tools/wmc/mcl.c

   1 /*
   2  * Wine Message Compiler lexical scanner
   3  *
   4  * Copyright 2000 Bertho A. Stultiens (BS)
   5  *
   6  * This library is free software; you can redistribute it and/or
   7  * modify it under the terms of the GNU Lesser General Public
   8  * License as published by the Free Software Foundation; either
   9  * version 2.1 of the License, or (at your option) any later version.
  10  *
  11  * This library 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 GNU
  14  * Lesser General Public License for more details.
  15  *
  16  * You should have received a copy of the GNU Lesser General Public
  17  * License along with this library; if not, write to the Free Software
  18  * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301, USA
  19  */
  20
  21 #include "config.h"
  22
  23 #include <stdio.h>
  24 #include <stdlib.h>
  25 #include <ctype.h>
  26 #include <assert.h>
  27 #include <string.h>
  28
  29 #include "utils.h"
  30 #include "wmc.h"
  31 #include "lang.h"
  32
  33 #include "mcy.tab.h"
  34
  35 /*
  36  * Keywords are case insensitive. All normal input is treated as
  37  * being in codepage iso-8859-1 for ascii input files (unicode
  38  * page 0) and as equivalent unicode if unicode input is selected.
  39  * All normal input, which is not part of a message text, is
  40  * enforced to be unicode page 0. Otherwise an error will be
  41  * generated. The normal file data should only be ASCII because
  42  * that is the basic definition of the grammar.
  43  *
  44  * Byteorder or unicode input is determined automatically by
  45  * reading the first 8 bytes and checking them against unicode
  46  * page 0 byteorder (hibyte must be 0).
  47  * -- FIXME --
  48  * Alternatively, the input is checked against a special byte
  49  * sequence to identify the file.
  50  * -- FIXME --
  51  *
  52  *
  53  * Keywords:
  54  *      Codepages
  55  *      Facility
  56  *      FacilityNames
  57  *      LanguageNames
  58  *      MessageId
  59  *      MessageIdTypedef
  60  *      Severity
  61  *      SeverityNames
  62  *      SymbolicName
  63  *
  64  * Default added identifiers for classes:
  65  * SeverityNames:
  66  *      Success         = 0x0
  67  *      Informational   = 0x1
  68  *      Warning         = 0x2
  69  *      Error           = 0x3
  70  * FacilityNames:
  71  *      System          = 0x0FF
  72  *      Application     = 0xFFF
  73  *
  74  * The 'Codepages' keyword is a wmc extension.
  75  */
  76
  77 static const WCHAR ustr_application[]   = { 'A', 'p', 'p', 'l', 'i', 'c', 'a', 't', 'i', 'o', 'n', 0 };
  78 static const WCHAR ustr_codepages[]     = { 'C', 'o', 'd', 'e', 'p', 'a', 'g', 'e', 's', 0 };
  79 static const WCHAR ustr_english[]       = { 'E', 'n', 'g', 'l', 'i', 's', 'h', 0 };
  80 static const WCHAR ustr_error[]         = { 'E', 'r', 'r', 'o', 'r', 0 };
  81 static const WCHAR ustr_facility[]      = { 'F', 'a', 'c', 'i', 'l', 'i', 't', 'y', 0 };
