~mdw / sgt / charset / blob
? search:
summary | shortlog | log | commit | commitdiff | tree
blame | history | raw | HEAD
Make read_utf8(), like read_sbcs(), accessible to the rest of the library,
[sgt/charset] / utf8.c
1 /*
2 * utf8.c - routines to handle UTF-8.
3 */
4
5 #ifndef ENUM_CHARSETS
6
7 #include "charset.h"
8 #include "internal.h"
9
10 /*
11 * UTF-8 has no associated data, so `charset' may be ignored.
12 */
13
14 void read_utf8(charset_spec const *charset, long int input_chr,
15 charset_state *state,
16 void (*emit)(void *ctx, long int output), void *emitctx)
17 {
18 UNUSEDARG(charset);
19
20 /*
21 * For reading UTF-8, the `state' word contains the character
22 * being accumulated. This is shifted left by six bits each
23 * time a character is added, and there's a single '1' bit
24 * in what would be bit 31 of the final character, which we
25 * use to detect when it's complete.
26 *
27 * As required, the state is zero when we are not in the middle
28 * of a multibyte character at all.
29 *
30 * For example, when reading E9 8D 8B, starting at state=0:
31 *
32 * - after E9, the state is 0x00080009
33 * - after 8D, the state is 0x0200024d
34 * - after 8B, the state conceptually becomes 0x8000934b, at
35 * which point we notice we've got as many characters as we
36 * were expecting, output U+934B, and reset the state to
37 * zero.
38 *
39 * If we detect an overlong sequence, we shift the marker bit
40 * right one bit. This is safe because an overlong sequence
41 * can't encode a top-bit-set character. Not that we worry
42 * about what overlong sequences are trying to encode, but
43 * it's nice to know that we could if we wanted to.
44 *
45 * Note that the maximum number of bits we might need to store
46 * in the character value field is 25 (U+7FFFFFFF contains 31
47 * bits, but we will never actually store its full value
48 * because when we receive the last 6 bits in the final
49 * continuation byte we will output it and revert the state to
50 * zero). Hence we need 26 bits in total.
51 */
52
53 if (input_chr < 0x80) {
54 /*
55 * Single-byte character. If the state is nonzero before
56 * coming here, output an error for an incomplete sequence.
57 * Then output the character.
58 */
59 if (state->s0 != 0) {
60 emit(emitctx, ERROR);
61 state->s0 = 0;
62 }
63 emit(emitctx, input_chr);
64 } else if (input_chr == 0xFE || input_chr == 0xFF) {
65 /*
66 * FE and FF bytes should _never_ occur in UTF-8. They are
67 * automatic errors; if the state was nonzero to start
68 * with, output a further error for an incomplete sequence.
69 */
70 if (state->s0 != 0) {
71 emit(emitctx, ERROR);
72 state->s0 = 0;
73 }
74 emit(emitctx, ERROR);
75 } else if (input_chr >= 0x80 && input_chr < 0xC0) {
76 /*
77 * Continuation byte. Output an error for an unexpected
78 * continuation byte, if the state is zero.
79 */
80 if (state->s0 == 0) {
81 emit(emitctx, ERROR);
82 } else {
83 unsigned long charval;
84
85 /*
86 * Otherwise, accumulate more of the character value.
87 */
88 charval = state->s0;
89 charval = (charval << 6) | (input_chr & 0x3F);
90
91 /*
92 * Detect overlong encodings. We're looking for too many
93 * leading zeroes given our position in the character. If
94 * we find an overlong encoding, clear the current marker
95 * bit and set the bit below it. Overlong two-byte
96 * encodings are a special case, and are detected when we
97 * read their inital byte.
98 */
99 if ((charval & 0xffffffe0L) == 0x02000000L)
100 charval ^= 0x03000000L;
101 else if ((charval & 0xfffffff0L) == 0x00080000L)
102 charval ^= 0x000c0000L;
103 else if ((charval & 0xfffffff8L) == 0x00002000L)
104 charval ^= 0x00003000L;
105 else if ((charval & 0xfffffffcL) == 0x00000080L)
106 charval ^= 0x000000c0L;
107
108 /*
109 * Check the byte counts; if we have not reached the
110 * end of the character, update the state and return.
111 */
112 if (!(charval & 0xc0000000L)) {
113 state->s0 = charval;
114 return;
115 }
116
117 /*
118 * Clear the marker bit, or set it if it's clear,
119 * indicating an overlong sequence.
120 */
121 charval ^= 0x80000000L;
122
123 /*
124 * Now we know we've reached the end of the character.
125 * `charval' is the Unicode value. We should check for
126 * various invalid things, and then either output
127 * charval or an error. In all cases we reset the state
128 * to zero.
129 */
130 state->s0 = 0;
131
132 if (charval & 0x80000000L) {
133 /* We got an overlong sequence. */
134 emit(emitctx, ERROR);
135 } else if (charval >= 0xD800 && charval < 0xE000) {
136 /*
137 * Surrogates (0xD800-0xDFFF) may never be encoded
138 * in UTF-8. A surrogate pair in Unicode should
139 * have been encoded as a single UTF-8 character
140 * occupying more than three bytes.
