2dc6356a |
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 | |
72ecd48a |
10 | void read_utf8(charset_spec const *, long int, charset_state *, |
11 | void (*)(void *, long int), void *); |
12 | void write_utf8(charset_spec const *, long int, |
13 | charset_state *, void (*)(void *, long int), void *); |
14 | |
2dc6356a |
15 | /* |
16 | * UTF-8 has no associated data, so `charset' may be ignored. |
17 | */ |
18 | |
19 | void read_utf8(charset_spec const *charset, long int input_chr, |
20 | charset_state *state, |
21 | void (*emit)(void *ctx, long int output), void *emitctx) |
22 | { |
23 | UNUSEDARG(charset); |
24 | |
25 | /* |
26 | * For reading UTF-8, the `state' word contains: |
27 | * |
28 | * - in bits 29-31, the number of bytes expected to be in the |
29 | * current multibyte character (which we can tell instantly |
30 | * from the first byte, of course). |
31 | * |
32 | * - in bits 26-28, the number of bytes _seen so far_ in the |
33 | * current multibyte character. |
34 | * |
35 | * - in the remainder of the word, the current value of the |
36 | * character, which is shifted upwards by 6 bits to |
37 | * accommodate each new byte. |
38 | * |
39 | * As required, the state is zero when we are not in the middle |
40 | * of a multibyte character at all. |
41 | * |
42 | * For example, when reading E9 8D 8B, starting at state=0: |
43 | * |
44 | * - after E9, the state is 0x64000009 |
45 | * - after 8D, the state is 0x6800024d |
46 | * - after 8B, the state conceptually becomes 0x6c00934b, at |
47 | * which point we notice we've got as many characters as we |
48 | * were expecting, output U+934B, and reset the state to |
49 | * zero. |
50 | * |
51 | * Note that the maximum number of bits we might need to store |
52 | * in the character value field is 25 (U+7FFFFFFF contains 31 |
53 | * bits, but we will never actually store its full value |
54 | * because when we receive the last 6 bits in the final |
55 | * continuation byte we will output it and revert the state to |
56 | * zero). Hence the character value field never collides with |
57 | * the byte counts. |
58 | */ |
59 | |
60 | if (input_chr < 0x80) { |
61 | /* |
62 | * Single-byte character. If the state is nonzero before |
63 | * coming here, output an error for an incomplete sequence. |
64 | * Then output the character. |
65 | */ |
66 | if (state->s0 != 0) { |
67 | emit(emitctx, ERROR); |
68 | state->s0 = 0; |
69 | } |
70 | emit(emitctx, input_chr); |
71 | } else if (input_chr == 0xFE || input_chr == 0xFF) { |
72 | /* |
73 | * FE and FF bytes should _never_ occur in UTF-8. They are |
74 | * automatic errors; if the state was nonzero to start |
75 | * with, output a further error for an incomplete sequence. |
76 | */ |
77 | if (state->s0 != 0) { |
78 | emit(emitctx, ERROR); |
79 | state->s0 = 0; |
80 | } |
81 | emit(emitctx, ERROR); |
82 | } else if (input_chr >= 0x80 && input_chr < 0xC0) { |