  82 static const WCHAR ustr_facilitynames[] = { 'F', 'a', 'c', 'i', 'l', 'i', 't', 'y', 'N', 'a', 'm', 'e', 's', 0 };
  83 static const WCHAR ustr_informational[] = { 'I', 'n', 'f', 'o', 'r', 'm', 'a', 't', 'i', 'o', 'n', 'a', 'l', 0 };
  84 static const WCHAR ustr_language[]      = { 'L', 'a', 'n', 'g', 'u', 'a', 'g', 'e', 0};
  85 static const WCHAR ustr_languagenames[] = { 'L', 'a', 'n', 'g', 'u', 'a', 'g', 'e', 'N', 'a', 'm', 'e', 's', 0};
  86 static const WCHAR ustr_messageid[]     = { 'M', 'e', 's', 's', 'a', 'g', 'e', 'I', 'd', 0 };
  87 static const WCHAR ustr_messageidtypedef[] = { 'M', 'e', 's', 's', 'a', 'g', 'e', 'I', 'd', 'T', 'y', 'p', 'e', 'd', 'e', 'f', 0 };
  88 static const WCHAR ustr_outputbase[]    = { 'O', 'u', 't', 'p', 'u', 't', 'B', 'a', 's', 'e', 0 };
  89 static const WCHAR ustr_severity[]      = { 'S', 'e', 'v', 'e', 'r', 'i', 't', 'y', 0 };
  90 static const WCHAR ustr_severitynames[] = { 'S', 'e', 'v', 'e', 'r', 'i', 't', 'y', 'N', 'a', 'm', 'e', 's', 0 };
  91 static const WCHAR ustr_success[]       = { 'S', 'u', 'c', 'c', 'e', 's', 's', 0 };
  92 static const WCHAR ustr_symbolicname[]  = { 'S', 'y', 'm', 'b', 'o', 'l', 'i', 'c', 'N', 'a', 'm', 'e', 0 };
  93 static const WCHAR ustr_system[]        = { 'S', 'y', 's', 't', 'e', 'm', 0 };
  94 static const WCHAR ustr_warning[]       = { 'W', 'a', 'r', 'n', 'i', 'n', 'g', 0 };
  95 static const WCHAR ustr_msg00001[]      = { 'm', 's', 'g', '0', '0', '0', '0', '1', 0 };
  96 /*
  97  * This table is to beat any form of "expression building" to check for
  98  * correct filename characters. It is also used for ident checks.
  99  * FIXME: use it more consistently.
 100  */
 101
 102 #define CH_SHORTNAME    0x01
 103 #define CH_LONGNAME     0x02
 104 #define CH_IDENT        0x04
 105 #define CH_NUMBER       0x08
 106 /*#define CH_WILDCARD   0x10*/
 107 /*#define CH_DOT        0x20*/
 108 #define CH_PUNCT        0x40
 109 #define CH_INVALID      0x80
 110
 111 static const char char_table[256] = {
 112         0x80, 0x80, 0x80, 0x80, 0x80, 0x80, 0x80, 0x80, /* 0x00 - 0x07 */
 113         0x80, 0x80, 0x80, 0x80, 0x80, 0x80, 0x80, 0x80, /* 0x08 - 0x0F */
 114         0x80, 0x80, 0x80, 0x80, 0x80, 0x80, 0x80, 0x80, /* 0x10 - 0x17 */
 115         0x80, 0x80, 0x80, 0x80, 0x80, 0x80, 0x80, 0x80, /* 0x18 - 0x1F */
 116         0x80, 0x03, 0x80, 0x03, 0x03, 0x03, 0x03, 0x03, /* 0x20 - 0x27 " !"#$%&'" */
 117         0x43, 0x43, 0x10, 0x80, 0x03, 0x03, 0x22, 0x80, /* 0x28 - 0x2F "()*+,-./" */
 118         0x0b, 0x0b, 0x0b, 0x0b, 0x0b, 0x0b, 0x0b, 0x0b, /* 0x30 - 0x37 "01234567" */
 119         0x0b, 0x0b, 0xc0, 0x80, 0x80, 0x80, 0x80, 0x10, /* 0x38 - 0x3F "89:;<=>?" */
 120         0x03, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, /* 0x40 - 0x47 "@ABCDEFG" */
 121         0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, /* 0x48 - 0x4F "HIJKLMNO" */
