47b675e0accfc231f6886e24d3df661be6ba66c5
[reactos.git] / reactos / tools / create_nls / create_nls.c
1 /*
2 * Tool for creating NT-like NLS files for Unicode <-> Codepage conversions.
3 * Tool for creating NT-like l_intl.nls file for case mapping of unicode
4 * characters.
5 * Copyright 2000 Timoshkov Dmitry
6 * Copyright 2001 Matei Alexandru
7 *
8 * Sources of information:
9 * Andrew Kozin's YAW project http://www.chat.ru/~stanson/yaw_en.html
10 * Ove K├Áven's investigations http://www.ping.uio.no/~ovehk/nls
11 */
12 #include <windows.h>
13 #include <stdio.h>
14 #include <stdlib.h>
15 #include <malloc.h>
16 #include <string.h>
17 #include <ctype.h>
18
19 static const WCHAR * const uprtable[256];
20 static const WCHAR * const lwrtable[256];
21
22 #define NLSDIR "../../media/nls"
23 #define LIBDIR "unicode.org/"
24
25 typedef struct {
26 WORD wSize; /* in words 0x000D */
27 WORD CodePage;
28 WORD MaxCharSize; /* 1 or 2 */
29 BYTE DefaultChar[MAX_DEFAULTCHAR];
30 WCHAR UnicodeDefaultChar;
31 WCHAR unknown1;
32 WCHAR unknown2;
33 BYTE LeadByte[MAX_LEADBYTES];
34 } __attribute__((packed)) NLS_FILE_HEADER;
35
36 /*
37 Support for translation from the multiple unicode chars
38 to the single code page char.
39
40 002D;HYPHEN-MINUS;Pd;0;ET;;;;;N;;;;;
41 00AD;SOFT HYPHEN;Pd;0;ON;;;;;N;;;;;
42 2010;HYPHEN;Pd;0;ON;;;;;N;;;;;
43 2011;NON-BREAKING HYPHEN;Pd;0;ON;<noBreak> 2010;;;;N;;;;;
44 2013;EN DASH;Pd;0;ON;;;;;N;;;;;
45 2014;EM DASH;Pd;0;ON;;;;;N;;;;;
46 2015;HORIZONTAL BAR;Pd;0;ON;;;;;N;QUOTATION DASH;;;;
47 */
48
49 /* HYPHEN-MINUS aliases */
50 static WCHAR hyphen_aliases[] = {0x00AD,0x2010,0x2011,0x2013,0x2014,0x2015,0};
51
52 static struct {
53 WCHAR cp_char;
54 WCHAR *alias; /* must be 0 terminated */
55 } u2cp_alias[] = {
56 /* HYPHEN-MINUS aliases */
57 {0x002D, hyphen_aliases}
58 };
59
60 static void patch_aliases(void *u2cp, CPINFOEXA *cpi)
61 {
62 int i, j;
63 WCHAR *wc, *alias;
64 BYTE *c;
65
66 if(cpi->MaxCharSize == 2) {
67 wc = (WCHAR *)u2cp;
68 for(i = 0; i < 65536; i++) {
69 for(j = 0; j < sizeof(u2cp_alias)/sizeof(u2cp_alias[0]); j++) {
70 alias = u2cp_alias[j].alias;
71 while(*alias) {
72 if(*alias == i && wc[i] == *(WCHAR *)cpi->DefaultChar) {
73 wc[i] = u2cp_alias[j].cp_char;
74 }
75 alias++;
76 }
77 }
78 }
79 }
80 else {
81 c = (BYTE *)u2cp;
82 for(i = 0; i < 65536; i++) {
83 for(j = 0; j < sizeof(u2cp_alias)/sizeof(u2cp_alias[0]); j++) {
84 alias = u2cp_alias[j].alias;
85 while(*alias) {
86 if(*alias == i && c[i] == cpi->DefaultChar[0] && u2cp_alias[j].cp_char < 256) {
87 c[i] = (BYTE)u2cp_alias[j].cp_char;
88 }
89 alias++;
90 }
91 }
92 }
93 }
94 }
95
96 static BOOL write_unicode2cp_table(FILE *out, CPINFOEXA *cpi, WCHAR *table)
97 {
98 void *u2cp;
99 WCHAR *wc;
100 CHAR *c;
101 int i;
102 BOOL ret = TRUE;
103
104 u2cp = malloc(cpi->MaxCharSize * 65536);
105 if(!u2cp) {
106 printf("Not enough memory for Unicode to Codepage table\n");
107 return FALSE;
108 }
109
110 if(cpi->MaxCharSize == 2) {
111 wc = (WCHAR *)u2cp;
112 for(i = 0; i < 65536; i++)
113 wc[i] = *(WCHAR *)cpi->DefaultChar;
114
115 for(i = 0; i < 65536; i++)
116 if (table[i] != '?')
117 wc[table[i]] = (WCHAR)i;
118 }
119 else {
120 c = (CHAR *)u2cp;
121 for(i = 0; i < 65536; i++)
122 c[i] = cpi->DefaultChar[0];
123
124 for(i = 0; i < 256; i++)
125 if (table[i] != '?')