141 */
142 emit(emitctx, ERROR);
143 } else if (charval == 0xFFFE || charval == 0xFFFF) {
144 /*
145 * U+FFFE and U+FFFF are invalid Unicode characters
146 * and may never be encoded in UTF-8. (This is one
147 * reason why U+FFFF is our way of signalling an
148 * error to our `emit' function :-)
149 */
150 emit(emitctx, ERROR);
151 } else {
152 /*
153 * Oh, all right. We'll let this one off.
154 */
155 emit(emitctx, charval);
156 }
157 }
158
159 } else {
160 /*
161 * Lead byte. First output an error for an incomplete
162 * sequence, if the state is nonzero.
163 */
164 if (state->s0 != 0)
165 emit(emitctx, ERROR);
166
167 /*
168 * Now deal with the lead byte: work out the number of
169 * bytes we expect to see in this character, and extract
170 * the initial bits of it too.
171 */
172 if (input_chr >= 0xC0 && input_chr < 0xC2) {
173 /* beginning of an overlong two-byte sequence */
174 state->s0 = 0x01000000L | (input_chr & 0x1F);
175 } else if (input_chr >= 0xC2 && input_chr < 0xE0) {
176 state->s0 = 0x02000000L | (input_chr & 0x1F);
177 } else if (input_chr >= 0xE0 && input_chr < 0xF0) {
178 state->s0 = 0x00080000L | (input_chr & 0x0F);
179 } else if (input_chr >= 0xF0 && input_chr < 0xF8) {
180 state->s0 = 0x00002000L | (input_chr & 0x07);
181 } else if (input_chr >= 0xF8 && input_chr < 0xFC) {
182 state->s0 = 0x00000080L | (input_chr & 0x03);
183 } else if (input_chr >= 0xFC && input_chr < 0xFE) {
184 state->s0 = 0x00000002L | (input_chr & 0x01);
185 }
186 }
187 }
188
189 /*
190 * UTF-8 is a stateless multi-byte encoding (in the sense that just
191 * after any character has been completed, the state is always the
192 * same); hence when writing it, there is no need to use the
193 * charset_state.
194 */
195
196 static int write_utf8(charset_spec const *charset, long int input_chr,
197 charset_state *state,
198 void (*emit)(void *ctx, long int output),
199 void *emitctx)
200 {
201 UNUSEDARG(charset);
202 UNUSEDARG(state);
203
204 if (input_chr == -1)
205 return TRUE; /* stateless; no cleanup required */
206
207 /*
208 * Refuse to output any illegal code points.
209 */
210 if (input_chr == 0xFFFE || input_chr == 0xFFFF ||
211 (input_chr >= 0xD800 && input_chr < 0xE000)) {
212 return FALSE;
213 } else if (input_chr < 0x80) { /* one-byte character */
214 emit(emitctx, input_chr);
215 return TRUE;
216 } else if (input_chr < 0x800) { /* two-byte character */
217 emit(emitctx, 0xC0 | (0x1F & (input_chr >> 6)));
218 emit(emitctx, 0x80 | (0x3F & (input_chr )));
219 return TRUE;
220 } else if (input_chr < 0x10000) { /* three-byte character */
221 emit(emitctx, 0xE0 | (0x0F & (input_chr >> 12)));
222 emit(emitctx, 0x80 | (0x3F & (input_chr >> 6)));
223 emit(emitctx, 0x80 | (0x3F & (input_chr )));
224 return TRUE;
225 } else if (input_chr < 0x200000) { /* four-byte character */
226 emit(emitctx, 0xF0 | (0x07 & (input_chr >> 18)));
227 emit(emitctx, 0x80 | (0x3F & (input_chr >> 12)));
228 emit(emitctx, 0x80 | (0x3F & (input_chr >> 6)));
229 emit(emitctx, 0x80 | (0x3F & (input_chr )));
230 return TRUE;
231 } else if (input_chr < 0x4000000) {/* five-byte character */
232 emit(emitctx, 0xF8 | (0x03 & (input_chr >> 24)));
233 emit(emitctx, 0x80 | (0x3F & (input_chr >> 18)));
234 emit(emitctx, 0x80 | (0x3F & (input_chr >> 12)));
235 emit(emitctx, 0x80 | (0x3F & (input_chr >> 6)));
236 emit(emitctx, 0x80 | (0x3F & (input_chr )));
237 return TRUE;
238 } else { /* six-byte character */
239 emit(emitctx, 0xFC | (0x01 & (input_chr >> 30)));
240 emit(emitctx, 0x80 | (0x3F & (input_chr >> 24)));
241 emit(emitctx, 0x80 | (0x3F & (input_chr >> 18)));
242 emit(emitctx, 0x80 | (0x3F & (input_chr >> 12)));
243 emit(emitctx, 0x80 | (0x3F & (input_chr >> 6)));
244 emit(emitctx, 0x80 | (0x3F & (input_chr )));
245 return TRUE;
246 }
247 }
248
249 #ifdef TESTMODE
250
251 #include <stdio.h>
252 #include <stdarg.h>
253
254 int total_errs = 0;
255
256 void utf8_emit(void *ctx, long output)
257 {
258 wchar_t **p = (wchar_t **)ctx;
259 *(*p)++ = output;
260 }
261
262 void utf8_read_test(int line, char *input, int inlen, ...)