 122         0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, /* 0x50 - 0x57 "PQRSTUVW" */
 123         0x07, 0x07, 0x07, 0x80, 0x80, 0x80, 0x80, 0x07, /* 0x58 - 0x5F "XYZ[\]^_" */
 124         0x03, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, /* 0x60 - 0x67 "`abcdefg" */
 125         0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, /* 0x68 - 0x6F "hijklmno" */
 126         0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, 0x07, /* 0x70 - 0x77 "pqrstuvw" */
 127         0x07, 0x07, 0x07, 0x03, 0x80, 0x03, 0x03, 0x80, /* 0x78 - 0x7F "xyz{|}~ " */
 128         0x02, 0x02, 0x02, 0x02, 0x02, 0x02, 0x02, 0x02, /* 0x80 - 0x87 */
 129         0x02, 0x02, 0x02, 0x02, 0x02, 0x02, 0x02, 0x02, /* 0x88 - 0x8F */
 130         0x02, 0x02, 0x02, 0x02, 0x02, 0x02, 0x02, 0x02, /* 0x90 - 0x97 */
 131         0x02, 0x02, 0x02, 0x02, 0x02, 0x02, 0x02, 0x02, /* 0x98 - 0x9F */
 132         0x02, 0x02, 0x02, 0x02, 0x02, 0x02, 0x02, 0x02, /* 0xA0 - 0xA7 */
 133         0x02, 0x02, 0x02, 0x02, 0x02, 0x02, 0x02, 0x02, /* 0xA8 - 0xAF */
 134         0x02, 0x02, 0x02, 0x02, 0x02, 0x02, 0x02, 0x02, /* 0xB0 - 0xB7 */
 135         0x02, 0x02, 0x02, 0x02, 0x02, 0x02, 0x02, 0x02, /* 0xB8 - 0xBF */
 136         0x02, 0x02, 0x02, 0x02, 0x02, 0x02, 0x02, 0x02, /* 0xC0 - 0xC7 */
 137         0x02, 0x02, 0x02, 0x02, 0x02, 0x02, 0x02, 0x02, /* 0xC8 - 0xCF */
 138         0x02, 0x02, 0x02, 0x02, 0x02, 0x02, 0x02, 0x02, /* 0xD0 - 0xD7 */
 139         0x02, 0x02, 0x02, 0x02, 0x02, 0x02, 0x02, 0x02, /* 0xD8 - 0xDF */
 140         0x02, 0x02, 0x02, 0x02, 0x02, 0x02, 0x02, 0x02, /* 0xE0 - 0xE7 */
 141         0x02, 0x02, 0x02, 0x02, 0x02, 0x02, 0x02, 0x02, /* 0xE8 - 0xEF */
 142         0x02, 0x02, 0x02, 0x02, 0x02, 0x02, 0x02, 0x02, /* 0xF0 - 0xF7 */
 143         0x02, 0x02, 0x02, 0x02, 0x02, 0x02, 0x02, 0x80, /* 0xF8 - 0xFF */
 144 };
 145
 146 static int isisochar(int ch)
 147 {
 148         return !(ch & (~0xff));
 149 }
 150
 151 static int codepage;
 152 static const union cptable *codepage_def;
 153
 154 void set_codepage(int cp)
 155 {
 156         codepage = cp;
 157         codepage_def = find_codepage(codepage);
 158         if(!codepage_def && codepage != CP_UTF8)
 159                 xyyerror("Codepage %d not found; cannot process\n", codepage);
 160 }
 161
 162 /*
 163  * Input functions
 164  */
 165 static int nungetstack = 0;
 166 static int allocungetstack = 0;
 167 static char *ungetstack = NULL;
 168 static int ninputbuffer = 0;
 169 static WCHAR *inputbuffer = NULL;
 170 static char *xlatebuffer = NULL;
 171
 172 #define INPUTBUFFER_SIZE        2048    /* Must be larger than 4 and approx. large enough to hold a line */
 173
 174 /*
 175  * Fill the input buffer with *one* line of input.
 176  * The line is '\n' terminated so that scanning
 177  * messages with translation works as expected
 178  * (otherwise we cannot pre-translate because the
 179  * language is first known one line before the
 180  * actual message).