83 | /* |
84 | * Continuation byte. Output an error for an unexpected |
85 | * continuation byte, if the state is zero. |
86 | */ |
87 | if (state->s0 == 0) { |
88 | emit(emitctx, ERROR); |
89 | } else { |
90 | unsigned long charval; |
91 | unsigned long topstuff; |
92 | int bytes; |
93 | |
94 | /* |
95 | * Otherwise, accumulate more of the character value. |
96 | */ |
97 | charval = state->s0 & 0x03ffffffL; |
98 | charval = (charval << 6) | (input_chr & 0x3F); |
99 | |
100 | /* |
101 | * Check the byte counts; if we have not reached the |
102 | * end of the character, update the state and return. |
103 | */ |
104 | topstuff = state->s0 & 0xfc000000L; |
105 | topstuff += 0x04000000L; /* add one to the byte count */ |
106 | if (((topstuff << 3) ^ topstuff) & 0xe0000000L) { |
107 | state->s0 = topstuff | charval; |
108 | return; |
109 | } |
110 | |
111 | /* |
112 | * Now we know we've reached the end of the character. |
113 | * `charval' is the Unicode value. We should check for |
114 | * various invalid things, and then either output |
115 | * charval or an error. In all cases we reset the state |
116 | * to zero. |
117 | */ |
118 | bytes = topstuff >> 29; |
119 | state->s0 = 0; |
120 | |
121 | if (charval >= 0xD800 && charval < 0xE000) { |
122 | /* |
123 | * Surrogates (0xD800-0xDFFF) may never be encoded |
124 | * in UTF-8. A surrogate pair in Unicode should |
125 | * have been encoded as a single UTF-8 character |
126 | * occupying more than three bytes. |
127 | */ |
128 | emit(emitctx, ERROR); |
129 | } else if (charval == 0xFFFE || charval == 0xFFFF) { |
130 | /* |
131 | * U+FFFE and U+FFFF are invalid Unicode characters |
132 | * and may never be encoded in UTF-8. (This is one |
133 | * reason why U+FFFF is our way of signalling an |
134 | * error to our `emit' function :-) |
135 | */ |
136 | emit(emitctx, ERROR); |
137 | } else if ((charval <= 0x7FL /* && bytes > 1 */) || |
138 | (charval <= 0x7FFL && bytes > 2) || |
139 | (charval <= 0xFFFFL && bytes > 3) || |
140 | (charval <= 0x1FFFFFL && bytes > 4) || |
141 | (charval <= 0x3FFFFFFL && bytes > 5)) { |
142 | /* |
143 | * Overlong sequences are not to be tolerated, |
144 | * under any circumstances. |
145 | */ |
146 | emit(emitctx, ERROR); |
147 | } else { |
148 | /* |
149 | * Oh, all right. We'll let this one off. |
150 | */ |
151 | emit(emitctx, charval); |
152 | } |
153 | } |
154 | |
155 | } else { |
156 | /* |
157 | * Lead byte. First output an error for an incomplete |
158 | * sequence, if the state is nonzero. |
159 | */ |
160 | if (state->s0 != 0) |
161 | emit(emitctx, ERROR); |
162 | |
163 | /* |
164 | * Now deal with the lead byte: work out the number of |
165 | * bytes we expect to see in this character, and extract |
166 | * the initial bits of it too. |
167 | */ |
168 | if (input_chr >= 0xC0 && input_chr < 0xE0) { |
169 | state->s0 = 0x44000000L | (input_chr & 0x1F); |