126 c[table[i]] = (CHAR)i;
127 }
128
129 patch_aliases(u2cp, cpi);
130
131 if(fwrite(u2cp, 1, cpi->MaxCharSize * 65536, out) != cpi->MaxCharSize * 65536)
132 ret = FALSE;
133
134 free(u2cp);
135
136 return ret;
137 }
138
139 static BOOL write_lb_ranges(FILE *out, CPINFOEXA *cpi, WCHAR *table)
140 {
141 WCHAR sub_table[256];
142 WORD offset, offsets[256];
143 int i, j, range;
144
145 memset(offsets, 0, sizeof(offsets));
146
147 offset = 0;
148
149 for(i = 0; i < MAX_LEADBYTES; i += 2) {
150 for(range = cpi->LeadByte[i]; range != 0 && range <= cpi->LeadByte[i + 1]; range++) {
151 offset += 256;
152 offsets[range] = offset;
153 }
154 }
155
156 if(fwrite(offsets, 1, sizeof(offsets), out) != sizeof(offsets))
157 return FALSE;
158
159 for(i = 0; i < MAX_LEADBYTES; i += 2) {
160 for(range = cpi->LeadByte[i]; range != 0 && range <= cpi->LeadByte[i + 1]; range++) {
161 /*printf("Writing sub table for LeadByte %02X\n", range);*/
162 for(j = MAKEWORD(0, range); j <= MAKEWORD(0xFF, range); j++) {
163 sub_table[j - MAKEWORD(0, range)] = table[j];
164 }
165
166 if(fwrite(sub_table, 1, sizeof(sub_table), out) != sizeof(sub_table))
167 return FALSE;
168 }
169 }
170
171 return TRUE;
172 }
173
174 static BOOL create_nls_file(char *name, CPINFOEXA *cpi, WCHAR *table, WCHAR *oemtable)
175 {
176 FILE *out;
177 NLS_FILE_HEADER nls;
178 WORD wValue, number_of_lb_ranges, number_of_lb_subtables, i;
179
180 printf("Creating NLS table \"%s\"\n", name);
181
182 if(!(out = fopen(name, "wb"))) {
183 printf("Could not create file \"%s\"\n", name);
184 return FALSE;
185 }
186
187 memset(&nls, 0, sizeof(nls));
188
189 nls.wSize = sizeof(nls) / sizeof(WORD);
190 nls.CodePage = cpi->CodePage;
191 nls.MaxCharSize = cpi->MaxCharSize;
192 memcpy(nls.DefaultChar, cpi->DefaultChar, MAX_DEFAULTCHAR);
193 nls.UnicodeDefaultChar = cpi->UnicodeDefaultChar;
194 nls.unknown1 = '?';
195 nls.unknown2 = '?';
196 memcpy(nls.LeadByte, cpi->LeadByte, MAX_LEADBYTES);
197
198 if(fwrite(&nls, 1, sizeof(nls), out) != sizeof(nls)) {
199 fclose(out);
200 printf("Could not write to file \"%s\"\n", name);
201 return FALSE;
202 }
203
204 number_of_lb_ranges = 0;
205 number_of_lb_subtables = 0;
206
207 for(i = 0; i < MAX_LEADBYTES; i += 2) {
208 if(cpi->LeadByte[i] != 0 && cpi->LeadByte[i + 1] > cpi->LeadByte[i]) {
209 number_of_lb_ranges++;
210 number_of_lb_subtables += cpi->LeadByte[i + 1] - cpi->LeadByte[i] + 1;
211 }
212 }
213
214 /*printf("Number of LeadByte ranges %d\n", number_of_lb_ranges);*/
215 /*printf("Number of LeadByte subtables %d\n", number_of_lb_subtables);*/
216
217 /* Calculate offset to Unicode to CP table in words:
218 * 1. (256 * sizeof(WORD)) primary CP to Unicode table +
219 * 2. (WORD) optional OEM glyph table size in words +
220 * 3. OEM glyph table size in words * sizeof(WORD) +
221 * 4. (WORD) Number of DBCS LeadByte ranges +
222 * 5. if (Number of DBCS LeadByte ranges != 0) 256 * sizeof(WORD) offsets of lead byte sub tables
223 * 6. (Number of DBCS LeadByte sub tables * 256 * sizeof(WORD)) LeadByte sub tables +
224 * 7. (WORD) Unknown flag
225 */
226
227 wValue = (256 * sizeof(WORD) + /* 1 */
228 sizeof(WORD) + /* 2 */
229 ((oemtable !=NULL) ? (256 * sizeof(WORD)) : 0) + /* 3 */
230 sizeof(WORD) + /* 4 */
231 ((number_of_lb_subtables != 0) ? 256 * sizeof(WORD) : 0) + /* 5 */
232 number_of_lb_subtables * 256 * sizeof(WORD) + /* 6 */
233 sizeof(WORD) /* 7 */
234 ) / sizeof(WORD);
235
236 /* offset of Unicode to CP table in words */
237 fwrite(&wValue, 1, sizeof(wValue), out);
238
239 /* primary CP to Unicode table */
240 if(fwrite(table, 1, 256 * sizeof(WCHAR), out) != 256 * sizeof(WCHAR)) {
241 fclose(out);