263 {
264 va_list ap;
265 wchar_t *p, str[512];
266 int i;
267 charset_state state;
268 unsigned long l;
269
270 state.s0 = 0;
271 p = str;
272
273 for (i = 0; i < inlen; i++)
274 read_utf8(NULL, input[i] & 0xFF, &state, utf8_emit, &p);
275
276 va_start(ap, inlen);
277 l = 0;
278 for (i = 0; i < p - str; i++) {
279 l = va_arg(ap, long int);
280 if (l == -1) {
281 printf("%d: correct string shorter than output\n", line);
282 total_errs++;
283 break;
284 }
285 if (l != str[i]) {
286 printf("%d: char %d came out as %08x, should be %08x\n",
287 line, i, str[i], l);
288 total_errs++;
289 }
290 }
291 if (l != -1) {
292 l = va_arg(ap, long int);
293 if (l != -1) {
294 printf("%d: correct string longer than output\n", line);
295 total_errs++;
296 }
297 }
298 va_end(ap);
299 }
300
301 void utf8_write_test(int line, const long *input, int inlen, ...)
302 {
303 va_list ap;
304 wchar_t *p, str[512];
305 int i;
306 charset_state state;
307 unsigned long l;
308
309 state.s0 = 0;
310 p = str;
311
312 for (i = 0; i < inlen; i++) {
313 if (!write_utf8(NULL, input[i], &state, utf8_emit, &p))
314 utf8_emit(&p, ERROR);
315 }
316
317 va_start(ap, inlen);
318 l = 0;
319 for (i = 0; i < p - str; i++) {
320 l = va_arg(ap, long int);
321 if (l == -1) {
322 printf("%d: correct string shorter than output\n", line);
323 total_errs++;
324 break;
325 }
326 if (l != str[i]) {
327 printf("%d: char %d came out as %08x, should be %08x\n",
328 line, i, str[i], l);
329 total_errs++;
330 }
331 }
332 if (l != -1) {
333 l = va_arg(ap, long int);
334 if (l != -1) {
335 printf("%d: correct string longer than output\n", line);
336 total_errs++;
337 }
338 }
339 va_end(ap);
340 }
341
342 /* Macro to concoct the first three parameters of utf8_read_test. */
343 #define TESTSTR(x) __LINE__, x, lenof(x)
344
345 int main(void)
346 {
347 printf("read tests beginning\n");
348 utf8_read_test(TESTSTR("\xCE\xBA\xE1\xBD\xB9\xCF\x83\xCE\xBC\xCE\xB5"),
349 0x000003BA, /* GREEK SMALL LETTER KAPPA */
350 0x00001F79, /* GREEK SMALL LETTER OMICRON WITH OXIA */
351 0x000003C3, /* GREEK SMALL LETTER SIGMA */
352 0x000003BC, /* GREEK SMALL LETTER MU */
353 0x000003B5, /* GREEK SMALL LETTER EPSILON */
354 0, -1);
355 utf8_read_test(TESTSTR("\x00"),
356 0x00000000, /* <control> */
357 0, -1);
358 utf8_read_test(TESTSTR("\xC2\x80"),
359 0x00000080, /* <control> */
360 0, -1);
361 utf8_read_test(TESTSTR("\xE0\xA0\x80"),
362 0x00000800, /* <no name available> */
363 0, -1);
364 utf8_read_test(TESTSTR("\xF0\x90\x80\x80"),
365 0x00010000, /* <no name available> */
366 0, -1);
367 utf8_read_test(TESTSTR("\xF8\x88\x80\x80\x80"),
368 0x00200000, /* <no name available> */
369 0, -1);
370 utf8_read_test(TESTSTR("\xFC\x84\x80\x80\x80\x80"),
371 0x04000000, /* <no name available> */
372 0, -1);
373 utf8_read_test(TESTSTR("\x7F"),
374 0x0000007F, /* <control> */
375 0, -1);
376 utf8_read_test(TESTSTR("\xDF\xBF"),
377 0x000007FF, /* <no name available> */
378 0, -1);
379 utf8_read_test(TESTSTR("\xEF\xBF\xBD"),
380 0x0000FFFD, /* REPLACEMENT CHARACTER */
381 0, -1);
382 utf8_read_test(TESTSTR("\xEF\xBF\xBF"),
383 ERROR, /* <no name available> (invalid char) */
384 0, -1);
385 utf8_read_test(TESTSTR("\xF7\xBF\xBF\xBF"),
386 0x001FFFFF, /* <no name available> */
387 0, -1);
388 utf8_read_test(TESTSTR("\xFB\xBF\xBF\xBF\xBF"),
389 0x03FFFFFF, /* <no name available> */
390 0, -1);
391 utf8_read_test(TESTSTR("\xFD\xBF\xBF\xBF\xBF\xBF"),
392 0x7FFFFFFF, /* <no name available> */
393 0, -1);
394 utf8_read_test(TESTSTR("\xED\x9F\xBF"),
395 0x0000D7FF, /* <no name available> */
396 0, -1);