170 | } else if (input_chr >= 0xE0 && input_chr < 0xF0) { |
171 | state->s0 = 0x64000000L | (input_chr & 0x0F); |
172 | } else if (input_chr >= 0xF0 && input_chr < 0xF8) { |
173 | state->s0 = 0x84000000L | (input_chr & 0x07); |
174 | } else if (input_chr >= 0xF8 && input_chr < 0xFC) { |
175 | state->s0 = 0xa4000000L | (input_chr & 0x03); |
176 | } else if (input_chr >= 0xFC && input_chr < 0xFE) { |
177 | state->s0 = 0xc4000000L | (input_chr & 0x01); |
178 | } |
179 | } |
180 | } |
181 | |
182 | /* |
183 | * UTF-8 is a stateless multi-byte encoding (in the sense that just |
184 | * after any character has been completed, the state is always the |
185 | * same); hence when writing it, there is no need to use the |
186 | * charset_state. |
187 | */ |
188 | |
189 | void write_utf8(charset_spec const *charset, long int input_chr, |
190 | charset_state *state, |
191 | void (*emit)(void *ctx, long int output), void *emitctx) |
192 | { |
193 | UNUSEDARG(charset); |
194 | UNUSEDARG(state); |
195 | |
196 | /* |
197 | * Refuse to output any illegal code points. |
198 | */ |
199 | if (input_chr == 0xFFFE || input_chr == 0xFFFF || |
200 | (input_chr >= 0xD800 && input_chr < 0xE000)) { |
201 | emit(emitctx, ERROR); |
202 | } else if (input_chr < 0x80) { /* one-byte character */ |
203 | emit(emitctx, input_chr); |
204 | } else if (input_chr < 0x800) { /* two-byte character */ |
205 | emit(emitctx, 0xC0 | (0x1F & (input_chr >> 6))); |
206 | emit(emitctx, 0x80 | (0x3F & (input_chr ))); |
207 | } else if (input_chr < 0x10000) { /* three-byte character */ |
208 | emit(emitctx, 0xE0 | (0x0F & (input_chr >> 12))); |
209 | emit(emitctx, 0x80 | (0x3F & (input_chr >> 6))); |
210 | emit(emitctx, 0x80 | (0x3F & (input_chr ))); |
211 | } else if (input_chr < 0x200000) { /* four-byte character */ |
212 | emit(emitctx, 0xF0 | (0x07 & (input_chr >> 18))); |
213 | emit(emitctx, 0x80 | (0x3F & (input_chr >> 12))); |
214 | emit(emitctx, 0x80 | (0x3F & (input_chr >> 6))); |
215 | emit(emitctx, 0x80 | (0x3F & (input_chr ))); |
216 | } else if (input_chr < 0x4000000) {/* five-byte character */ |
217 | emit(emitctx, 0xF8 | (0x03 & (input_chr >> 24))); |
218 | emit(emitctx, 0x80 | (0x3F & (input_chr >> 18))); |
219 | emit(emitctx, 0x80 | (0x3F & (input_chr >> 12))); |
220 | emit(emitctx, 0x80 | (0x3F & (input_chr >> 6))); |
221 | emit(emitctx, 0x80 | (0x3F & (input_chr ))); |
222 | } else { /* six-byte character */ |
223 | emit(emitctx, 0xFC | (0x01 & (input_chr >> 30))); |
224 | emit(emitctx, 0x80 | (0x3F & (input_chr >> 24))); |
225 | emit(emitctx, 0x80 | (0x3F & (input_chr >> 18))); |
226 | emit(emitctx, 0x80 | (0x3F & (input_chr >> 12))); |
227 | emit(emitctx, 0x80 | (0x3F & (input_chr >> 6))); |
228 | emit(emitctx, 0x80 | (0x3F & (input_chr ))); |
229 | } |
230 | } |
231 | |
232 | #ifdef TESTMODE |
233 | |
234 | #include <stdio.h> |
235 | #include <stdarg.h> |
236 | |