242 printf("Could not write to file \"%s\"\n", name);
243 return FALSE;
244 }
245
246 /* optional OEM glyph table size in words */
247 wValue = (oemtable != NULL) ? (256 * sizeof(WORD)) : 0;
248 fwrite(&wValue, 1, sizeof(wValue), out);
249
250 /* optional OEM to Unicode table */
251 if (oemtable) {
252 if(fwrite(oemtable, 1, 256 * sizeof(WCHAR), out) != 256 * sizeof(WCHAR)) {
253 fclose(out);
254 printf("Could not write to file \"%s\"\n", name);
255 return FALSE;
256 }
257 }
258
259 /* Number of DBCS LeadByte ranges */
260 fwrite(&number_of_lb_ranges, 1, sizeof(number_of_lb_ranges), out);
261
262 /* offsets of lead byte sub tables and lead byte sub tables */
263 if(number_of_lb_ranges > 0) {
264 if(!write_lb_ranges(out, cpi, table)) {
265 fclose(out);
266 printf("Could not write to file \"%s\"\n", name);
267 return FALSE;
268 }
269 }
270
271 /* Unknown flag */
272 wValue = 0;
273 fwrite(&wValue, 1, sizeof(wValue), out);
274
275 if(!write_unicode2cp_table(out, cpi, table)) {
276 fclose(out);
277 printf("Could not write to file \"%s\"\n", name);
278 return FALSE;
279 }
280
281 fclose(out);
282 return TRUE;
283 }
284
285 /* correct the codepage information such as default chars */
286 static void patch_codepage_info(CPINFOEXA *cpi)
287 {
288 /* currently nothing */
289 }
290
291 static WCHAR *Load_CP2Unicode_Table(char *table_name, UINT cp, CPINFOEXA *cpi)
292 {
293 char buf[256];
294 char *p;
295 DWORD n, value;
296 FILE *file;
297 WCHAR *table;
298 int lb_ranges, lb_range_started, line;
299
300 printf("Loading translation table \"%s\"\n", table_name);
301
302 /* Init to default values */
303 memset(cpi, 0, sizeof(CPINFOEXA));
304 cpi->CodePage = cp;
305 *(WCHAR *)cpi->DefaultChar = '?';
306 cpi->MaxCharSize = 1;
307 cpi->UnicodeDefaultChar = '?';
308
309 patch_codepage_info(cpi);
310
311 table = (WCHAR *)malloc(sizeof(WCHAR) * 65536);
312 if(!table) {
313 printf("Not enough memory for Codepage to Unicode table\n");
314 return NULL;
315 }
316
317 for(n = 0; n < 256; n++)
318 table[n] = (WCHAR)n;
319
320 for(n = 256; n < 65536; n++)
321 table[n] = cpi->UnicodeDefaultChar;
322
323 file = fopen(table_name, "r");
324 if(file == NULL) {
325 free(table);
326 return NULL;
327 }
328
329 line = 0;
330 lb_ranges = 0;
331 lb_range_started = 0;
332
333 while(fgets(buf, sizeof(buf), file)) {
334 line++;
335 p = buf;
336 while(isspace(*p)) p++;
337
338 if(!*p || p[0] == '#')
339 continue;
340
341 n = strtol(p, &p, 0);
342 if(n > 0xFFFF) {
343 printf("Line %d: Entry 0x%06lX: File \"%s\" corrupted\n", line, n, table_name);
344 continue;
345 }
346
347 if(n > 0xFF && cpi->MaxCharSize != 2) {
348 /*printf("Line %d: Entry 0x%04lX: Switching to DBCS\n", line, n);*/
349 cpi->MaxCharSize = 2;
350 }
351
352 while(isspace(*p)) p++;
353
354 if(!*p || p[0] == '#') {
355 /*printf("Line %d: Entry 0x%02lX has no Unicode value\n", line, n);*/
356 }
357 else {
358 value = strtol(p, &p, 0);
359 if(value > 0xFFFF) {
360 printf("Line %d: Entry 0x%06lX unicode value: File \"%s\" corrupted\n", line, n, table_name);
361 }
362 table[n] = (WCHAR)value;
363 }
364
365 /* wait for comment */
366 while(*p && *p != '#') p++;
367
368 if(*p == '#' && strstr(p, "DBCS LEAD BYTE")) {
369 /*printf("Line %d, entry 0x%02lX DBCS LEAD BYTE\n", line, n);*/
370 if(n > 0xFF) {
371 printf("Line %d: Entry 0x%04lX: Error: DBCS lead byte overflowed\n", line, n);
372 continue;
373 }
374
375 table[n] = (WCHAR)0;
376
377 if(lb_range_started) {
378 cpi->LeadByte[(lb_ranges - 1) * 2 + 1] = (BYTE)n;
379 }
380 else {
381 /*printf("Line %d: Starting new DBCS lead byte range, entry 0x%02lX\n", line, n);*/
382 if(lb_ranges < MAX_LEADBYTES/2) {
383 lb_ranges++;
