   int t;
1232
1233   debug2("Decoding char %02x for encoding %d\n", c, encoding);
1234 #ifdef UTF8
1235   if (encoding == UTF8)
1236     return FromUtf8(c, statep);
1237 #endif
1238   if (encoding == SJIS)
1239     {
1240       if (!*statep)
1241         {
1242           if ((0x81 <= c && c <= 0x9f) || (0xe0 <= c && c <= 0xef))
1243             {
1244               *statep = c;
1245               return -1;
1246             }
1247           if (c < 0x80)
1248             return c;
1249           return c | (KANA << 16);
1250         }
1251       t = c;
1252       c = *statep;
1253       *statep = 0;
1254       if (0x40 <= t && t <= 0xfc && t != 0x7f)
1255         {
1256           if (c <= 0x9f)
1257             c = (c - 0x81) * 2 + 0x21;
1258           else
1259             c = (c - 0xc1) * 2 + 0x21;
1260           if (t <= 0x7e)
1261             t -= 0x1f;
1262           else if (t <= 0x9e)
1263             t -= 0x20;
1264           else
1265              t -= 0x7e, c++;
1266           return (c << 8) | t | (KANJI << 16);
1267         }
1268       return t;
1269     }
1270   if (encoding == EUC_JP || encoding == EUC_KR || encoding == EUC_CN)
1271     {
1272       if (!*statep)
1273         {
1274           if (c & 0x80)
1275             {
1276               *statep = c;
1277               return -1;
1278             }
1279           return c;
1280         }
1281       t = c;
1282       c = *statep;
1283       *statep = 0;
1284       if (encoding == EUC_JP)
1285         {
1286           if (c == 0x8e)
1287             return t | (KANA << 16);
1288           if (c == 0x8f)
1289             {
1290               *statep = t | (KANJI0212 << 8);
1291               return -1;
1292             }
1293         }
1294       c &= 0xff7f;
1295       t &= 0x7f;
1296       c = c << 8 | t;
1297       if (encoding == EUC_KR)
1298         return c | (3 << 16);
1299       if (encoding == EUC_CN)
1300         return c | (1 << 16);
1301       if (c & (KANJI0212 << 16))
1302         return c;
1303       else
1304         return c | (KANJI << 16);
1305     }
1306   if (encoding == BIG5 || encoding == GBK)
1307     {
1308       if (!*statep)
1309         {
1310           if (c & 0x80)
1311             {
1312               if (encoding == GBK && c == 0x80)
1313                 return 0xa4 | (('b'|0x80) << 16);
1314               *statep = c;
1315               return -1;
1316             }
1317           return c;
1318         }
1319       t = c;
1320       c = *statep;
1321       *statep = 0;
1322       c &= 0x7f;
1323       return c << 8 | t | (encoding == BIG5  ? 030 << 16 : 031 << 16);
1324     }
1325   return c | (encodings[encoding].deffont << 16);
1326 }
1327
1328 int
1329 EncodeChar(bp, c, encoding, fontp)
1330 char *bp;
1331 int c;
1332 int encoding;
1333 int *fontp;
1334 {
1335   int t, f, l;
1336
1337   debug2("Encoding char %02x for encoding %d\n", c, encoding);
1338   if (c == -1 && fontp)
1339     {
1340       if (*fontp == 0)
1341         return 0;
1342       if (bp)
1343         {
1344           *bp++ = 033;
1345           *bp++ = '(';
1346           *bp++ = 'B';
1347         }
1348       return 3;
1349     }
1350   f = c >> 16;
1351
1352 #ifdef UTF8
1353   if (encoding == UTF8)
1354     {
1355       if (f)
1356         {
1357 # ifdef DW_CHARS
1358           if (is_dw_font(f))
1359             {
1360               int c2 = c & 0xff;
1361               c = (c >> 8 & 0xff) | (f << 8);
1362               c = recode_char_dw_to_encoding(c, &c2, encoding);
1363             }
1364           else
1365 # endif
1366             {
1367               c = (c & 0xff) | (f << 8);
1368               c = recode_char_to_encoding(c, encoding);