397 utf8_read_test(TESTSTR("\xEE\x80\x80"),
398 0x0000E000, /* <Private Use, First> */
399 0, -1);
400 utf8_read_test(TESTSTR("\xEF\xBF\xBD"),
401 0x0000FFFD, /* REPLACEMENT CHARACTER */
402 0, -1);
403 utf8_read_test(TESTSTR("\xF4\x8F\xBF\xBF"),
404 0x0010FFFF, /* <no name available> */
405 0, -1);
406 utf8_read_test(TESTSTR("\xF4\x90\x80\x80"),
407 0x00110000, /* <no name available> */
408 0, -1);
409 utf8_read_test(TESTSTR("\x80"),
410 ERROR, /* (unexpected continuation byte) */
411 0, -1);
412 utf8_read_test(TESTSTR("\xBF"),
413 ERROR, /* (unexpected continuation byte) */
414 0, -1);
415 utf8_read_test(TESTSTR("\x80\xBF"),
416 ERROR, /* (unexpected continuation byte) */
417 ERROR, /* (unexpected continuation byte) */
418 0, -1);
419 utf8_read_test(TESTSTR("\x80\xBF\x80"),
420 ERROR, /* (unexpected continuation byte) */
421 ERROR, /* (unexpected continuation byte) */
422 ERROR, /* (unexpected continuation byte) */
423 0, -1);
424 utf8_read_test(TESTSTR("\x80\xBF\x80\xBF"),
425 ERROR, /* (unexpected continuation byte) */
426 ERROR, /* (unexpected continuation byte) */
427 ERROR, /* (unexpected continuation byte) */
428 ERROR, /* (unexpected continuation byte) */
429 0, -1);
430 utf8_read_test(TESTSTR("\x80\xBF\x80\xBF\x80"),
431 ERROR, /* (unexpected continuation byte) */
432 ERROR, /* (unexpected continuation byte) */
433 ERROR, /* (unexpected continuation byte) */
434 ERROR, /* (unexpected continuation byte) */
435 ERROR, /* (unexpected continuation byte) */
436 0, -1);
437 utf8_read_test(TESTSTR("\x80\xBF\x80\xBF\x80\xBF"),
438 ERROR, /* (unexpected continuation byte) */
439 ERROR, /* (unexpected continuation byte) */
440 ERROR, /* (unexpected continuation byte) */
441 ERROR, /* (unexpected continuation byte) */
442 ERROR, /* (unexpected continuation byte) */
443 ERROR, /* (unexpected continuation byte) */
444 0, -1);
445 utf8_read_test(TESTSTR("\x80\xBF\x80\xBF\x80\xBF\x80"),
446 ERROR, /* (unexpected continuation byte) */
447 ERROR, /* (unexpected continuation byte) */
448 ERROR, /* (unexpected continuation byte) */
449 ERROR, /* (unexpected continuation byte) */
450 ERROR, /* (unexpected continuation byte) */
451 ERROR, /* (unexpected continuation byte) */
452 ERROR, /* (unexpected continuation byte) */
453 0, -1);
454 utf8_read_test(TESTSTR("\x80\x81\x82\x83\x84\x85\x86\x87\x88\x89\x8A\x8B\x8C\x8D\x8E\x8F\x90\x91\x92\x93\x94\x95\x96\x97\x98\x99\x9A\x9B\x9C\x9D\x9E\x9F\xA0\xA1\xA2\xA3\xA4\xA5\xA6\xA7\xA8\xA9\xAA\xAB\xAC\xAD\xAE\xAF\xB0\xB1\xB2\xB3\xB4\xB5\xB6\xB7\xB8\xB9\xBA\xBB\xBC\xBD\xBE\xBF"),
455 ERROR, /* (unexpected continuation byte) */
456 ERROR, /* (unexpected continuation byte) */
457 ERROR, /* (unexpected continuation byte) */
458 ERROR, /* (unexpected continuation byte) */
459 ERROR, /* (unexpected continuation byte) */
460 ERROR, /* (unexpected continuation byte) */
461 ERROR, /* (unexpected continuation byte) */
462 ERROR, /* (unexpected continuation byte) */
463 ERROR, /* (unexpected continuation byte) */
464 ERROR, /* (unexpected continuation byte) */
465 ERROR, /* (unexpected continuation byte) */
466 ERROR, /* (unexpected continuation byte) */
467 ERROR, /* (unexpected continuation byte) */
468 ERROR, /* (unexpected continuation byte) */
469 ERROR, /* (unexpected continuation byte) */
470 ERROR, /* (unexpected continuation byte) */
471 ERROR, /* (unexpected continuation byte) */
472 ERROR, /* (unexpected continuation byte) */
473 ERROR, /* (unexpected continuation byte) */
474 ERROR, /* (unexpected continuation byte) */
475 ERROR, /* (unexpected continuation byte) */