384 lb_range_started = 1;
385 cpi->LeadByte[(lb_ranges - 1) * 2] = (BYTE)n;
386 }
387 else
388 printf("Line %d: Error: could not start new lead byte range\n", line);
389 }
390 }
391 else {
392 if(lb_range_started)
393 lb_range_started = 0;
394 }
395 }
396
397 fclose(file);
398
399 return table;
400 }
401
402 static WCHAR *Load_OEM2Unicode_Table(char *table_name, WCHAR *def_table, UINT cp, CPINFOEXA *cpi)
403 {
404 char buf[256];
405 char *p;
406 DWORD n, value;
407 FILE *file;
408 WCHAR *table;
409 int line;
410
411 printf("Loading oem glyph table \"%s\"\n", table_name);
412
413 table = (WCHAR *)malloc(sizeof(WCHAR) * 65536);
414 if(!table) {
415 printf("Not enough memory for Codepage to Unicode table\n");
416 return NULL;
417 }
418
419 memcpy(table, def_table, 65536 * sizeof(WCHAR));
420
421 file = fopen(table_name, "r");
422 if(file == NULL) {
423 free(table);
424 return NULL;
425 }
426
427 while(fgets(buf, sizeof(buf), file)) {
428 line++;
429 p = buf;
430 while(isspace(*p)) p++;
431
432 if(!*p || p[0] == '#')
433 continue;
434
435 value = strtol(p, &p, 16);
436 if(value > 0xFFFF) {
437 printf("Line %d: Entry 0x%06lX: File \"%s\" corrupted\n", line, value, table_name);
438 continue;
439 }
440
441 while(isspace(*p)) p++;
442
443 if(!*p || p[0] == '#') {
444 /*printf("Line %d: Entry 0x%02lX has no Unicode value\n", line, n);*/
445 continue;
446 }
447 else {
448 n = strtol(p, &p, 16);
449 if(n > 0xFFFF) {
450 printf("Line %d: Entry 0x%06lX unicode value: File \"%s\" corrupted\n", line, value, table_name);
451 continue;
452 }
453 }
454
455 if (cpi->CodePage == 864) {
456 while(isspace(*p)) p++;
457
458 if(!*p || p[0] == '#' || p[0] == '-') {
459 /*printf("Line %d: Entry 0x%02lX has no Unicode value\n", line, n);*/
460 continue;
461 }
462 else {
463 n = strtol(p, &p, 16);
464 if(n > 0xFFFF) {
465 printf("Line %d: Entry 0x%06lX oem value: File \"%s\" corrupted\n", line, value, table_name);
466 }
467 continue;
468 }
469 }
470
471 table[n] = (WCHAR)value;
472 }
473
474 fclose(file);
475
476 return table;
477 }
478
479 int write_nls_files()
480 {
481 WCHAR *table;
482 WCHAR *oemtable;
483 char nls_filename[256];
484 CPINFOEXA cpi;
485 int i;
486 struct code_page {
487 UINT cp;
488 BOOL oem;
489 char *table_filename;
490 char *comment;
491 } pages[] = {
492 {37, FALSE, LIBDIR"MAPPINGS/VENDORS/MICSFT/EBCDIC/CP037.TXT", "IBM EBCDIC US Canada"},
493 {424, FALSE, LIBDIR"MAPPINGS/VENDORS/MISC/CP424.TXT", "IBM EBCDIC Hebrew"},
494 {437, TRUE, LIBDIR"MAPPINGS/VENDORS/MICSFT/PC/CP437.TXT", "OEM United States"},
495 {500, FALSE, LIBDIR"MAPPINGS/VENDORS/MICSFT/EBCDIC/CP500.TXT", "IBM EBCDIC International"},
496 /*{708, FALSE, "", "Arabic ASMO"},*/
497 /*{720, FALSE, "", "Arabic Transparent ASMO"},*/
498 {737, TRUE, LIBDIR"MAPPINGS/VENDORS/MICSFT/PC/CP737.TXT", "OEM Greek 437G"},
499 {775, TRUE, LIBDIR"MAPPINGS/VENDORS/MICSFT/PC/CP775.TXT", "OEM Baltic"},
500 {850, TRUE, LIBDIR"MAPPINGS/VENDORS/MICSFT/PC/CP850.TXT", "OEM Multilingual Latin 1"},
501 {852, TRUE, LIBDIR"MAPPINGS/VENDORS/MICSFT/PC/CP852.TXT", "OEM Slovak Latin 2"},
502 {855, TRUE, LIBDIR"MAPPINGS/VENDORS/MICSFT/PC/CP855.TXT", "OEM Cyrillic" },
503 {856, TRUE, LIBDIR"MAPPINGS/VENDORS/MISC/CP856.TXT", "Hebrew PC"},
504 {857, TRUE, LIBDIR"MAPPINGS/VENDORS/MICSFT/PC/CP857.TXT", "OEM Turkish"},
505 {860, TRUE, LIBDIR"MAPPINGS/VENDORS/MICSFT/PC/CP860.TXT", "OEM Portuguese"},
506 {861, TRUE, LIBDIR"MAPPINGS/VENDORS/MICSFT/PC/CP861.TXT", "OEM Icelandic"},
507 {862, TRUE, LIBDIR"MAPPINGS/VENDORS/MICSFT/PC/CP862.TXT", "OEM Hebrew"},
508 {863, TRUE, LIBDIR"MAPPINGS/VENDORS/MICSFT/PC/CP863.TXT", "OEM Canadian French"},