237 | int total_errs = 0; |
238 | |
239 | void utf8_emit(void *ctx, long output) |
240 | { |
241 | wchar_t **p = (wchar_t **)ctx; |
242 | *(*p)++ = output; |
243 | } |
244 | |
245 | void utf8_read_test(int line, char *input, int inlen, ...) |
246 | { |
247 | va_list ap; |
248 | wchar_t *p, str[512]; |
249 | int i; |
250 | charset_state state; |
251 | unsigned long l; |
252 | |
253 | state.s0 = 0; |
254 | p = str; |
255 | |
256 | for (i = 0; i < inlen; i++) |
257 | read_utf8(NULL, input[i] & 0xFF, &state, utf8_emit, &p); |
258 | |
259 | va_start(ap, inlen); |
260 | l = 0; |
261 | for (i = 0; i < p - str; i++) { |
262 | l = va_arg(ap, long int); |
263 | if (l == -1) { |
264 | printf("%d: correct string shorter than output\n", line); |
265 | total_errs++; |
266 | break; |
267 | } |
268 | if (l != str[i]) { |
269 | printf("%d: char %d came out as %08x, should be %08x\n", |
270 | line, i, str[i], l); |
271 | total_errs++; |
272 | } |
273 | } |
274 | if (l != -1) { |
275 | l = va_arg(ap, long int); |
276 | if (l != -1) { |
277 | printf("%d: correct string longer than output\n", line); |
278 | total_errs++; |
279 | } |
280 | } |
281 | va_end(ap); |
282 | } |
283 | |
284 | void utf8_write_test(int line, const long *input, int inlen, ...) |
285 | { |
286 | va_list ap; |
287 | wchar_t *p, str[512]; |
288 | int i; |
289 | charset_state state; |
290 | unsigned long l; |
291 | |
292 | state.s0 = 0; |
293 | p = str; |
294 | |
295 | for (i = 0; i < inlen; i++) |
296 | write_utf8(NULL, input[i], &state, utf8_emit, &p); |
297 | |
298 | va_start(ap, inlen); |
299 | l = 0; |
300 | for (i = 0; i < p - str; i++) { |
301 | l = va_arg(ap, long int); |
302 | if (l == -1) { |
303 | printf("%d: correct string shorter than output\n", line); |
304 | total_errs++; |
305 | break; |
306 | } |
307 | if (l != str[i]) { |
308 | printf("%d: char %d came out as %08x, should be %08x\n", |
309 | line, i, str[i], l); |
310 | total_errs++; |
311 | } |
312 | } |
313 | if (l != -1) { |
314 | l = va_arg(ap, long int); |
315 | if (l != -1) { |
316 | printf("%d: correct string longer than output\n", line); |
317 | total_errs++; |
318 | } |
319 | } |
320 | va_end(ap); |
321 | } |
322 | |
323 | /* Macro to concoct the first three parameters of utf8_read_test. */ |
324 | #define TESTSTR(x) __LINE__, x, lenof(x) |
325 | |
326 | int main(void) |
327 | { |
328 | printf("read tests beginning\n"); |
329 | utf8_read_test(TESTSTR("\xCE\xBA\xE1\xBD\xB9\xCF\x83\xCE\xBC\xCE\xB5"), |
330 | 0x000003BA, /* GREEK SMALL LETTER KAPPA */ |
331 | 0x00001F79, /* GREEK SMALL LETTER OMICRON WITH OXIA */ |
332 | 0x000003C3, /* GREEK SMALL LETTER SIGMA */ |
333 | 0x000003BC, /* GREEK SMALL LETTER MU */ |
334 | 0x000003B5, /* GREEK SMALL LETTER EPSILON */ |
335 | 0, -1); |
336 | utf8_read_test(TESTSTR("\x00"), |
337 | 0x00000000, /* <control> */ |
338 | 0, -1); |
339 | utf8_read_test(TESTSTR("\xC2\x80"), |