 181  */
 182 static int fill_inputbuffer(void)
 183 {
 184         int n;
 185         static const char err_fatalread[] = "Fatal: reading input failed";
 186         static int endian = -1;
 187
 188         if(!inputbuffer)
 189         {
 190                 inputbuffer = xmalloc(INPUTBUFFER_SIZE*sizeof(WCHAR));
 191                 xlatebuffer = xmalloc(INPUTBUFFER_SIZE);
 192         }
 193
 194 try_again:
 195         if(!unicodein)
 196         {
 197                 char *cptr;
 198                 cptr = fgets(xlatebuffer, INPUTBUFFER_SIZE, yyin);
 199                 if(!cptr && ferror(yyin))
 200                         xyyerror(err_fatalread);
 201                 else if(!cptr)
 202                         return 0;
 203                 if (codepage_def)
 204                     n = wine_cp_mbstowcs(codepage_def, 0, xlatebuffer, strlen(xlatebuffer)+1, inputbuffer, INPUTBUFFER_SIZE);
 205                 else
 206                     n = wine_utf8_mbstowcs(0, xlatebuffer, strlen(xlatebuffer)+1, inputbuffer, INPUTBUFFER_SIZE);
 207                 if(n < 0)
 208                         internal_error(__FILE__, __LINE__, "Could not translate to unicode (%d)\n", n);
 209                 if(n <= 1)
 210                         goto try_again; /* Should not happen */
 211                 n--;    /* Strip added conversion '\0' from input length */
 212                 /*
 213                  * FIXME:
 214                  * Detect UTF-8 in the first time we read some bytes by
 215                  * checking the special sequence "FE..." or something like
 216                  * that. I need to check www.unicode.org for details.
 217                  */
 218         }
 219         else
 220         {
 221                 if(endian == -1)
 222                 {
 223                         n = fread(inputbuffer, 1, 8, yyin);
 224                         if(n != 8)
 225                         {
 226                                 if(!n && ferror(yyin))
 227                                         xyyerror(err_fatalread);
 228                                 else
 229                                         xyyerror("Fatal: file to short to determine byteorder (should never happen)\n");
 230                         }
 231                         if(isisochar(inputbuffer[0]) &&
 232                                 isisochar(inputbuffer[1]) &&
 233                                 isisochar(inputbuffer[2]) &&
 234                                 isisochar(inputbuffer[3]))
 235                         {
 236 #ifdef WORDS_BIGENDIAN
 237                                 endian = WMC_BO_BIG;
 238 #else
 239                                 endian = WMC_BO_LITTLE;
 240 #endif
 241                         }
 242                         else if(isisochar(BYTESWAP_WORD(inputbuffer[0])) &&
 243                                 isisochar(BYTESWAP_WORD(inputbuffer[1])) &&
 244                                 isisochar(BYTESWAP_WORD(inputbuffer[2])) &&
 245                                 isisochar(BYTESWAP_WORD(inputbuffer[3])))
 246                         {
 247 #ifdef WORDS_BIGENDIAN
 248                                 endian = WMC_BO_LITTLE;
 249 #else
 250                                 endian = WMC_BO_BIG;
 251 #endif
 252                         }
 253                         else
 254                                 xyyerror("Fatal: cannot determine file's byteorder\n");
 255                         /* FIXME:
 256                          * Determine the file-endian with the leader-bytes
 257                          * "FF FE..."; can't remember the exact sequence.