509 {864, TRUE, LIBDIR"MAPPINGS/VENDORS/MICSFT/PC/CP864.TXT", "OEM Arabic"},
510 {865, TRUE, LIBDIR"MAPPINGS/VENDORS/MICSFT/PC/CP865.TXT", "OEM Nordic"},
511 {866, TRUE, LIBDIR"MAPPINGS/VENDORS/MICSFT/PC/CP866.TXT", "OEM Russian"},
512 {869, TRUE, LIBDIR"MAPPINGS/VENDORS/MICSFT/PC/CP869.TXT", "OEM Greek"},
513 /*{870, FALSE, "", "IBM EBCDIC Multilingual/ROECE (Latin 2)"},*/
514 {874, FALSE, LIBDIR"MAPPINGS/VENDORS/MICSFT/PC/CP874.TXT", "ANSI/OEM Thai"},
515 {875, FALSE, LIBDIR"MAPPINGS/VENDORS/MICSFT/EBCDIC/CP875.TXT", "IBM EBCDIC Greek"},
516 {878, FALSE, LIBDIR"MAPPINGS/VENDORS/MISC/KOI8-R.TXT", "Russian KOI8"},
517 {932, FALSE, LIBDIR"MAPPINGS/VENDORS/MICSFT/WINDOWS/CP932.TXT", "ANSI/OEM Japanese Shift-JIS"},
518 {936, FALSE, LIBDIR"MAPPINGS/VENDORS/MICSFT/WINDOWS/CP936.TXT", "ANSI/OEM Simplified Chinese GBK"},
519 {949, FALSE, LIBDIR"MAPPINGS/VENDORS/MICSFT/WINDOWS/CP949.TXT", "ANSI/OEM Korean Unified Hangul"},
520 {950, FALSE, LIBDIR"MAPPINGS/VENDORS/MICSFT/WINDOWS/CP950.TXT", "ANSI/OEM Traditional Chinese Big5"},
521 {1006, FALSE, LIBDIR"MAPPINGS/VENDORS/MISC/CP1006.TXT", "IBM Arabic"},
522 {1026, FALSE, LIBDIR"MAPPINGS/VENDORS/MICSFT/EBCDIC/CP1026.TXT", "IBM EBCDIC Latin 5 Turkish"},
523 {1250, FALSE, LIBDIR"MAPPINGS/VENDORS/MICSFT/WINDOWS/CP1250.TXT", "ANSI Eastern Europe"},
524 {1251, FALSE, LIBDIR"MAPPINGS/VENDORS/MICSFT/WINDOWS/CP1251.TXT", "ANSI Cyrillic"},
525 {1252, FALSE, LIBDIR"MAPPINGS/VENDORS/MICSFT/WINDOWS/CP1252.TXT", "ANSI Latin 1"},
526 {1253, FALSE, LIBDIR"MAPPINGS/VENDORS/MICSFT/WINDOWS/CP1253.TXT", "ANSI Greek"},
527 {1254, FALSE, LIBDIR"MAPPINGS/VENDORS/MICSFT/WINDOWS/CP1254.TXT", "ANSI Turkish"},
528 {1255, FALSE, LIBDIR"MAPPINGS/VENDORS/MICSFT/WINDOWS/CP1255.TXT", "ANSI Hebrew"},
529 {1256, FALSE, LIBDIR"MAPPINGS/VENDORS/MICSFT/WINDOWS/CP1256.TXT", "ANSI Arabic"},
530 {1257, FALSE, LIBDIR"MAPPINGS/VENDORS/MICSFT/WINDOWS/CP1257.TXT", "ANSI Baltic"},
531 {1258, FALSE, LIBDIR"MAPPINGS/VENDORS/MICSFT/WINDOWS/CP1258.TXT", "ANSI/OEM Viet Nam"},
532 {10000, FALSE, LIBDIR"MAPPINGS/VENDORS/MICSFT/MAC/ROMAN.TXT", "Mac Roman"},
533 {10006, FALSE, LIBDIR"MAPPINGS/VENDORS/MICSFT/MAC/GREEK.TXT", "Mac Greek"},
534 {10007, FALSE, LIBDIR"MAPPINGS/VENDORS/MICSFT/MAC/CYRILLIC.TXT", "Mac Cyrillic"},
535 {10029, FALSE, LIBDIR"MAPPINGS/VENDORS/MICSFT/MAC/LATIN2.TXT", "Mac Latin 2"},
536 {10079, FALSE, LIBDIR"MAPPINGS/VENDORS/MICSFT/MAC/ICELAND.TXT", "Mac Icelandic"},
537 {10081, FALSE, LIBDIR"MAPPINGS/VENDORS/MICSFT/MAC/TURKISH.TXT", "Mac Turkish"},
538 /*{20000, FALSE, "", "CNS Taiwan"},*/
539 /*{20001, FALSE, "", "TCA Taiwan"},*/
540 /*{20002, FALSE, "", "Eten Taiwan"},*/
541 /*{20003, FALSE, "", "IBM5550 Taiwan"},*/
542 /*{20004, FALSE, "", "TeleText Taiwan"},*/
543 /*{20005, FALSE, "", "Wang Taiwan"},*/
544 /*{20105, FALSE, "", "IA5 IRV International Alphabet No.5"},*/
545 /*{20106, FALSE, "", "IA5 German"},*/
546 /*{20107, FALSE, "", "IA5 Swedish"},*/
547 /*{20108, FALSE, "", "IA5 Norwegian"},*/
548 /*{20127, FALSE, "", "US ASCII"},*/
549 /*{20261, FALSE, "", "T.61"},*/
550 /*{20269, FALSE, "", "ISO 6937 NonSpacing Accent"},*/
551 /*{20273, FALSE, "", "IBM EBCDIC Germany"},*/
552 /*{20277, FALSE, "", "IBM EBCDIC Denmark/Norway"},*/
553 /*{20278, FALSE, "", "IBM EBCDIC Finland/Sweden"},*/
554 /*{20280, FALSE, "", "IBM EBCDIC Italy"},*/
555 /*{20284, FALSE, "", "IBM EBCDIC Latin America/Spain"},*/
556 /*{20285, FALSE, "", "IBM EBCDIC United Kingdom"},*/
557 /*{20290, FALSE, "", "IBM EBCDIC Japanese Katakana Extended"},*/
558 /*{20297, FALSE, "", "IBM EBCDIC France"},*/
559 /*{20420, FALSE, "", "IBM EBCDIC Arabic"},*/
560 /*{20423, FALSE, "", "IBM EBCDIC Greek"},*/
561 /*{20424, FALSE, "", "IBM EBCDIC Hebrew"},*/