340 | 0x00000080, /* <control> */ |
341 | 0, -1); |
342 | utf8_read_test(TESTSTR("\xE0\xA0\x80"), |
343 | 0x00000800, /* <no name available> */ |
344 | 0, -1); |
345 | utf8_read_test(TESTSTR("\xF0\x90\x80\x80"), |
346 | 0x00010000, /* <no name available> */ |
347 | 0, -1); |
348 | utf8_read_test(TESTSTR("\xF8\x88\x80\x80\x80"), |
349 | 0x00200000, /* <no name available> */ |
350 | 0, -1); |
351 | utf8_read_test(TESTSTR("\xFC\x84\x80\x80\x80\x80"), |
352 | 0x04000000, /* <no name available> */ |
353 | 0, -1); |
354 | utf8_read_test(TESTSTR("\x7F"), |
355 | 0x0000007F, /* <control> */ |
356 | 0, -1); |
357 | utf8_read_test(TESTSTR("\xDF\xBF"), |
358 | 0x000007FF, /* <no name available> */ |
359 | 0, -1); |
360 | utf8_read_test(TESTSTR("\xEF\xBF\xBD"), |
361 | 0x0000FFFD, /* REPLACEMENT CHARACTER */ |
362 | 0, -1); |
363 | utf8_read_test(TESTSTR("\xEF\xBF\xBF"), |
364 | ERROR, /* <no name available> (invalid char) */ |
365 | 0, -1); |
366 | utf8_read_test(TESTSTR("\xF7\xBF\xBF\xBF"), |
367 | 0x001FFFFF, /* <no name available> */ |
368 | 0, -1); |
369 | utf8_read_test(TESTSTR("\xFB\xBF\xBF\xBF\xBF"), |
370 | 0x03FFFFFF, /* <no name available> */ |
371 | 0, -1); |
372 | utf8_read_test(TESTSTR("\xFD\xBF\xBF\xBF\xBF\xBF"), |
373 | 0x7FFFFFFF, /* <no name available> */ |
374 | 0, -1); |
375 | utf8_read_test(TESTSTR("\xED\x9F\xBF"), |
376 | 0x0000D7FF, /* <no name available> */ |
377 | 0, -1); |
378 | utf8_read_test(TESTSTR("\xEE\x80\x80"), |
379 | 0x0000E000, /* <Private Use, First> */ |
380 | 0, -1); |
381 | utf8_read_test(TESTSTR("\xEF\xBF\xBD"), |
382 | 0x0000FFFD, /* REPLACEMENT CHARACTER */ |
383 | 0, -1); |
384 | utf8_read_test(TESTSTR("\xF4\x8F\xBF\xBF"), |
385 | 0x0010FFFF, /* <no name available> */ |
386 | 0, -1); |
387 | utf8_read_test(TESTSTR("\xF4\x90\x80\x80"), |
388 | 0x00110000, /* <no name available> */ |
389 | 0, -1); |
390 | utf8_read_test(TESTSTR("\x80"), |
391 | ERROR, /* (unexpected continuation byte) */ |
392 | 0, -1); |
393 | utf8_read_test(TESTSTR("\xBF"), |
394 | ERROR, /* (unexpected continuation byte) */ |
395 | 0, -1); |
396 | utf8_read_test(TESTSTR("\x80\xBF"), |
397 | ERROR, /* (unexpected continuation byte) */ |
398 | ERROR, /* (unexpected continuation byte) */ |
399 | 0, -1); |
400 | utf8_read_test(TESTSTR("\x80\xBF\x80"), |
401 | ERROR, /* (unexpected continuation byte) */ |
402 | ERROR, /* (unexpected continuation byte) */ |
403 | ERROR, /* (unexpected continuation byte) */ |
404 | 0, -1); |
405 | utf8_read_test(TESTSTR("\x80\xBF\x80\xBF"), |
406 | ERROR, /* (unexpected continuation byte) */ |
407 | ERROR, /* (unexpected continuation byte) */ |
408 | ERROR, /* (unexpected continuation byte) */ |
409 | ERROR, /* (unexpected continuation byte) */ |
410 | 0, -1); |
411 | utf8_read_test(TESTSTR("\x80\xBF\x80\xBF\x80"), |
412 | ERROR, /* (unexpected continuation byte) */ |
413 | ERROR, /* (unexpected continuation byte) */ |
