Prefer https URLs where possible.
[libiconv.git] / tools / 8bit_tab_to_h.c
blob94d9f393cb95f7354c8b93d5d4832b4d82c44e5d
1 /* Copyright (C) 1999-2002, 2011-2012, 2016, 2018 Free Software Foundation, Inc.
2 This file is part of the GNU LIBICONV Tools.
4 This program is free software: you can redistribute it and/or modify
5 it under the terms of the GNU General Public License as published by
6 the Free Software Foundation; either version 3 of the License, or
7 (at your option) any later version.
9 This program is distributed in the hope that it will be useful,
10 but WITHOUT ANY WARRANTY; without even the implied warranty of
11 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
12 GNU General Public License for more details.
14 You should have received a copy of the GNU General Public License
15 along with this program; if not, see <https://www.gnu.org/licenses/>. */
18 * Generates an 8-bit character set table from a .TXT table as found on
19 * ftp.unicode.org or from a table containing the 256 Unicode values as
20 * hexadecimal integers.
21 * Examples:
23 * ./8bit_tab_to_h ISO-8859-1 iso8859_1 < tab8859_1
24 * ./8bit_tab_to_h ISO-8859-2 iso8859_2 < tab8859_2
25 * ./8bit_tab_to_h ISO-8859-3 iso8859_3 < tab8859_3
26 * ./8bit_tab_to_h ISO-8859-4 iso8859_4 < tab8859_4
27 * ./8bit_tab_to_h ISO-8859-5 iso8859_5 < tab8859_5
28 * ./8bit_tab_to_h ISO-8859-6 iso8859_6 < tab8859_6
29 * ./8bit_tab_to_h ISO-8859-7 iso8859_7 < tab8859_7
30 * ./8bit_tab_to_h ISO-8859-8 iso8859_8 < tab8859_8
31 * ./8bit_tab_to_h ISO-8859-9 iso8859_9 < tab8859_9
32 * ./8bit_tab_to_h ISO-8859-10 iso8859_10 < tab8859_10
33 * ./8bit_tab_to_h ISO-8859-14 iso8859_14 < tab8859_14
34 * ./8bit_tab_to_h ISO-8859-15 iso8859_15 < tab8859_15
35 * ./8bit_tab_to_h JISX0201.1976-0 jisx0201 < jis0201
36 * ./8bit_tab_to_h TIS620.2533-1 tis620 < tabtis620
37 * ./8bit_tab_to_h KOI8-R koi8_r < tabkoi8_r
38 * ./8bit_tab_to_h KOI8-U koi8_u < tabkoi8_u
39 * ./8bit_tab_to_h ARMSCII-8 armscii_8 < tabarmscii_8
40 * ./8bit_tab_to_h CP1133 cp1133 < tabibm_cp1133
41 * ./8bit_tab_to_h MULELAO-1 mulelao < tabmulelao_1
42 * ./8bit_tab_to_h VISCII1.1-1 viscii1 < tabviscii
43 * ./8bit_tab_to_h TCVN-5712 tcvn < tabtcvn
44 * ./8bit_tab_to_h GEORGIAN-ACADEMY georgian_ac < tabgeorgian_academy
45 * ./8bit_tab_to_h GEORGIAN-PS georgian_ps < tabgeorgian_ps
47 * ./8bit_tab_to_h ISO-8859-1 iso8859_1 < 8859-1.TXT
48 * ./8bit_tab_to_h ISO-8859-2 iso8859_2 < 8859-2.TXT
49 * ./8bit_tab_to_h ISO-8859-3 iso8859_3 < 8859-3.TXT
50 * ./8bit_tab_to_h ISO-8859-4 iso8859_4 < 8859-4.TXT
51 * ./8bit_tab_to_h ISO-8859-5 iso8859_5 < 8859-5.TXT
52 * ./8bit_tab_to_h ISO-8859-6 iso8859_6 < 8859-6.TXT
53 * ./8bit_tab_to_h ISO-8859-7 iso8859_7 < 8859-7.TXT
54 * ./8bit_tab_to_h ISO-8859-8 iso8859_8 < 8859-8.TXT