562 /*{20833, FALSE, "", "IBM EBCDIC Korean Extended"},*/
563 /*{20838, FALSE, "", "IBM EBCDIC Thai"},*/
564 /*{20871, FALSE, "", "IBM EBCDIC Icelandic"},*/
565 /*{20880, FALSE, "", "IBM EBCDIC Cyrillic (Russian)"},*/
566 {20866, FALSE, LIBDIR"MAPPINGS/VENDORS/MISC/KOI8-R.TXT", "Russian KOI8"},
567 /*{20905, FALSE, "", "IBM EBCDIC Turkish"},*/
568 /*{21025, FALSE, "", "IBM EBCDIC Cyrillic (Serbian, Bulgarian)"},*/
569 /*{21027, FALSE, "", "Ext Alpha Lowercase"},*/
570 {28591, FALSE, LIBDIR"MAPPINGS/ISO8859/8859-1.TXT", "ISO 8859-1 Latin 1"},
571 {28592, FALSE, LIBDIR"MAPPINGS/ISO8859/8859-2.TXT", "ISO 8859-2 Eastern Europe"},
572 {28593, FALSE, LIBDIR"MAPPINGS/ISO8859/8859-3.TXT", "ISO 8859-3 Turkish"},
573 {28594, FALSE, LIBDIR"MAPPINGS/ISO8859/8859-4.TXT", "ISO 8859-4 Baltic"},
574 {28595, FALSE, LIBDIR"MAPPINGS/ISO8859/8859-5.TXT", "ISO 8859-5 Cyrillic"},
575 {28596, FALSE, LIBDIR"MAPPINGS/ISO8859/8859-6.TXT", "ISO 8859-6 Arabic"},
576 {28597, FALSE, LIBDIR"MAPPINGS/ISO8859/8859-7.TXT", "ISO 8859-7 Greek"},
577 {28598, FALSE, LIBDIR"MAPPINGS/ISO8859/8859-8.TXT", "ISO 8859-8 Hebrew"},
578 {28599, FALSE, LIBDIR"MAPPINGS/ISO8859/8859-9.TXT", "ISO 8859-9 Latin 5"}
579 };
580
581 for(i = 0; i < sizeof(pages)/sizeof(pages[0]); i++) {
582 table = Load_CP2Unicode_Table(pages[i].table_filename, pages[i].cp, &cpi);
583 if(!table) {
584 printf("Could not load \"%s\" (%s)\n", pages[i].table_filename, pages[i].comment);
585 continue;
586 }
587
588 if (pages[i].oem) {
589 oemtable = Load_OEM2Unicode_Table(LIBDIR"MAPPINGS/VENDORS/MISC/IBMGRAPH.TXT", table, pages[i].cp, &cpi);
590 if(!oemtable) {
591 printf("Could not load \"%s\" (%s)\n", LIBDIR"MAPPINGS/VENDORS/MISC/IBMGRAPH.TXT", "IBM OEM glyph table");
592 continue;
593 }
594 }
595
596 sprintf(nls_filename, "%s/c_%03d.nls", NLSDIR, cpi.CodePage);
597 if(!create_nls_file(nls_filename, &cpi, table, pages[i].oem ? oemtable : NULL)) {
598 printf("Could not write \"%s\" (%s)\n", nls_filename, pages[i].comment);
599 }
600
601 if (pages[i].oem)
602 free(oemtable);
603
604 free(table);
605 }
606
607 return 0;
608 }
609
610
611
612 static WORD *to_upper_org = NULL, *to_lower_org = NULL;
613
614 #if 0
615 static WORD diffs[256];
616 static int number_of_diffs;
617 #endif
618
619 static WORD number_of_subtables_with_diffs;
620 /* pointers to subtables with 16 elements in each to the main table */
621 static WORD *subtables_with_diffs[4096];
622
623 static WORD number_of_subtables_with_offsets;
624 /* subtables with 16 elements */
625 static WORD subtables_with_offsets[4096 * 16];
626
627 static void test_packed_table(WCHAR *table)
628 {
629 WCHAR test_str[] = L"This is an English text. \x0CF\x0EE-\x0F0\x0F3\x0F1\x0F1\x0EA\x0E8 \x0FF \x0EF\x0E8\x0F1\x0E0\x0F2\x0FC \x0F3\x0EC\x0E5\x0FE \x0ED\x0E5\x0EC\x0ED\x0EE\x0E6\x0EA\x0EE. 1234567890";
630 //WORD diff, off;
631 //WORD *sub_table;
632 DWORD i, len;
633
634 len = lstrlenW(test_str);
635
636 for(i = 0; i < len + 1; i++) {
637 /*off = table[HIBYTE(test_str[i])];
638
639 sub_table = table + off;
640 off = sub_table[LOBYTE(test_str[i]) >> 4];
641
642 sub_table = table + off;
643 off = LOBYTE(test_str[i]) & 0x0F;
644
645 diff = sub_table[off];
646
647 test_str[i] += diff;*/
648 test_str[i] += table[table[table[HIBYTE(test_str[i])] + (LOBYTE(test_str[i]) >> 4)] + (LOBYTE(test_str[i]) & 0x0F)];
649 }
650 /*
651 {
652 FILE *file;
653 static int n = 0;
654 char name[20];
655
656 sprintf(name, "text%02d.dat", n++);
657 file = fopen(name, "wb");
658 fwrite(test_str, len * sizeof(WCHAR), 1, file);
659 fclose(file);
660 }*/
661 }
662
663 static BOOL CreateCaseDiff(char *table_name)
664 {
665 char buf[256];