476 ERROR, /* (unexpected continuation byte) */
477 ERROR, /* (unexpected continuation byte) */
478 ERROR, /* (unexpected continuation byte) */
479 ERROR, /* (unexpected continuation byte) */
480 ERROR, /* (unexpected continuation byte) */
481 ERROR, /* (unexpected continuation byte) */
482 ERROR, /* (unexpected continuation byte) */
483 ERROR, /* (unexpected continuation byte) */
484 ERROR, /* (unexpected continuation byte) */
485 ERROR, /* (unexpected continuation byte) */
486 ERROR, /* (unexpected continuation byte) */
487 ERROR, /* (unexpected continuation byte) */
488 ERROR, /* (unexpected continuation byte) */
489 ERROR, /* (unexpected continuation byte) */
490 ERROR, /* (unexpected continuation byte) */
491 ERROR, /* (unexpected continuation byte) */
492 ERROR, /* (unexpected continuation byte) */
493 ERROR, /* (unexpected continuation byte) */
494 ERROR, /* (unexpected continuation byte) */
495 ERROR, /* (unexpected continuation byte) */
496 ERROR, /* (unexpected continuation byte) */
497 ERROR, /* (unexpected continuation byte) */
498 ERROR, /* (unexpected continuation byte) */
499 ERROR, /* (unexpected continuation byte) */
500 ERROR, /* (unexpected continuation byte) */
501 ERROR, /* (unexpected continuation byte) */
502 ERROR, /* (unexpected continuation byte) */
503 ERROR, /* (unexpected continuation byte) */
504 ERROR, /* (unexpected continuation byte) */
505 ERROR, /* (unexpected continuation byte) */
506 ERROR, /* (unexpected continuation byte) */
507 ERROR, /* (unexpected continuation byte) */
508 ERROR, /* (unexpected continuation byte) */
509 ERROR, /* (unexpected continuation byte) */
510 ERROR, /* (unexpected continuation byte) */
511 ERROR, /* (unexpected continuation byte) */
512 ERROR, /* (unexpected continuation byte) */
513 ERROR, /* (unexpected continuation byte) */
514 ERROR, /* (unexpected continuation byte) */
515 ERROR, /* (unexpected continuation byte) */
516 ERROR, /* (unexpected continuation byte) */
517 ERROR, /* (unexpected continuation byte) */
518 ERROR, /* (unexpected continuation byte) */
519 0, -1);
520 utf8_read_test(TESTSTR("\xC0\x20\xC1\x20\xC2\x20\xC3\x20\xC4\x20\xC5\x20\xC6\x20\xC7\x20"),
521 ERROR, /* (incomplete sequence) */
522 0x00000020, /* SPACE */
523 ERROR, /* (incomplete sequence) */
524 0x00000020, /* SPACE */
525 ERROR, /* (incomplete sequence) */
526 0x00000020, /* SPACE */
527 ERROR, /* (incomplete sequence) */
528 0x00000020, /* SPACE */
529 ERROR, /* (incomplete sequence) */
530 0x00000020, /* SPACE */
531 ERROR, /* (incomplete sequence) */
532 0x00000020, /* SPACE */
533 ERROR, /* (incomplete sequence) */
534 0x00000020, /* SPACE */
535 ERROR, /* (incomplete sequence) */
536 0x00000020, /* SPACE */
537 0, -1);
538 utf8_read_test(TESTSTR("\xE0\x20\xE1\x20\xE2\x20\xE3\x20\xE4\x20\xE5\x20\xE6\x20\xE7\x20\xE8\x20\xE9\x20\xEA\x20\xEB\x20\xEC\x20\xED\x20\xEE\x20\xEF\x20"),
539 ERROR, /* (incomplete sequence) */
540 0x00000020, /* SPACE */
541 ERROR, /* (incomplete sequence) */
542 0x00000020, /* SPACE */
543 ERROR, /* (incomplete sequence) */
544 0x00000020, /* SPACE */
545 ERROR, /* (incomplete sequence) */
546 0x00000020, /* SPACE */
547 ERROR, /* (incomplete sequence) */
548 0x00000020, /* SPACE */
549 ERROR, /* (incomplete sequence) */
550 0x00000020, /* SPACE */
551 ERROR, /* (incomplete sequence) */
552 0x00000020, /* SPACE */
553 ERROR, /* (incomplete sequence) */
554 0x00000020, /* SPACE */
555 ERROR, /* (incomplete sequence) */
556 0x00000020, /* SPACE */
557 ERROR, /* (incomplete sequence) */
558 0x00000020, /* SPACE */