55 * ./8bit_tab_to_h ISO-8859-9 iso8859_9 < 8859-9.TXT
56 * ./8bit_tab_to_h ISO-8859-10 iso8859_10 < 8859-10.TXT
57 * ./8bit_tab_to_h ISO-8859-14 iso8859_14 < 8859-14.TXT
58 * ./8bit_tab_to_h ISO-8859-15 iso8859_15 < 8859-15.TXT
59 * ./8bit_tab_to_h JISX0201.1976-0 jisx0201 < JIS0201.TXT
60 * ./8bit_tab_to_h KOI8-R koi8_r < KOI8-R.TXT
62 * ./8bit_tab_to_h 'CP50221 JISX0208 extensions' cp50221_0208_ext < CP50221-0208-EXT.TXT
63 * ./8bit_tab_to_h 'CP50221 JISX0212 extensions' cp50221_0212_ext < CP50221-0212-EXT.TXT
66 #include <stdio.h>
67 #include <stdlib.h>
68 #include <stdbool.h>
69 #include <string.h>
71 int main (int argc, char *argv[])
73 const char* charsetname;
74 const char* c_charsetname;
75 const char* filename;
76 const char* directory;
77 int charset2uni[0x100];
79 if (argc != 3 && argc != 4 && argc != 5)
80 exit(1);
81 charsetname = argv[1];
82 c_charsetname = argv[2];
83 if (argc > 3) {
84 filename = argv[3];
85 } else {
86 char* s = (char*) malloc(strlen(c_charsetname)+strlen(".h")+1);
87 strcpy(s,c_charsetname); strcat(s,".h");
88 filename = s;
90 directory = (argc > 4 ? argv[4] : "");
92 fprintf(stderr, "Creating %s%s\n", directory, filename);
95 int i, c;
96 c = getc(stdin);
97 ungetc(c,stdin);
98 if (c == '#') {
99 /* Read a unicode.org style .TXT file. */
100 for (i = 0; i < 0x100; i++)
101 charset2uni[i] = 0xfffd;
102 for (;;) {
103 c = getc(stdin);
104 if (c == EOF)
105 break;
106 if (c == '\n' || c == ' ' || c == '\t')
107 continue;
108 if (c == '#') {
109 do { c = getc(stdin); } while (!(c == EOF || c == '\n'));
110 continue;
112 ungetc(c,stdin);
113 if (scanf("0x%x", &i) != 1 || !(i >= 0 && i < 0x100))
114 exit(1);
115 do { c = getc(stdin); } while (c == ' ' || c == '\t');
116 if (c != EOF)
117 ungetc(c,stdin);
118 if (c == '\n' || c == '#')
119 continue;
120 if (scanf("0x%x", &charset2uni[i]) != 1)
121 exit(1);
123 } else {
124 /* Read a table of hexadecimal Unicode values. */
125 for (i = 0; i < 0x100; i++) {
126 if (scanf("%x", &charset2uni[i]) != 1)
127 exit(1);
128 if (charset2uni[i] < 0 || charset2uni[i] == 0xffff)
129 charset2uni[i] = 0xfffd;
131 if (scanf("%x", &i) != EOF)
132 exit(1);
136 /* Write the output file. */
138 FILE* f;
141 char* fname = malloc(strlen(directory)+strlen(filename)+1);
142 strcpy(fname,directory); strcat(fname,filename);
143 f = fopen(fname,"w");
144 if (f == NULL)
145 exit(1);
148 fprintf(f, "/*\n");
149 fprintf(f, " * Copyright (C) 1999-2002 Free Software Foundation, Inc.\n");
150 fprintf(f, " * This file is part of the GNU LIBICONV Library.\n");
151 fprintf(f, " *\n");
152 fprintf(f, " * The GNU LIBICONV Library is free software; you can redistribute it\n");
153 fprintf(f, " * and/or modify it under the terms of the GNU Library General Public\n");