666 char *p;
667 WORD code, case_mapping;
668 FILE *file;
669 int line;
670
671 to_upper_org = (WORD *)calloc(65536, sizeof(WORD));
672 if(!to_upper_org) {
673 printf("Not enough memory for to upper table\n");
674 return FALSE;
675 }
676
677 to_lower_org = (WORD *)calloc(65536, sizeof(WORD));
678 if(!to_lower_org) {
679 printf("Not enough memory for to lower table\n");
680 return FALSE;
681 }
682
683 file = fopen(table_name, "r");
684 if(file == NULL) {
685 printf("Could not open file \"%s\"\n", table_name);
686 return FALSE;
687 }
688
689 line = 0;
690
691 while(fgets(buf, sizeof(buf), file)) {
692 line++;
693 p = buf;
694 while(*p && isspace(*p)) p++;
695
696 if(!*p)
697 continue;
698
699 /* 0. Code value */
700 code = (WORD)strtol(p, &p, 16);
701
702 //if(code != 0x9A0 && code != 0xBA0)
703 //continue;
704
705 while(*p && *p != ';') p++;
706 if(!*p)
707 continue;
708 p++;
709
710 /* 1. Character name */
711 while(*p && *p != ';') p++;
712 if(!*p)
713 continue;
714 p++;
715
716 /* 2. General Category */
717 while(*p && *p != ';') p++;
718 if(!*p)
719 continue;
720 p++;
721
722 /* 3. Canonical Combining Classes */
723 while(*p && *p != ';') p++;
724 if(!*p)
725 continue;
726 p++;
727
728 /* 4. Bidirectional Category */
729 while(*p && *p != ';') p++;
730 if(!*p)
731 continue;
732 p++;
733
734 /* 5. Character Decomposition Mapping */
735 while(*p && *p != ';') p++;
736 if(!*p)
737 continue;
738 p++;
739
740 /* 6. Decimal digit value */
741 while(*p && *p != ';') p++;
742 if(!*p)
743 continue;
744 p++;
745
746 /* 7. Digit value */
747 while(*p && *p != ';') p++;
748 if(!*p)
749 continue;
750 p++;
751
752 /* 8. Numeric value */
753 while(*p && *p != ';') p++;
754 if(!*p)
755 continue;
756 p++;
757
758 /* 9. Mirrored */
759 while(*p && *p != ';') p++;
760 if(!*p)
761 continue;
762 p++;
763
764 /* 10. Unicode 1.0 Name */
765 while(*p && *p != ';') p++;
766 if(!*p)
767 continue;
768 p++;
769
770 /* 11. 10646 comment field */
771 while(*p && *p != ';') p++;
772 if(!*p)
773 continue;
774 p++;
775
776 /* 12. Uppercase Mapping */
777 while(*p && isspace(*p)) p++;
778 if(!*p) continue;
779 if(*p != ';') {
780 case_mapping = (WORD)strtol(p, &p, 16);
781 to_upper_org[code] = case_mapping - code;
782 while(*p && *p != ';') p++;
783 }
784 else
785 p++;
786
787 /* 13. Lowercase Mapping */
788 while(*p && isspace(*p)) p++;
789 if(!*p) continue;
790 if(*p != ';') {
791 case_mapping = (WORD)strtol(p, &p, 16);
792 to_lower_org[code] = case_mapping - code;
793 while(*p && *p != ';') p++;
794 }
795 else
796 p++;
797
798 /* 14. Titlecase Mapping */
799 while(*p && *p != ';') p++;
800 if(!*p)
801 continue;
802 p++;
803 }
804
805 fclose(file);
806
807 return TRUE;
808 }
809
810 #if 0
811 static int find_diff(WORD diff)
812 {
813 int i;
814
815 for(i = 0; i < number_of_diffs; i++) {
816 if(diffs[i] == diff)
817 return i;
818 }
819
820 return -1;
821 }
822 #endif
823
824 static WORD find_subtable_with_diffs(WORD *table, WORD *subtable)
825 {
826 WORD index;
827
828 for(index = 0; index < number_of_subtables_with_diffs; index++) {
829 if(memcmp(subtables_with_diffs[index], subtable, 16 * sizeof(WORD)) == 0) {
830 return index;
831 }
832 }
833
834 if(number_of_subtables_with_diffs >= 4096) {
835 printf("Could not add new subtable with diffs, storage is full\n");
836 return 0;
837 }
838
839 subtables_with_diffs[number_of_subtables_with_diffs] = subtable;
840 number_of_subtables_with_diffs++;
841
842 return index;
843 }
844
845 static WORD find_subtable_with_offsets(WORD *subtable)
846 {
847 WORD index;
848
849 for(index = 0; index < number_of_subtables_with_offsets; index++) {