 258                          */
 259                         n /= 2;
 260 #ifdef WORDS_BIGENDIAN
 261                         if(endian == WMC_BO_LITTLE)
 262 #else
 263                         if(endian == WMC_BO_BIG)
 264 #endif
 265                         {
 266                                 inputbuffer[0] = BYTESWAP_WORD(inputbuffer[0]);
 267                                 inputbuffer[1] = BYTESWAP_WORD(inputbuffer[1]);
 268                                 inputbuffer[2] = BYTESWAP_WORD(inputbuffer[2]);
 269                                 inputbuffer[3] = BYTESWAP_WORD(inputbuffer[3]);
 270                         }
 271
 272                 }
 273                 else
 274                 {
 275                         int i;
 276                         n = 0;
 277                         for(i = 0; i < INPUTBUFFER_SIZE; i++)
 278                         {
 279                                 int t;
 280                                 t = fread(&inputbuffer[i], 2, 1, yyin);
 281                                 if(!t && ferror(yyin))
 282                                         xyyerror(err_fatalread);
 283                                 else if(!t && n)
 284                                         break;
 285                                 n++;
 286 #ifdef WORDS_BIGENDIAN
 287                                 if(endian == WMC_BO_LITTLE)
 288 #else
 289                                 if(endian == WMC_BO_BIG)
 290 #endif
 291                                 {
 292                                         if((inputbuffer[i] = BYTESWAP_WORD(inputbuffer[i])) == '\n')
 293                                                 break;
 294                                 }
 295                                 else
 296                                 {
 297                                         if(inputbuffer[i] == '\n')
 298                                                 break;
 299                                 }
 300                         }
 301                 }
 302
 303         }
 304
 305         if(!n)
 306         {
 307                 mcy_warning("Re-read line (input was or converted to zilch)\n");
 308                 goto try_again; /* Should not happen, but could be due to stdin reading and a signal */
 309         }
 310
 311         ninputbuffer += n;
 312         return 1;
 313 }
 314
 315 static int get_unichar(void)
 316 {
 317         static WCHAR *b = NULL;
 318         char_number++;
 319
 320         if(nungetstack)
 321                 return ungetstack[--nungetstack];
 322
 323         if(!ninputbuffer)
 324         {
 325                 if(!fill_inputbuffer())
 326                         return EOF;
 327                 b = inputbuffer;
 328         }
 329
 330         ninputbuffer--;
 331         return (int)(*b++ & 0xffff);
 332 }
 333
 334 static void unget_unichar(int ch)
 335 {
 336         if(ch == EOF)
 337                 return;
 338
 339         char_number--;
 340
 341         if(nungetstack == allocungetstack)
 342         {
 343                 allocungetstack += 32;
 344                 ungetstack = xrealloc(ungetstack, allocungetstack * sizeof(*ungetstack));
 345         }
 346
 347         ungetstack[nungetstack++] = (WCHAR)ch;
 348 }
 349
 350
 351 /*
 352  * Normal character stack.
 353  * Used for number scanning.
 354  */
 355 static int ncharstack = 0;
 356 static int alloccharstack = 0;
 357 static char *charstack = NULL;
 358
 359 static void empty_char_stack(void)
 360 {
 361         ncharstack = 0;
 362 }
 363
 364 static void push_char(int ch)
 365 {
 366         if(ncharstack == alloccharstack)
 367         {
 368                 alloccharstack += 32;
 369                 charstack = xrealloc(charstack, alloccharstack * sizeof(*charstack));
 370         }
 371         charstack[ncharstack++] = (char)ch;
 372 }
 373
 374 static int tos_char_stack(void)
 375 {
 376         if(!ncharstack)
 377                 return 0;
 378         else
 379                 return (int)(charstack[ncharstack-1] & 0xff);
 380 }
 381
 382 static char *get_char_stack(void)
 383 {
 384         return charstack;
 385 }
 386
 387 /*
 388  * Unicode character stack.
 389  * Used for general scanner.