1369             }
1370         }
1371       return ToUtf8(bp, c);
1372     }
1373   if ((c & 0xff00) && f == 0)   /* is_utf8? */
1374     {
1375 # ifdef DW_CHARS
1376       if (utf8_isdouble(c))
1377         {
1378           int c2 = 0xffff;
1379           c = recode_char_dw_to_encoding(c, &c2, encoding);
1380           c = (c << 8) | (c2 & 0xff);
1381         }
1382       else
1383 # endif
1384         {
1385           c = recode_char_to_encoding(c, encoding);
1386           c = ((c & 0xff00) << 8) | (c & 0xff);
1387         }
1388       debug1("Encode: char mapped from utf8 to %x\n", c);
1389       f = c >> 16;
1390     }
1391 #endif
1392   if (f & 0x80)         /* map special 96-fonts to latin1 */
1393     f = 0;
1394
1395   if (encoding == SJIS)
1396     {
1397       if (f == KANA)
1398         c = (c & 0xff) | 0x80;
1399       else if (f == KANJI)
1400         {
1401           if (!bp)
1402             return 2;
1403           t = c & 0xff;
1404           c = (c >> 8) & 0xff;
1405           t += (c & 1) ? ((t <= 0x5f) ? 0x1f : 0x20) : 0x7e;
1406           c = (c - 0x21) / 2 + ((c < 0x5f) ? 0x81 : 0xc1);
1407           *bp++ = c;
1408           *bp++ = t;
1409           return 2;
1410         }
1411     }
1412   if (encoding == EUC)
1413     {
1414       if (f == KANA)
1415         {
1416           if (bp)
1417             {
1418               *bp++ = 0x8e;
1419               *bp++ = c;
1420             }
1421           return 2;
1422         }
1423       if (f == KANJI)
1424         {
1425           if (bp)
1426             {
1427               *bp++ = (c >> 8) | 0x80;
1428               *bp++ = c | 0x80;
1429             }
1430           return 2;
1431         }
1432       if (f == KANJI0212)
1433         {
1434           if (bp)
1435             {
1436               *bp++ = 0x8f;
1437               *bp++ = c >> 8;
1438               *bp++ = c;
1439             }
1440           return 3;
1441         }
1442     }
1443   if ((encoding == EUC_KR && f == 3) || (encoding == EUC_CN && f == 1))
1444     {
1445       if (bp)
1446         {
1447           *bp++ = (c >> 8) | 0x80;
1448           *bp++ = c | 0x80;
1449         }
1450       return 2;
1451     }
1452   if ((encoding == BIG5 && f == 030) || (encoding == GBK && f == 031))
1453     {
1454       if (bp)
1455         {
1456           *bp++ = (c >> 8) | 0x80;
1457           *bp++ = c;
1458         }
1459       return 2;
1460     }
1461   if (encoding == GBK && f == 0 && c == 0xa4)
1462     c = 0x80;
1463
1464   l = 0;
1465   if (fontp && f != *fontp)
1466     {
1467       *fontp = f;
1468       if (f && f < ' ')
1469         {
1470           if (bp)
1471            {
1472              *bp++ = 033;
1473              *bp++ = '$';
1474              if (f > 2)
1475                *bp++ = '(';
1476              *bp++ = '@' + f;
1477            }
1478           l += f > 2 ? 4 : 3;
1479         }
1480       else if (f < 128)
1481         {
1482           if (f == 0)
1483             f = 'B';
1484           if (bp)
1485             {
1486               *bp++ = 033;
1487               *bp++ = '(';
1488               *bp++ = f;
1489             }
1490           l += 3;
1491         }
1492     }
1493   if (c & 0xff00)
1494     {
1495       if (bp)
1496         *bp++ = c >> 8;
1497       l++;
1498     }
1499   if (bp)
1500     *bp++ = c;
1501   return l + 1;
1502 }
1503
1504 int
1505 CanEncodeFont(encoding, f)
1506 int encoding, f;
1507 {
1508   switch(encoding)
1509     {
1510 #ifdef UTF8
1511     case UTF8:
1512       return 1;
1513 #endif
1514     case SJIS:
1515       return f == KANJI || f == KANA;
1516     case EUC:
1517       return f == KANJI || f == KANA || f == KANJI0212;
1518     case EUC_KR:
1519       return f == 3;
1520     case EUC_CN:
1521       return f == 1;
1522     case BIG5:
1523       return f == 030;
1524     case GBK:
1525       return f == 031;
1526     default:
1527       break;
1528     }
1529   return 0;
1530 }
1531
1532 #ifdef DW_CHARS
1533 int
1534 PrepareEncodedChar(c)
1535 int c;
1536 {
1537   int encoding;
1538   int t = 0;
1539   int f;
1540
1541   encoding = D_encoding;
1542   f = D_rend.font;
1543   t = D_mbcs;
1544   if (encoding == SJIS)
1545     {
1546       if (f == KANA)
1547         return c | 0x80;
1548       else if (f == KANJI)
1549         {
1550           t += (c & 1) ? ((t <= 0x5f) ? 0x1f : 0x20) : 0x7e;
1551           c = (c - 0x21) / 2 + ((c < 0x5f) ? 0x81 : 0xc1);
1552           D_mbcs = t;
1553         }
1554       return c;
1555     }
1556   if (encoding == EUC)
1557     {
1558       if (f == KANA)
1559         {
1560           AddChar(0x8e);
1561           return c | 0x80;
1562         }
1563       if (f == KANJI)
1564         {
1565           D_mbcs = t | 0x80;
1566           return c | 0x80;
1567         }
1568       if (f == KANJI0212)
1569         {
1570           AddChar(0x8f);
1571           D_mbcs = t | 0x80;
1572           return c | 0x80;
1573         }
1574     }
1575   if ((encoding == EUC_KR && f == 3) || (encoding == EUC_CN && f == 1))
1576     {
1577       D_mbcs = t | 0x80;
1578       return c | 0x80;
1579     }
1580   if ((encoding == BIG5 && f == 030) || (encoding == GBK && f == 031))
1581     return c | 0x80;
1582   return c;
1583 }
1584 #endif
1585
1586 int
1587 RecodeBuf(fbuf, flen, fenc, tenc, tbuf)
1588 unsigned char *fbuf;
1589 int flen;
1590 int fenc, tenc;
1591 unsigned char *tbuf;
1592 {
1593   int c, i, j;
1594   int decstate = 0, font = 0;
1595
1596   for (i = j = 0; i < flen; i++)
1597     {
1598       c = fbuf[i];
1599       c = DecodeChar(c, fenc, &decstate);
1600       if (c == -2)
1601         i--;
1602       if (c < 0)
1603         continue;
1604       j += EncodeChar(tbuf ? (char *)tbuf + j : 0, c, tenc, &font);
1605     }
1606   j += EncodeChar(tbuf ? (char *)tbuf + j : 0, -1, tenc, &font);
1607   return j;
1608 }
1609
1610 #ifdef UTF8
1611 int
1612 ContainsSpecialDeffont(ml, xs, xe, encoding)
1613 struct mline *ml;
1614 int xs, xe;
1615 int encoding;
1616 {
1617   unsigned char *f, *i;
1618   int c, x, dx;
1619
1620   if (encoding == UTF8 || encodings[encoding].deffont == 0)
1621     return 0;
1622   i = ml->image + xs;
1623   f = ml->font + xs;
1624   dx = xe - xs + 1;
1625   while (dx-- > 0)
1626     {
1627       if (*f++)
1628         continue;
1629       c = *i++;
1630       x = recode_char_to_encoding(c | (encodings[encoding].deffont << 8), UTF8);
1631       if (c != x)
1632         {
1633           debug2("ContainsSpecialDeffont: yes %02x != %02x\n", c, x);
1634           return 1;
1635         }
1636     }
1637   debug("ContainsSpecialDeffont: no\n");
1638   return 0;
1639 }
1640
1641
1642 int
1643 LoadFontTranslation(font, file)
1644 int font;
1645 char *file;
1646 {
1647   char buf[1024], *myfile;
1648   FILE *f;
1649   int i;
1650   int fo;
1651   int x, u, c, ok;
1652   unsigned short (*p)[2], (*tab)[2];
1653
1654   myfile = file;
1655   if (myfile == 0)
1656     {
1657       if (font == 0 || screenencodings == 0)
1658         return -1;
1659       if (strlen(screenencodings) > sizeof(buf) - 10)
1660         return -1;
1661       sprintf(buf, "%s/%02x", screenencodings, font & 0xff);
1662       myfile = buf;
1663     }
1664   debug1("LoadFontTranslation: trying %s\n", myfile);
1665   if ((f = secfopen(myfile, "r")) == 0)
1666     return -1;