154 fprintf(f, " * License as published by the Free Software Foundation; either version 2\n");
155 fprintf(f, " * of the License, or (at your option) any later version.\n");
156 fprintf(f, " *\n");
157 fprintf(f, " * The GNU LIBICONV Library is distributed in the hope that it will be\n");
158 fprintf(f, " * useful, but WITHOUT ANY WARRANTY; without even the implied warranty of\n");
159 fprintf(f, " * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU\n");
160 fprintf(f, " * Library General Public License for more details.\n");
161 fprintf(f, " *\n");
162 fprintf(f, " * You should have received a copy of the GNU Library General Public\n");
163 fprintf(f, " * License along with the GNU LIBICONV Library; see the file COPYING.LIB.\n");
164 fprintf(f, " * If not, see <https://www.gnu.org/licenses/>.\n");
165 fprintf(f, " */\n");
166 fprintf(f, "\n");
167 fprintf(f, "/*\n");
168 fprintf(f, " * %s\n", charsetname);
169 fprintf(f, " */\n");
170 fprintf(f, "\n");
173 int i, i1, i2, i3;
174 int line[16];
175 int tableno;
176 struct { int minline; int maxline; } tables[16];
177 bool some_invalid;
178 bool final_ret_reached;
180 for (i1 = 0; i1 < 16; i1++) {
181 bool all_invalid = true;
182 bool all_identity = true;
183 for (i2 = 0; i2 < 16; i2++) {
184 i = 16*i1+i2;
185 if (charset2uni[i] != 0xfffd)
186 all_invalid = false;
187 if (charset2uni[i] != i)
188 all_identity = false;
190 if (all_invalid)
191 line[i1] = -2;
192 else if (all_identity)
193 line[i1] = -1;
194 else
195 line[i1] = 0;
197 tableno = 0;
198 for (i1 = 0; i1 < 16; i1++) {
199 if (line[i1] >= 0) {
200 if (i1 > 0 && tableno > 0 && line[i1-1] == tableno-1) {
201 line[i1] = tableno-1;
202 tables[tableno-1].maxline = i1;
203 } else {
204 tableno++;
205 line[i1] = tableno-1;
206 tables[tableno-1].minline = tables[tableno-1].maxline = i1;
210 some_invalid = false;
211 for (i = 0; i < 0x100; i++)
212 if (charset2uni[i] == 0xfffd)
213 some_invalid = true;
214 if (tableno > 0) {
215 int t;
216 for (t = 0; t < tableno; t++) {
217 fprintf(f, "static const unsigned short %s_2uni", c_charsetname);
218 if (tableno > 1)
219 fprintf(f, "_%d", t+1);
220 fprintf(f, "[%d] = {\n", 16*(tables[t].maxline-tables[t].minline+1));
221 for (i1 = tables[t].minline; i1 <= tables[t].maxline; i1++) {
222 fprintf(f, " /* 0x%02x */\n", 16*i1);
223 for (i2 = 0; i2 < 2; i2++) {
224 fprintf(f, " ");
225 for (i3 = 0; i3 < 8; i3++) {
226 i = 16*i1+8*i2+i3;
227 fprintf(f, " 0x%04x,", charset2uni[i]);
229 fprintf(f, "\n");
232 fprintf(f, "};\n");
234 fprintf(f, "\n");
236 final_ret_reached = false;
237 fprintf(f, "static int\n%s_mbtowc (conv_t conv, ucs4_t *pwc, const unsigned char *s, size_t n)\n", c_charsetname);
238 fprintf(f, "{\n");
239 fprintf(f, " unsigned char c = *s;\n");