414 | ERROR, /* (unexpected continuation byte) */ |
415 | ERROR, /* (unexpected continuation byte) */ |
416 | ERROR, /* (unexpected continuation byte) */ |
417 | 0, -1); |
418 | utf8_read_test(TESTSTR("\x80\xBF\x80\xBF\x80\xBF"), |
419 | ERROR, /* (unexpected continuation byte) */ |
420 | ERROR, /* (unexpected continuation byte) */ |
421 | ERROR, /* (unexpected continuation byte) */ |
422 | ERROR, /* (unexpected continuation byte) */ |
423 | ERROR, /* (unexpected continuation byte) */ |
424 | ERROR, /* (unexpected continuation byte) */ |
425 | 0, -1); |
426 | utf8_read_test(TESTSTR("\x80\xBF\x80\xBF\x80\xBF\x80"), |
427 | ERROR, /* (unexpected continuation byte) */ |
428 | ERROR, /* (unexpected continuation byte) */ |
429 | ERROR, /* (unexpected continuation byte) */ |
430 | ERROR, /* (unexpected continuation byte) */ |
431 | ERROR, /* (unexpected continuation byte) */ |
432 | ERROR, /* (unexpected continuation byte) */ |
433 | ERROR, /* (unexpected continuation byte) */ |
434 | 0, -1); |
435 | 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"), |
436 | ERROR, /* (unexpected continuation byte) */ |
437 | ERROR, /* (unexpected continuation byte) */ |
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 | ERROR, /* (unexpected continuation byte) */ |
445 | ERROR, /* (unexpected continuation byte) */ |
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 | ERROR, /* (unexpected continuation byte) */ |
454 | ERROR, /* (unexpected continuation byte) */ |
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 | 0, -1); |
501 | utf8_read_test(TESTSTR("\xC0\x20\xC1\x20\xC2\x20\xC3\x20\xC4\x20\xC5\x20\xC6\x20\xC7\x20"), |
502 | ERROR, /* (incomplete sequence) */ |
503 | 0x00000020, /* SPACE */ |
504 | ERROR, /* (incomplete sequence) */ |
505 | 0x00000020, /* SPACE */ |
506 | ERROR, /* (incomplete sequence) */ |
507 | 0x00000020, /* SPACE */ |
508 | ERROR, /* (incomplete sequence) */ |
509 | 0x00000020, /* SPACE */ |
510 | ERROR, /* (incomplete sequence) */ |
511 | 0x00000020, /* SPACE */ |
512 | ERROR, /* (incomplete sequence) */ |
513 | 0x00000020, /* SPACE */ |
514 | ERROR, /* (incomplete sequence) */ |
515 | 0x00000020, /* SPACE */ |
516 | ERROR, /* (incomplete sequence) */ |
517 | 0x00000020, /* SPACE */ |
518 | 0, -1); |
519 | 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"), |
520 | ERROR, /* (incomplete sequence) */ |
521 | 0x00000020, /* SPACE */ |
522 | ERROR, /* (incomplete sequence) */ |
523 | 0x00000020, /* SPACE */ |
524 | ERROR, /* (incomplete sequence) */ |
525 | 0x00000020, /* SPACE */ |
526 | ERROR, /* (incomplete sequence) */ |
527 | 0x00000020, /* SPACE */ |
528 | ERROR, /* (incomplete sequence) */ |
529 | 0x00000020, /* SPACE */ |
530 | ERROR, /* (incomplete sequence) */ |
531 | 0x00000020, /* SPACE */ |
532 | ERROR, /* (incomplete sequence) */ |
533 | 0x00000020, /* SPACE */ |
534 | ERROR, /* (incomplete sequence) */ |
535 | 0x00000020, /* SPACE */ |