559 ERROR, /* (incomplete sequence) */
560 0x00000020, /* SPACE */
561 ERROR, /* (incomplete sequence) */
562 0x00000020, /* SPACE */
563 ERROR, /* (incomplete sequence) */
564 0x00000020, /* SPACE */
565 ERROR, /* (incomplete sequence) */
566 0x00000020, /* SPACE */
567 ERROR, /* (incomplete sequence) */
568 0x00000020, /* SPACE */
569 ERROR, /* (incomplete sequence) */
570 0x00000020, /* SPACE */
571 0, -1);
572 utf8_read_test(TESTSTR("\xF0\x20\xF1\x20\xF2\x20\xF3\x20\xF4\x20\xF5\x20\xF6\x20\xF7\x20"),
573 ERROR, /* (incomplete sequence) */
574 0x00000020, /* SPACE */
575 ERROR, /* (incomplete sequence) */
576 0x00000020, /* SPACE */
577 ERROR, /* (incomplete sequence) */
578 0x00000020, /* SPACE */
579 ERROR, /* (incomplete sequence) */
580 0x00000020, /* SPACE */
581 ERROR, /* (incomplete sequence) */
582 0x00000020, /* SPACE */
583 ERROR, /* (incomplete sequence) */
584 0x00000020, /* SPACE */
585 ERROR, /* (incomplete sequence) */
586 0x00000020, /* SPACE */
587 ERROR, /* (incomplete sequence) */
588 0x00000020, /* SPACE */
589 0, -1);
590 utf8_read_test(TESTSTR("\xF8\x20\xF9\x20\xFA\x20\xFB\x20"),
591 ERROR, /* (incomplete sequence) */
592 0x00000020, /* SPACE */
593 ERROR, /* (incomplete sequence) */
594 0x00000020, /* SPACE */
595 ERROR, /* (incomplete sequence) */
596 0x00000020, /* SPACE */
597 ERROR, /* (incomplete sequence) */
598 0x00000020, /* SPACE */
599 0, -1);
600 utf8_read_test(TESTSTR("\xFC\x20\xFD\x20"),
601 ERROR, /* (incomplete sequence) */
602 0x00000020, /* SPACE */
603 ERROR, /* (incomplete sequence) */
604 0x00000020, /* SPACE */
605 0, -1);
606 utf8_read_test(TESTSTR("\xC0"),
607 ERROR, /* (incomplete sequence) */
608 0, -1);
609 utf8_read_test(TESTSTR("\xE0\x80"),
610 ERROR, /* (incomplete sequence) */
611 0, -1);
612 utf8_read_test(TESTSTR("\xF0\x80\x80"),
613 ERROR, /* (incomplete sequence) */
614 0, -1);
615 utf8_read_test(TESTSTR("\xF8\x80\x80\x80"),
616 ERROR, /* (incomplete sequence) */
617 0, -1);
618 utf8_read_test(TESTSTR("\xFC\x80\x80\x80\x80"),
619 ERROR, /* (incomplete sequence) */
620 0, -1);
621 utf8_read_test(TESTSTR("\xDF"),
622 ERROR, /* (incomplete sequence) */
623 0, -1);
624 utf8_read_test(TESTSTR("\xEF\xBF"),
625 ERROR, /* (incomplete sequence) */
626 0, -1);
627 utf8_read_test(TESTSTR("\xF7\xBF\xBF"),
628 ERROR, /* (incomplete sequence) */
629 0, -1);
630 utf8_read_test(TESTSTR("\xFB\xBF\xBF\xBF"),
631 ERROR, /* (incomplete sequence) */
632 0, -1);
633 utf8_read_test(TESTSTR("\xFD\xBF\xBF\xBF\xBF"),
634 ERROR, /* (incomplete sequence) */
635 0, -1);
636 utf8_read_test(TESTSTR("\xC0\xE0\x80\xF0\x80\x80\xF8\x80\x80\x80\xFC\x80\x80\x80\x80\xDF\xEF\xBF\xF7\xBF\xBF\xFB\xBF\xBF\xBF\xFD\xBF\xBF\xBF\xBF"),
637 ERROR, /* (incomplete sequence) */
638 ERROR, /* (incomplete sequence) */
639 ERROR, /* (incomplete sequence) */
640 ERROR, /* (incomplete sequence) */
641 ERROR, /* (incomplete sequence) */
642 ERROR, /* (incomplete sequence) */
643 ERROR, /* (incomplete sequence) */
644 ERROR, /* (incomplete sequence) */
645 ERROR, /* (incomplete sequence) */
646 ERROR, /* (incomplete sequence) */
647 0, -1);
648 utf8_read_test(TESTSTR("\xFE"),
649 ERROR, /* (invalid UTF-8 byte) */
650 0, -1);
651 utf8_read_test(TESTSTR("\xFF"),
652 ERROR, /* (invalid UTF-8 byte) */
653 0, -1);
654 utf8_read_test(TESTSTR("\xFE\xFE\xFF\xFF"),
655 ERROR, /* (invalid UTF-8 byte) */
656 ERROR, /* (invalid UTF-8 byte) */
657 ERROR, /* (invalid UTF-8 byte) */
658 ERROR, /* (invalid UTF-8 byte) */
659 0, -1);
660 utf8_read_test(TESTSTR("\xC0\xAF"),
661 ERROR, /* SOLIDUS (overlong form of 2F) */
662 0, -1);