240 if (some_invalid) {
241 for (i1 = 0; i1 < 16;) {
242 int t = line[i1];
243 const char* indent;
244 for (i2 = i1; i2 < 16 && line[i2] == t; i2++);
245 indent = (i1 == 0 && i2 == 16 ? " " : " ");
246 if (i1 == 0) {
247 if (i2 == 16) {
248 } else {
249 fprintf(f, " if (c < 0x%02x) {\n", 16*i2);
251 } else {
252 if (i2 == 16) {
253 fprintf(f, " else {\n");
254 } else {
255 fprintf(f, " else if (c < 0x%02x) {\n", 16*i2);
258 if (t == -2) {
259 final_ret_reached = true;
260 } else if (t == -1) {
261 fprintf(f, "%s*pwc = (ucs4_t) c;\n", indent);
262 fprintf(f, "%sreturn 1;\n", indent);
263 } else {
264 fprintf(f, "%s", indent);
265 some_invalid = false;
266 for (i = 16*i1; i < 16*i2; i++)
267 if (charset2uni[i] == 0xfffd)
268 some_invalid = true;
269 if (some_invalid)
270 fprintf(f, "unsigned short wc = ");
271 else
272 fprintf(f, "*pwc = (ucs4_t) ");
273 fprintf(f, "%s_2uni", c_charsetname);
274 if (tableno > 1)
275 fprintf(f, "_%d", t+1);
276 fprintf(f, "[c");
277 if (tables[t].minline > 0)
278 fprintf(f, "-0x%02x", 16*tables[t].minline);
279 fprintf(f, "];\n");
280 if (some_invalid) {
281 fprintf(f, "%sif (wc != 0xfffd) {\n", indent);
282 fprintf(f, "%s *pwc = (ucs4_t) wc;\n", indent);
283 fprintf(f, "%s return 1;\n", indent);
284 fprintf(f, "%s}\n", indent);
285 final_ret_reached = true;
286 } else {
287 fprintf(f, "%sreturn 1;\n", indent);
290 if (!(i1 == 0 && i2 == 16))
291 fprintf(f, " }\n");
292 i1 = i2;
294 if (final_ret_reached)
295 fprintf(f, " return RET_ILSEQ;\n");
296 } else {
297 for (i1 = 0; i1 < 16;) {
298 int t = line[i1];
299 for (i2 = i1; i2 < 16 && line[i2] == t; i2++);
300 if (i1 == 0) {
301 if (i2 == 16) {
302 fprintf(f, " ");
303 } else {
304 fprintf(f, " if (c < 0x%02x)\n ", 16*i2);
306 } else {
307 if (i2 == 16) {
308 fprintf(f, " else\n ");
309 } else {
310 fprintf(f, " else if (c < 0x%02x)\n ", 16*i2);
313 if (t == -1)
314 fprintf(f, "*pwc = (ucs4_t) c;\n");
315 else {
316 fprintf(f, "*pwc = (ucs4_t) %s_2uni", c_charsetname);
317 if (tableno > 1)
318 fprintf(f, "_%d", t+1);
319 fprintf(f, "[c");
320 if (tables[t].minline > 0)
321 fprintf(f, "-0x%02x", 16*tables[t].minline);
322 fprintf(f, "];\n");
324 i1 = i2;
326 fprintf(f, " return 1;\n");
328 fprintf(f, "}\n");
332 fprintf(f, "\n");
335 int uni2charset[0x10000];
336 bool pages[0x100];
337 int line[0x2000];
338 int tableno;
339 struct { int minline; int maxline; int usecount; const char* suffix; } tables[0x2000];
340 bool need_c;
341 bool fix_0000;
342 int i, j, p, j1, j2, t;
344 for (j = 0; j < 0x10000; j++)
345 uni2charset[j] = 0;
346 for (p = 0; p < 0x100; p++)
347 pages[p] = false;
348 for (i = 0; i < 0x100; i++) {
349 j = charset2uni[i];
350 if (j != 0xfffd) {
351 uni2charset[j] = i;
352 pages[j>>8] = true;