 390  */
 391 static int nunicharstack = 0;
 392 static int allocunicharstack = 0;
 393 static WCHAR *unicharstack = NULL;
 394
 395 static void empty_unichar_stack(void)
 396 {
 397         nunicharstack = 0;
 398 }
 399
 400 static void push_unichar(int ch)
 401 {
 402         if(nunicharstack == allocunicharstack)
 403         {
 404                 allocunicharstack += 128;
 405                 unicharstack = xrealloc(unicharstack, allocunicharstack * sizeof(*unicharstack));
 406         }
 407         unicharstack[nunicharstack++] = (WCHAR)ch;
 408 }
 409
 410 #if 0
 411 static int tos_unichar_stack(void)
 412 {
 413         if(!nunicharstack)
 414                 return 0;
 415         else
 416                 return (int)(unicharstack[nunicharstack-1] & 0xffff);
 417 }
 418 #endif
 419
 420 static WCHAR *get_unichar_stack(void)
 421 {
 422         return unicharstack;
 423 }
 424
 425 /*
 426  * Number scanner
 427  *
 428  * state |      ch         | next state
 429  * ------+-----------------+--------------------------
 430  *   0   | [0]             | 1
 431  *   0   | [1-9]           | 4
 432  *   0   | .               | error (should never occur)
 433  *   1   | [xX]            | 2
 434  *   1   | [0-7]           | 3
 435  *   1   | [89a-wyzA-WYZ_] | error invalid digit
 436  *   1   | .               | return 0
 437  *   2   | [0-9a-fA-F]     | 2
 438  *   2   | [g-zG-Z_]       | error invalid hex digit
 439  *   2   | .               | return (hex-number) if TOS != [xX] else error
 440  *   3   | [0-7]           | 3
 441  *   3   | [89a-zA-Z_]     | error invalid octal digit
 442  *   3   | .               | return (octal-number)
 443  *   4   | [0-9]           | 4
 444  *   4   | [a-zA-Z_]       | error invalid decimal digit
 445  *   4   | .               | return (decimal-number)
 446  *
 447  * All non-identifier characters [^a-zA-Z_0-9] terminate the scan
 448  * and return the value. This is not entirely correct, but close
 449  * enough (should check punctuators as trailing context, but the
 450  * char_table is not adapted to that and it is questionable whether
 451  * it is worth the trouble).
 452  * All non-iso-8859-1 characters are an error.
 453  */
 454 static int scan_number(int ch)
 455 {
 456         int state = 0;
 457         int base = 10;
 458         empty_char_stack();
 459
 460         while(1)
 461         {
 462                 if(!isisochar(ch))
 463                         xyyerror("Invalid digit\n");
 464
 465                 switch(state)
 466                 {
 467                 case 0:
 468                         if(isdigit(ch))
 469                         {
 470                                 push_char(ch);
 471                                 if(ch == '0')
 472                                         state = 1;
 473                                 else
 474                                         state = 4;
 475                         }
 476                         else
 477                                 internal_error(__FILE__, __LINE__, "Non-digit in first number-scanner state\n");
 478                         break;
 479                 case 1:
 480                         if(ch == 'x' || ch == 'X')
 481                         {
 482                                 push_char(ch);
 483                                 state = 2;
 484                         }
 485                         else if(ch >= '0' && ch <= '7')
 486                         {
 487                                 push_char(ch);
 488                                 state = 3;
 489                         }
 490                         else if(isalpha(ch) || ch == '_')
 491                                 xyyerror("Invalid number digit\n");
 492                         else
 493                         {
 494                                 unget_unichar(ch);
 495                                 mcy_lval.num = 0;
 496                                 return tNUMBER;
 497                         }
 498                         break;
 499                 case 2:
 500                         if(isxdigit(ch))
 501                                 push_char(ch);
 502                         else if(isalpha(ch) || ch == '_' || !isxdigit(tos_char_stack()))
 503                                 xyyerror("Invalid hex digit\n");
 504                         else
 505                         {
 506                                 base = 16;
 507                                 goto finish;
 508                         }
 509                         break;
 510                 case 3:
 511                         if(ch >= '0' && ch <= '7')
 512                                 push_char(ch);
 513                         else if(isalnum(ch) || ch == '_')
 514                                 xyyerror("Invalid octal digit\n");
 515                         else
 516                         {
 517                                 base = 8;
 518                                 goto finish;
 519                         }
 520                         break;
 521                 case 4:
 522                         if(isdigit(ch))
 523                                 push_char(ch);
 524                         else if(isalnum(ch) || ch == '_')
 525                                 xyyerror("Invalid decimal digit\n");
 526                         else
 527                         {
 528                                 base = 10;
 529                                 goto finish;
 530                         }
 531                         break;
 532                 default:
 533                         internal_error(__FILE__, __LINE__, "Invalid state in number-scanner\n");
 534                 }
 535                 ch = get_unichar();
 536         }
 537 finish:
 538         unget_unichar(ch);
 539         push_char(0);
 540         mcy_lval.num = strtoul(get_char_stack(), NULL, base);
 541         return tNUMBER;
 542 }
 543
 544 static void newline(void)
 545 {
 546         line_number++;
 547         char_number = 1;
 548 }
 549
 550 static int unisort(const void *p1, const void *p2)
 551 {
 552         return unistricmp(((const token_t *)p1)->name, ((const token_t *)p2)->name);
 553 }
 554
 555 static token_t *tokentable = NULL;
 556 static int ntokentable = 0;
 557
 558 token_t *lookup_token(const WCHAR *s)
 559 {
 560         token_t tok;
 561
 562         tok.name = s;
 563         return (token_t *)bsearch(&tok, tokentable, ntokentable, sizeof(*tokentable), unisort);
 564 }
 565
 566 void add_token(tok_e type, const WCHAR *name, int tok, int cp, const WCHAR *alias, int fix)
 567 {
 568         ntokentable++;
 569         tokentable = xrealloc(tokentable, ntokentable * sizeof(*tokentable));
 570         tokentable[ntokentable-1].type = type;
 571         tokentable[ntokentable-1].name = name;
 572         tokentable[ntokentable-1].token = tok;
 573         tokentable[ntokentable-1].codepage = cp;
 574         tokentable[ntokentable-1].alias = alias;
 575         tokentable[ntokentable-1].fixed = fix;
 576         qsort(tokentable, ntokentable, sizeof(*tokentable), unisort);
 577 }
 578
 579 void get_tokentable(token_t **tab, int *len)
 580 {
 581         assert(tab != NULL);
 582         assert(len != NULL);
 583         *tab = tokentable;
 584         *len = ntokentable;
 585 }
 586
 587 /*
 588  * The scanner
 589  *
 590  */
 591 int mcy_lex(void)
 592 {
 593         static const WCHAR ustr_dot1[] = { '.', '\n', 0 };
 594         static const WCHAR ustr_dot2[] = { '.', '\r', '\n', 0 };
 595         static int isinit = 0;
 596         int ch;
 597
 598         if(!isinit)
 599         {
 600                 isinit++;
 601                 set_codepage(WMC_DEFAULT_CODEPAGE);
 602                 add_token(tok_keyword,  ustr_codepages,         tCODEPAGE,      0, NULL, 0);
 603                 add_token(tok_keyword,  ustr_facility,          tFACILITY,      0, NULL, 1);
 604                 add_token(tok_keyword,  ustr_facilitynames,     tFACNAMES,      0, NULL, 1);
 605                 add_token(tok_keyword,  ustr_language,          tLANGUAGE,      0, NULL, 1);
 606                 add_token(tok_keyword,  ustr_languagenames,     tLANNAMES,      0, NULL, 1);
 607                 add_token(tok_keyword,  ustr_messageid,         tMSGID,         0, NULL, 1);
 608                 add_token(tok_keyword,  ustr_messageidtypedef,  tTYPEDEF,       0, NULL, 1);
 609                 add_token(tok_keyword,  ustr_outputbase,        tBASE,          0, NULL, 1);
 610                 add_token(tok_keyword,  ustr_severity,          tSEVERITY,      0, NULL, 1);
 611                 add_token(tok_keyword,  ustr_severitynames,     tSEVNAMES,      0, NULL, 1);
 612                 add_token(tok_keyword,  ustr_symbolicname,      tSYMNAME,       0, NULL, 1);
 613                 add_token(tok_severity, ustr_error,             0x03,           0, NULL, 0);
 614                 add_token(tok_severity, ustr_warning,           0x02,           0, NULL, 0);
 615                 add_token(tok_severity, ustr_informational,     0x01,           0, NULL, 0);
 616                 add_token(tok_severity, ustr_success,           0x00,           0, NULL, 0);
 617                 add_token(tok_facility, ustr_application,       0xFFF,          0, NULL, 0);
 618                 add_token(tok_facility, ustr_system,            0x0FF,          0, NULL, 0);
 619                 add_token(tok_language, ustr_english,           0x409,          437, ustr_msg00001, 0);
 620         }
 621
 622         empty_unichar_stack();
 623
 624         while(1)
 625         {
 626                 if(want_line)
 627                 {
 628                         while((ch = get_unichar()) != '\n')
 629                         {
 630                                 if(ch == EOF)
 631                                         xyyerror("Unexpected EOF\n");
 632                                 push_unichar(ch);
 633                         }
 634                         newline();
 635                         push_unichar(ch);
 636                         push_unichar(0);
 637                         if(!unistrcmp(ustr_dot1, get_unichar_stack()) || !unistrcmp(ustr_dot2, get_unichar_stack()))