663 utf8_read_test(TESTSTR("\xE0\x80\xAF"),
664 ERROR, /* SOLIDUS (overlong form of 2F) */
665 0, -1);
666 utf8_read_test(TESTSTR("\xF0\x80\x80\xAF"),
667 ERROR, /* SOLIDUS (overlong form of 2F) */
668 0, -1);
669 utf8_read_test(TESTSTR("\xF8\x80\x80\x80\xAF"),
670 ERROR, /* SOLIDUS (overlong form of 2F) */
671 0, -1);
672 utf8_read_test(TESTSTR("\xFC\x80\x80\x80\x80\xAF"),
673 ERROR, /* SOLIDUS (overlong form of 2F) */
674 0, -1);
675 utf8_read_test(TESTSTR("\xC1\xBF"),
676 ERROR, /* <control> (overlong form of 7F) */
677 0, -1);
678 utf8_read_test(TESTSTR("\xE0\x9F\xBF"),
679 ERROR, /* <no name available> (overlong form of DF BF) */
680 0, -1);
681 utf8_read_test(TESTSTR("\xF0\x8F\xBF\xBF"),
682 ERROR, /* <no name available> (overlong form of EF BF BF) (invalid char) */
683 0, -1);
684 utf8_read_test(TESTSTR("\xF8\x87\xBF\xBF\xBF"),
685 ERROR, /* <no name available> (overlong form of F7 BF BF BF) */
686 0, -1);
687 utf8_read_test(TESTSTR("\xFC\x83\xBF\xBF\xBF\xBF"),
688 ERROR, /* <no name available> (overlong form of FB BF BF BF BF) */
689 0, -1);
690 utf8_read_test(TESTSTR("\xC0\x80"),
691 ERROR, /* <control> (overlong form of 00) */
692 0, -1);
693 utf8_read_test(TESTSTR("\xE0\x80\x80"),
694 ERROR, /* <control> (overlong form of 00) */
695 0, -1);
696 utf8_read_test(TESTSTR("\xF0\x80\x80\x80"),
697 ERROR, /* <control> (overlong form of 00) */
698 0, -1);
699 utf8_read_test(TESTSTR("\xF8\x80\x80\x80\x80"),
700 ERROR, /* <control> (overlong form of 00) */
701 0, -1);
702 utf8_read_test(TESTSTR("\xFC\x80\x80\x80\x80\x80"),
703 ERROR, /* <control> (overlong form of 00) */
704 0, -1);
705 utf8_read_test(TESTSTR("\xED\xA0\x80"),
706 ERROR, /* <Non Private Use High Surrogate, First> (surrogate) */
707 0, -1);
708 utf8_read_test(TESTSTR("\xED\xAD\xBF"),
709 ERROR, /* <Non Private Use High Surrogate, Last> (surrogate) */
710 0, -1);
711 utf8_read_test(TESTSTR("\xED\xAE\x80"),
712 ERROR, /* <Private Use High Surrogate, First> (surrogate) */
713 0, -1);
714 utf8_read_test(TESTSTR("\xED\xAF\xBF"),
715 ERROR, /* <Private Use High Surrogate, Last> (surrogate) */
716 0, -1);
717 utf8_read_test(TESTSTR("\xED\xB0\x80"),
718 ERROR, /* <Low Surrogate, First> (surrogate) */
719 0, -1);
720 utf8_read_test(TESTSTR("\xED\xBE\x80"),
721 ERROR, /* <no name available> (surrogate) */
722 0, -1);
723 utf8_read_test(TESTSTR("\xED\xBF\xBF"),
724 ERROR, /* <Low Surrogate, Last> (surrogate) */
725 0, -1);
726 utf8_read_test(TESTSTR("\xED\xA0\x80\xED\xB0\x80"),
727 ERROR, /* <Non Private Use High Surrogate, First> (surrogate) */
728 ERROR, /* <Low Surrogate, First> (surrogate) */
729 0, -1);
730 utf8_read_test(TESTSTR("\xED\xA0\x80\xED\xBF\xBF"),
731 ERROR, /* <Non Private Use High Surrogate, First> (surrogate) */
732 ERROR, /* <Low Surrogate, Last> (surrogate) */
733 0, -1);
734 utf8_read_test(TESTSTR("\xED\xAD\xBF\xED\xB0\x80"),
735 ERROR, /* <Non Private Use High Surrogate, Last> (surrogate) */
736 ERROR, /* <Low Surrogate, First> (surrogate) */
737 0, -1);
738 utf8_read_test(TESTSTR("\xED\xAD\xBF\xED\xBF\xBF"),
739 ERROR, /* <Non Private Use High Surrogate, Last> (surrogate) */
740 ERROR, /* <Low Surrogate, Last> (surrogate) */
741 0, -1);
742 utf8_read_test(TESTSTR("\xED\xAE\x80\xED\xB0\x80"),
743 ERROR, /* <Private Use High Surrogate, First> (surrogate) */
744 ERROR, /* <Low Surrogate, First> (surrogate) */
745 0, -1);
746 utf8_read_test(TESTSTR("\xED\xAE\x80\xED\xBF\xBF"),
747 ERROR, /* <Private Use High Surrogate, First> (surrogate) */
748 ERROR, /* <Low Surrogate, Last> (surrogate) */
749 0, -1);