355 for (j1 = 0; j1 < 0x2000; j1++) {
356 bool all_invalid = true;
357 bool all_identity = true;
358 for (j2 = 0; j2 < 8; j2++) {
359 j = 8*j1+j2;
360 if (uni2charset[j] != 0)
361 all_invalid = false;
362 if (uni2charset[j] != j)
363 all_identity = false;
365 if (all_invalid)
366 line[j1] = -2;
367 else if (all_identity)
368 line[j1] = -1;
369 else
370 line[j1] = 0;
372 tableno = 0;
373 for (j1 = 0; j1 < 0x2000; j1++) {
374 if (line[j1] >= 0) {
375 if (tableno > 0
376 && ((j1 > 0 && line[j1-1] == tableno-1)
377 || ((tables[tableno-1].maxline >> 5) == (j1 >> 5)
378 && j1 - tables[tableno-1].maxline <= 8))) {
379 line[j1] = tableno-1;
380 tables[tableno-1].maxline = j1;
381 } else {
382 tableno++;
383 line[j1] = tableno-1;
384 tables[tableno-1].minline = tables[tableno-1].maxline = j1;
388 for (t = 0; t < tableno; t++) {
389 tables[t].usecount = 0;
390 j1 = 8*tables[t].minline;
391 j2 = 8*(tables[t].maxline+1);
392 for (j = j1; j < j2; j++)
393 if (uni2charset[j] != 0)
394 tables[t].usecount++;
396 for (t = 0, p = -1, i = 0; t < tableno; t++) {
397 if (tables[t].usecount > 1) {
398 char* s;
399 if (p == tables[t].minline >> 5) {
400 s = (char*) malloc(5+1);
401 sprintf(s, "%02x_%d", p, ++i);
402 } else {
403 p = tables[t].minline >> 5;
404 s = (char*) malloc(2+1);
405 sprintf(s, "%02x", p);
407 tables[t].suffix = s;
408 } else
409 tables[t].suffix = NULL;
412 p = -1;
413 for (t = 0; t < tableno; t++)
414 if (tables[t].usecount > 1) {
415 p = 0;
416 fprintf(f, "static const unsigned char %s_page%s[%d] = {\n", c_charsetname, tables[t].suffix, 8*(tables[t].maxline-tables[t].minline+1));
417 for (j1 = tables[t].minline; j1 <= tables[t].maxline; j1++) {
418 if ((j1 % 0x20) == 0 && j1 > tables[t].minline)
419 fprintf(f, " /* 0x%04x */\n", 8*j1);
420 fprintf(f, " ");
421 for (j2 = 0; j2 < 8; j2++) {
422 j = 8*j1+j2;
423 fprintf(f, " 0x%02x,", uni2charset[j]);
425 fprintf(f, " /* 0x%02x-0x%02x */\n", 8*(j1 % 0x20), 8*(j1 % 0x20)+7);
427 fprintf(f, "};\n");
429 if (p >= 0)
430 fprintf(f, "\n");
432 need_c = false;
433 for (j1 = 0; j1 < 0x2000;) {
434 t = line[j1];
435 for (j2 = j1; j2 < 0x2000 && line[j2] == t; j2++);
436 if (t >= 0)
437 j2 = tables[t].maxline+1;
438 if (!(t == -2 || (t == -1 && j1 == 0)))
439 need_c = true;
440 j1 = j2;
442 fix_0000 = false;
443 fprintf(f, "static int\n%s_wctomb (conv_t conv, unsigned char *r, ucs4_t wc, size_t n)\n", c_charsetname);
444 fprintf(f, "{\n");
445 if (need_c)
446 fprintf(f, " unsigned char c = 0;\n");
447 for (j1 = 0; j1 < 0x2000;) {
448 t = line[j1];
449 for (j2 = j1; j2 < 0x2000 && line[j2] == t; j2++);
450 if (t >= 0) {
451 if (j1 != tables[t].minline) abort();
452 if (j2 > tables[t].maxline+1) abort();
453 j2 = tables[t].maxline+1;
455 if (t == -2) {
456 } else {