1667   i = ok = 0;
1668   for (;;)
1669     {
1670       for(; i < 12; i++)
1671         if (getc(f) != "ScreenI2UTF8"[i])
1672           break;
1673       if (getc(f) != 0)         /* format */
1674         break;
1675       fo = getc(f);             /* id */
1676       if (fo == EOF)
1677         break;
1678       if (font != -1 && font != fo)
1679         break;
1680       i = getc(f);
1681       x = getc(f);
1682       if (x == EOF)
1683         break;
1684       i = i << 8 | x;
1685       getc(f);
1686       while ((x = getc(f)) && x != EOF)
1687         getc(f);        /* skip font name (padded to 2 bytes) */
1688       if ((p = malloc(sizeof(*p) * (i + 1))) == 0)
1689         break;
1690       tab = p;
1691       while(i > 0)
1692         {
1693           x = getc(f);
1694           x = x << 8 | getc(f);
1695           u = getc(f);
1696           c = getc(f);
1697           u = u << 8 | c;
1698           if (c == EOF)
1699             break;
1700           (*p)[0] = x;
1701           (*p)[1] = u;
1702           p++;
1703           i--;
1704         }
1705       (*p)[0] = 0;
1706       (*p)[1] = 0;
1707       if (i || (tab[0][0] & 0x8000))
1708         {
1709           free(tab);
1710           break;
1711         }
1712       if (recodetabs[fo].tab && (recodetabs[fo].flags & RECODETAB_ALLOCED) != 0)
1713         free(recodetabs[fo].tab);
1714       recodetabs[fo].tab = tab;
1715       recodetabs[fo].flags = RECODETAB_ALLOCED;
1716       debug1("Successful load of recodetab %02x\n", fo);
1717       c = getc(f);
1718       if (c == EOF)
1719         {
1720           ok = 1;
1721           break;
1722         }
1723       if (c != 'S')
1724         break;
1725       i = 1;
1726     }
1727   fclose(f);
1728   if (font != -1 && file == 0 && recodetabs[font].flags == 0)
1729     recodetabs[font].flags = RECODETAB_TRIED;
1730   return ok ? 0 : -1;
1731 }
1732
1733 void
1734 LoadFontTranslationsForEncoding(encoding)
1735 int encoding;
1736 {
1737   char *c;
1738   int f;
1739
1740   debug1("LoadFontTranslationsForEncoding: encoding %d\n", encoding);
1741   if ((c = encodings[encoding].fontlist) != 0)
1742     while ((f = (unsigned char)*c++) != 0)
1743       if (recodetabs[f].flags == 0)
1744           LoadFontTranslation(f, 0);
1745   f = encodings[encoding].deffont;
1746   if (f > 0 && recodetabs[f].flags == 0)
1747     LoadFontTranslation(f, 0);
1748 }
1749
1750 #endif /* UTF8 */
1751
1752 #else /* !ENCODINGS */
1753
1754 /* Simple version of EncodeChar to encode font changes for
1755  * copy/paste mode
1756  */
1757 int
1758 EncodeChar(bp, c, encoding, fontp)
1759 char *bp;
1760 int c;
1761 int encoding;
1762 int *fontp;
1763 {
1764   int f, l;
1765   f = (c == -1) ? 0 : c >> 16;
1766   l = 0;
1767   if (fontp && f != *fontp)
1768     {
1769       *fontp = f;
1770       if (f && f < ' ')
1771         {
1772           if (bp)
1773            {
1774              *bp++ = 033;
1775              *bp++ = '$';
1776              if (f > 2)
1777                *bp++ = '(';
1778              *bp++ = '@' + f;
1779            }
1780           l += f > 2 ? 4 : 3;
1781         }
1782       else if (f < 128)
1783         {
1784           if (f == 0)
1785             f = 'B';
1786           if (bp)
1787             {
1788               *bp++ = 033;
1789               *bp++ = '(';
1790               *bp++ = f;
1791             }
1792           l += 3;
1793         }
1794     }
1795   if (c == -1)
1796     return l;
1797   if (c & 0xff00)
1798     {
1799       if (bp)
1800         *bp++ = c >> 8;
1801       l++;
1802     }
1803   if (bp)
1804     *bp++ = c;
1805   return l + 1;
1806 }
1807
1808 #endif /* ENCODINGS */