750 utf8_read_test(TESTSTR("\xED\xAF\xBF\xED\xB0\x80"),
751 ERROR, /* <Private Use High Surrogate, Last> (surrogate) */
752 ERROR, /* <Low Surrogate, First> (surrogate) */
753 0, -1);
754 utf8_read_test(TESTSTR("\xED\xAF\xBF\xED\xBF\xBF"),
755 ERROR, /* <Private Use High Surrogate, Last> (surrogate) */
756 ERROR, /* <Low Surrogate, Last> (surrogate) */
757 0, -1);
758 utf8_read_test(TESTSTR("\xEF\xBF\xBE"),
759 ERROR, /* <no name available> (invalid char) */
760 0, -1);
761 utf8_read_test(TESTSTR("\xEF\xBF\xBF"),
762 ERROR, /* <no name available> (invalid char) */
763 0, -1);
764 printf("read tests completed\n");
765 printf("write tests beginning\n");
766 {
767 const static long str[] =
768 {0x03BAL, 0x1F79L, 0x03C3L, 0x03BCL, 0x03B5L, 0};
769 utf8_write_test(TESTSTR(str),
770 0xCE, 0xBA,
771 0xE1, 0xBD, 0xB9,
772 0xCF, 0x83,
773 0xCE, 0xBC,
774 0xCE, 0xB5,
775 0, -1);
776 }
777 {
778 const static long str[] = {0x0000L, 0};
779 utf8_write_test(TESTSTR(str),
780 0x00,
781 0, -1);
782 }
783 {
784 const static long str[] = {0x0080L, 0};
785 utf8_write_test(TESTSTR(str),
786 0xC2, 0x80,
787 0, -1);
788 }
789 {
790 const static long str[] = {0x0800L, 0};
791 utf8_write_test(TESTSTR(str),
792 0xE0, 0xA0, 0x80,
793 0, -1);
794 }
795 {
796 const static long str[] = {0x00010000L, 0};
797 utf8_write_test(TESTSTR(str),
798 0xF0, 0x90, 0x80, 0x80,
799 0, -1);
800 }
801 {
802 const static long str[] = {0x00200000L, 0};
803 utf8_write_test(TESTSTR(str),
804 0xF8, 0x88, 0x80, 0x80, 0x80,
805 0, -1);
806 }
807 {
808 const static long str[] = {0x04000000L, 0};
809 utf8_write_test(TESTSTR(str),
810 0xFC, 0x84, 0x80, 0x80, 0x80, 0x80,
811 0, -1);
812 }
813 {
814 const static long str[] = {0x007FL, 0};
815 utf8_write_test(TESTSTR(str),
816 0x7F,
817 0, -1);
818 }
819 {
820 const static long str[] = {0x07FFL, 0};
821 utf8_write_test(TESTSTR(str),
822 0xDF, 0xBF,
823 0, -1);
824 }
825 {
826 const static long str[] = {0xFFFDL, 0};
827 utf8_write_test(TESTSTR(str),
828 0xEF, 0xBF, 0xBD,
829 0, -1);
830 }
831 {
832 const static long str[] = {0xFFFFL, 0};
833 utf8_write_test(TESTSTR(str),
834 ERROR,
835 0, -1);
836 }
837 {
838 const static long str[] = {0x001FFFFFL, 0};
839 utf8_write_test(TESTSTR(str),
840 0xF7, 0xBF, 0xBF, 0xBF,
841 0, -1);
842 }
843 {
844 const static long str[] = {0x03FFFFFFL, 0};
845 utf8_write_test(TESTSTR(str),
846 0xFB, 0xBF, 0xBF, 0xBF, 0xBF,
847 0, -1);
848 }
849 {
850 const static long str[] = {0x7FFFFFFFL, 0};
851 utf8_write_test(TESTSTR(str),
852 0xFD, 0xBF, 0xBF, 0xBF, 0xBF, 0xBF,
853 0, -1);
854 }
855 {
856 const static long str[] = {0xD7FFL, 0};
857 utf8_write_test(TESTSTR(str),
858 0xED, 0x9F, 0xBF,
859 0, -1);
860 }
861 {
862 const static long str[] = {0xD800L, 0};
863 utf8_write_test(TESTSTR(str),
864 ERROR,
865 0, -1);
866 }
867 {
868 const static long str[] = {0xD800L, 0xDC00L, 0};
869 utf8_write_test(TESTSTR(str),
870 ERROR,
871 ERROR,
872 0, -1);
873 }
874 {
875 const static long str[] = {0xDFFFL, 0};
876 utf8_write_test(TESTSTR(str),
877 ERROR,
878 0, -1);
879 }
880 {
881 const static long str[] = {0xE000L, 0};
882 utf8_write_test(TESTSTR(str),
883 0xEE, 0x80, 0x80,
884 0, -1);
885 }
886 printf("write tests completed\n");
887
888 printf("total: %d errors\n", total_errs);
889 return (total_errs != 0);
890 }
891 #endif /* TESTMODE */
892
893 const charset_spec charset_CS_UTF8 = {
894 CS_UTF8, read_utf8, write_utf8, NULL
895 };
896
897 #else /* ENUM_CHARSETS */
898
899 ENUM_CHARSET(CS_UTF8)
900
901 #endif /* ENUM_CHARSETS */
Simon Tatham's `charset' library; import of svn://svn.tartarus.org/sgt/charset