457 if (j1 == 0)
458 fprintf(f, " ");
459 else
460 fprintf(f, " else ");
461 if (t >= 0 && tables[t].usecount == 0) abort();
462 if (t >= 0 && tables[t].usecount == 1) {
463 if (j2 != j1+1) abort();
464 for (j = 8*j1; j < 8*j2; j++)
465 if (uni2charset[j] != 0) {
466 fprintf(f, "if (wc == 0x%04x)\n c = 0x%02x;\n", j, uni2charset[j]);
467 break;
469 } else {
470 if (j1 == 0) {
471 fprintf(f, "if (wc < 0x%04x)", 8*j2);
472 } else {
473 fprintf(f, "if (wc >= 0x%04x && wc < 0x%04x)", 8*j1, 8*j2);
475 if (t == -1) {
476 if (j1 == 0)
477 /* If wc == 0, the function must return 1, not -1. */
478 fprintf(f, " {\n *r = wc;\n return 1;\n }\n");
479 else
480 fprintf(f, "\n c = wc;\n");
481 } else {
482 fprintf(f, "\n c = %s_page%s[wc", c_charsetname, tables[t].suffix);
483 if (tables[t].minline > 0)
484 fprintf(f, "-0x%04x", 8*j1);
485 fprintf(f, "];\n");
486 if (j1 == 0 && uni2charset[0] == 0)
487 /* If wc == 0, the function must return 1, not -1. */
488 fix_0000 = true;
492 j1 = j2;
494 if (need_c) {
495 if (fix_0000)
496 fprintf(f, " if (c != 0 || wc == 0) {\n");
497 else
498 fprintf(f, " if (c != 0) {\n");
499 fprintf(f, " *r = c;\n");
500 fprintf(f, " return 1;\n");
501 fprintf(f, " }\n");
503 fprintf(f, " return RET_ILUNI;\n");
504 fprintf(f, "}\n");
508 if (ferror(f) || fclose(f))
509 exit(1);
512 #if 0
514 int i1, i2, i3, i1_min, i1_max, j1, j2;
516 i1_min = 16;
517 i1_max = -1;
518 for (i1 = 0; i1 < 16; i1++)
519 for (i2 = 0; i2 < 16; i2++)
520 if (charset2uni[16*i1+i2] != 0xfffd) {
521 if (i1_min > i1) i1_min = i1;
522 if (i1_max < i1) i1_max = i1;
524 printf("static const unsigned short %s_2uni[%d] = {\n",
525 name, 16*(i1_max-i1_min+1));
526 for (i1 = i1_min; i1 <= i1_max; i1++) {
527 printf(" /""* 0x%02x *""/\n", 16*i1);
528 for (i2 = 0; i2 < 2; i2++) {
529 printf(" ");
530 for (i3 = 0; i3 < 8; i3++) {
531 if (i3 > 0) printf(" ");
532 printf("0x%04x,", charset2uni[16*i1+8*i2+i3]);
534 printf("\n");
537 printf("};\n");
538 printf("\n");
540 for (p = 0; p < 0x100; p++)
541 pages[p] = 0;
542 for (i = 0; i < 0x100; i++)
543 if (charset2uni[i] != 0xfffd)
544 pages[charset2uni[i]>>8] = 1;
545 for (p = 0; p < 0x100; p++)
546 if (pages[p]) {
547 int j1_min = 32;
548 int j1_max = -1;
549 for (j1 = 0; j1 < 32; j1++)
550 for (j2 = 0; j2 < 8; j2++)
551 if (uni2charset[256*p+8*j1+j2] != 0) {
552 if (j1_min > j1) j1_min = j1;
553 if (j1_max < j1) j1_max = j1;
555 printf("static const unsigned char %s_page%02x[%d] = {\n",
556 name, p, 8*(j1_max-j1_min+1));
557 for (j1 = j1_min; j1 <= j1_max; j1++) {
558 printf(" ");
559 for (j2 = 0; j2 < 8; j2++)
560 printf("0x%02x, ", uni2charset[256*p+8*j1+j2]);
561 printf("/""* 0x%02x-0x%02x *""/\n", 8*j1, 8*j1+7);
563 printf("};\n");
565 printf("\n");
568 #endif
570 exit(0);