536 | ERROR, /* (incomplete sequence) */ |
537 | 0x00000020, /* SPACE */ |
538 | ERROR, /* (incomplete sequence) */ |
539 | 0x00000020, /* SPACE */ |
540 | ERROR, /* (incomplete sequence) */ |
541 | 0x00000020, /* SPACE */ |
542 | ERROR, /* (incomplete sequence) */ |
543 | 0x00000020, /* SPACE */ |
544 | ERROR, /* (incomplete sequence) */ |
545 | 0x00000020, /* SPACE */ |
546 | ERROR, /* (incomplete sequence) */ |
547 | 0x00000020, /* SPACE */ |
548 | ERROR, /* (incomplete sequence) */ |
549 | 0x00000020, /* SPACE */ |
550 | ERROR, /* (incomplete sequence) */ |
551 | 0x00000020, /* SPACE */ |
552 | 0, -1); |
553 | utf8_read_test(TESTSTR("\xF0\x20\xF1\x20\xF2\x20\xF3\x20\xF4\x20\xF5\x20\xF6\x20\xF7\x20"), |
554 | ERROR, /* (incomplete sequence) */ |
555 | 0x00000020, /* SPACE */ |
556 | ERROR, /* (incomplete sequence) */ |
557 | 0x00000020, /* SPACE */ |
558 | ERROR, /* (incomplete sequence) */ |
559 | 0x00000020, /* SPACE */ |
560 | ERROR, /* (incomplete sequence) */ |
561 | 0x00000020, /* SPACE */ |
562 | ERROR, /* (incomplete sequence) */ |
563 | 0x00000020, /* SPACE */ |
564 | ERROR, /* (incomplete sequence) */ |
565 | 0x00000020, /* SPACE */ |
566 | ERROR, /* (incomplete sequence) */ |
567 | 0x00000020, /* SPACE */ |
568 | ERROR, /* (incomplete sequence) */ |
569 | 0x00000020, /* SPACE */ |
570 | 0, -1); |
571 | utf8_read_test(TESTSTR("\xF8\x20\xF9\x20\xFA\x20\xFB\x20"), |
572 | ERROR, /* (incomplete sequence) */ |
573 | 0x00000020, /* SPACE */ |
574 | ERROR, /* (incomplete sequence) */ |
575 | 0x00000020, /* SPACE */ |
576 | ERROR, /* (incomplete sequence) */ |
577 | 0x00000020, /* SPACE */ |
578 | ERROR, /* (incomplete sequence) */ |
579 | 0x00000020, /* SPACE */ |
580 | 0, -1); |
581 | utf8_read_test(TESTSTR("\xFC\x20\xFD\x20"), |
582 | ERROR, /* (incomplete sequence) */ |
583 | 0x00000020, /* SPACE */ |
584 | ERROR, /* (incomplete sequence) */ |
585 | 0x00000020, /* SPACE */ |
586 | 0, -1); |
587 | utf8_read_test(TESTSTR("\xC0"), |
588 | ERROR, /* (incomplete sequence) */ |
589 | 0, -1); |
590 | utf8_read_test(TESTSTR("\xE0\x80"), |
591 | ERROR, /* (incomplete sequence) */ |
592 | 0, -1); |
593 | utf8_read_test(TESTSTR("\xF0\x80\x80"), |
594 | ERROR, /* (incomplete sequence) */ |
595 | 0, -1); |
596 | utf8_read_test(TESTSTR("\xF8\x80\x80\x80"), |
597 | ERROR, /* (incomplete sequence) */ |
598 | 0, -1); |
599 | utf8_read_test(TESTSTR("\xFC\x80\x80\x80\x80"), |
600 | ERROR, /* (incomplete sequence) */ |
601 | 0, -1); |
602 | utf8_read_test(TESTSTR("\xDF"), |
603 | ERROR, /* (incomplete sequence) */ |
604 | 0, -1); |
605 | utf8_read_test(TESTSTR("\xEF\xBF"), |
606 | ERROR, /* (incomplete sequence) */ |
607 | 0, -1); |
608 | utf8_read_test(TESTSTR("\xF7\xBF\xBF"), |
609 | ERROR, /* (incomplete sequence) */ |
610 | 0, -1); |
611 | utf8_read_test(TESTSTR("\xFB\xBF\xBF\xBF"), |
612 | ERROR, /* (incomplete sequence) */ |
613 | 0, -1); |