850 if(memcmp(&subtables_with_offsets[index * 16], subtable, 16 * sizeof(WORD)) == 0) {
851 return index;
852 }
853 }
854
855 if(number_of_subtables_with_offsets >= 4096) {
856 printf("Could not add new subtable with offsets, storage is full\n");
857 return 0;
858 }
859
860 memcpy(&subtables_with_offsets[number_of_subtables_with_offsets * 16], subtable, 16 * sizeof(WORD));
861 number_of_subtables_with_offsets++;
862
863 return index;
864 }
865
866 static WORD *pack_table(WORD *table, WORD *packed_size_in_words)
867 {
868 WORD high, low4, index;
869 WORD main_index[256];
870 WORD temp_subtable[16];
871 WORD *packed_table;
872 WORD *subtable_src, *subtable_dst;
873
874 memset(subtables_with_diffs, 0, sizeof(subtables_with_diffs));
875 number_of_subtables_with_diffs = 0;
876
877 memset(subtables_with_offsets, 0, sizeof(subtables_with_offsets));
878 number_of_subtables_with_offsets = 0;
879
880 for(high = 0; high < 256; high++) {
881 for(low4 = 0; low4 < 256; low4 += 16) {
882 index = find_subtable_with_diffs(table, &table[MAKEWORD(low4, high)]);
883
884 temp_subtable[low4 >> 4] = index;
885 }
886
887 index = find_subtable_with_offsets(temp_subtable);
888 main_index[high] = index;
889 }
890
891 *packed_size_in_words = 0x100 + number_of_subtables_with_offsets * 16 + number_of_subtables_with_diffs * 16;
892 packed_table = calloc(*packed_size_in_words, sizeof(WORD));
893
894 /* fill main index according to the subtables_with_offsets */
895 for(high = 0; high < 256; high++) {
896 packed_table[high] = 0x100 + main_index[high] * 16;
897 }
898
899 //memcpy(sub_table, subtables_with_offsets, number_of_subtables_with_offsets * 16);
900
901 /* fill subtable index according to the subtables_with_diffs */
902 for(index = 0; index < number_of_subtables_with_offsets; index++) {
903 subtable_dst = packed_table + 0x100 + index * 16;
904 subtable_src = &subtables_with_offsets[index * 16];
905
906 for(low4 = 0; low4 < 16; low4++) {
907 subtable_dst[low4] = 0x100 + number_of_subtables_with_offsets * 16 + subtable_src[low4] * 16;
908 }
909 }
910
911
912 for(index = 0; index < number_of_subtables_with_diffs; index++) {
913 subtable_dst = packed_table + 0x100 + number_of_subtables_with_offsets * 16 + index * 16;
914 memcpy(subtable_dst, subtables_with_diffs[index], 16 * sizeof(WORD));
915
916 }
917
918
919 test_packed_table(packed_table);
920
921 return packed_table;
922 }
923
924 int write_casemap_file(void)
925 {
926 WORD packed_size_in_words, offset_to_next_table_in_words;
927 WORD *packed_table, value;
928 FILE *file;
929
930 if(!CreateCaseDiff(LIBDIR"UnicodeData.txt"))
931 return -1;
932
933 file = fopen(NLSDIR"/l_intl.nls", "wb");
934
935 /* write version number */
936 value = 1;
937 fwrite(&value, 1, sizeof(WORD), file);
938
939 /* pack upper case table */
940 packed_table = pack_table(to_upper_org, &packed_size_in_words);
941 offset_to_next_table_in_words = packed_size_in_words + 1;
942 fwrite(&offset_to_next_table_in_words, 1, sizeof(WORD), file);
943 /* write packed upper case table */
944 fwrite(packed_table, sizeof(WORD), packed_size_in_words, file);
945 free(packed_table);
946
947 /* pack lower case table */
948 packed_table = pack_table(to_lower_org, &packed_size_in_words);
949 offset_to_next_table_in_words = packed_size_in_words + 1;
950 fwrite(&offset_to_next_table_in_words, 1, sizeof(WORD), file);
951 /* write packed lower case table */
952 fwrite(packed_table, sizeof(WORD), packed_size_in_words, file);
953 free(packed_table);
954
955 fclose(file);
956
957 free(to_upper_org);
958 free(to_lower_org);
959
960 return 0;
961 }
962
963 int main()
964 {
965 write_nls_files();
966 write_casemap_file();
967
968 return 0;
969 }