 638                         {
 639                                 want_line = 0;
 640                                 /* Reset the codepage to our default after each message */
 641                                 set_codepage(WMC_DEFAULT_CODEPAGE);
 642                                 return tMSGEND;
 643                         }
 644                         mcy_lval.str = xunistrdup(get_unichar_stack());
 645                         return tLINE;
 646                 }
 647
 648                 ch = get_unichar();
 649
 650                 if(ch == EOF)
 651                         return EOF;
 652
 653                 if(ch == '\n')
 654                 {
 655                         newline();
 656                         if(want_nl)
 657                         {
 658                                 want_nl = 0;
 659                                 return tNL;
 660                         }
 661                         continue;
 662                 }
 663
 664                 if(isisochar(ch))
 665                 {
 666                         if(want_file)
 667                         {
 668                                 int n = 0;
 669                                 while(n < 8 && isisochar(ch))
 670                                 {
 671                                         int t = char_table[ch];
 672                                         if((t & CH_PUNCT) || !(t & CH_SHORTNAME))
 673                                                 break;
 674
 675                                         push_unichar(ch);
 676                                         n++;
 677                                         ch = get_unichar();
 678                                 }
 679                                 unget_unichar(ch);
 680                                 push_unichar(0);
 681                                 want_file = 0;
 682                                 mcy_lval.str = xunistrdup(get_unichar_stack());
 683                                 return tFILE;
 684                         }
 685
 686                         if(char_table[ch] & CH_IDENT)
 687                         {
 688                                 token_t *tok;
 689                                 while(isisochar(ch) && (char_table[ch] & (CH_IDENT|CH_NUMBER)))
 690                                 {
 691                                         push_unichar(ch);
 692                                         ch = get_unichar();
 693                                 }
 694                                 unget_unichar(ch);
 695                                 push_unichar(0);
 696                                 if(!(tok = lookup_token(get_unichar_stack())))
 697                                 {
 698                                         mcy_lval.str = xunistrdup(get_unichar_stack());
 699                                         return tIDENT;
 700                                 }
 701                                 switch(tok->type)
 702                                 {
 703                                 case tok_keyword:
 704                                         return tok->token;
 705
 706                                 case tok_language:
 707                                         codepage = tok->codepage;
 708                                         /* Fall through */
 709                                 case tok_severity:
 710                                 case tok_facility:
 711                                         mcy_lval.tok = tok;
 712                                         return tTOKEN;
 713
 714                                 default:
 715                                         internal_error(__FILE__, __LINE__, "Invalid token type encountered\n");
 716                                 }
 717                         }
 718
 719                         if(isspace(ch)) /* Ignore space */
 720                                 continue;
 721
 722                         if(isdigit(ch))
 723                                 return scan_number(ch);
 724                 }
 725
 726                 switch(ch)
 727                 {
 728                 case ':':
 729                 case '=':
 730                 case '+':
 731                 case '(':
 732                 case ')':
 733                         return ch;
 734                 case ';':
 735                         while(ch != '\n' && ch != EOF)
 736                         {
 737                                 push_unichar(ch);
 738                                 ch = get_unichar();
 739                         }
 740                         newline();
 741                         push_unichar(ch);       /* Include the newline */
 742                         push_unichar(0);
 743                         mcy_lval.str = xunistrdup(get_unichar_stack());
 744                         return tCOMMENT;
 745                 default:
 746                         xyyerror("Invalid character '%c' (0x%04x)\n", isisochar(ch) && isprint(ch) ? ch : '.', ch);
 747                 }
 748         }
 749 }