614 | utf8_read_test(TESTSTR("\xFD\xBF\xBF\xBF\xBF"), |
615 | ERROR, /* (incomplete sequence) */ |
616 | 0, -1); |
617 | 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"), |
618 | ERROR, /* (incomplete sequence) */ |
619 | ERROR, /* (incomplete sequence) */ |
620 | ERROR, /* (incomplete sequence) */ |
621 | ERROR, /* (incomplete sequence) */ |
622 | ERROR, /* (incomplete sequence) */ |
623 | ERROR, /* (incomplete sequence) */ |
624 | ERROR, /* (incomplete sequence) */ |
625 | ERROR, /* (incomplete sequence) */ |
626 | ERROR, /* (incomplete sequence) */ |
627 | ERROR, /* (incomplete sequence) */ |
628 | 0, -1); |
629 | utf8_read_test(TESTSTR("\xFE"), |
630 | ERROR, /* (invalid UTF-8 byte) */ |
631 | 0, -1); |
632 | utf8_read_test(TESTSTR("\xFF"), |
633 | ERROR, /* (invalid UTF-8 byte) */ |
634 | 0, -1); |
635 | utf8_read_test(TESTSTR("\xFE\xFE\xFF\xFF"), |
636 | ERROR, /* (invalid UTF-8 byte) */ |
637 | ERROR, /* (invalid UTF-8 byte) */ |
638 | ERROR, /* (invalid UTF-8 byte) */ |
639 | ERROR, /* (invalid UTF-8 byte) */ |
640 | 0, -1); |
641 | utf8_read_test(TESTSTR("\xC0\xAF"), |
642 | ERROR, /* SOLIDUS (overlong form of 2F) */ |
643 | 0, -1); |
644 | utf8_read_test(TESTSTR("\xE0\x80\xAF"), |
645 | ERROR, /* SOLIDUS (overlong form of 2F) */ |
646 | 0, -1); |
647 | utf8_read_test(TESTSTR("\xF0\x80\x80\xAF"), |
648 | ERROR, /* SOLIDUS (overlong form of 2F) */ |
649 | 0, -1); |
650 | utf8_read_test(TESTSTR("\xF8\x80\x80\x80\xAF"), |
651 | ERROR, /* SOLIDUS (overlong form of 2F) */ |
652 | 0, -1); |
653 | utf8_read_test(TESTSTR("\xFC\x80\x80\x80\x80\xAF"), |
654 | ERROR, /* SOLIDUS (overlong form of 2F) */ |
655 | 0, -1); |
656 | utf8_read_test(TESTSTR("\xC1\xBF"), |
657 | ERROR, /* <control> (overlong form of 7F) */ |
658 | 0, -1); |
659 | utf8_read_test(TESTSTR("\xE0\x9F\xBF"), |
660 | ERROR, /* <no name available> (overlong form of DF BF) */ |
661 | 0, -1); |
662 | utf8_read_test(TESTSTR("\xF0\x8F\xBF\xBF"), |
663 | ERROR, /* <no name available> (overlong form of EF BF BF) (invalid char) */ |
664 | 0, -1); |
665 | utf8_read_test(TESTSTR("\xF8\x87\xBF\xBF\xBF"), |
666 | ERROR, /* <no name available> (overlong form of F7 BF BF BF) */ |
667 | 0, -1); |
668 | utf8_read_test(TESTSTR("\xFC\x83\xBF\xBF\xBF\xBF"), |
669 | ERROR, /* <no name available> (overlong form of FB BF BF BF BF) */ |
670 | 0, -1); |
671 | utf8_read_test(TESTSTR("\xC0\x80"), |
672 | ERROR, /* <control> (overlong form of 00) */ |
673 | 0, -1); |
674 | utf8_read_test(TESTSTR("\xE0\x80\x80"), |
675 | ERROR, /* <control> (overlong form of 00) */ |
676 | 0, -1); |
677 | utf8_read_test(TESTSTR("\xF0\x80\x80\x80"), |
678 | ERROR, /* <control> (overlong form of 00) */ |
679 | 0, -1); |
680 | utf8_read_test(TESTSTR("\xF8\x80\x80\x80\x80"), |
681 | ERROR, /* <control> (overlong form of 00) */ |
682 | 0, -1); |
683 | utf8_read_test(TESTSTR("\xFC\x80\x80\x80\x80\x80"), |
684 | ERROR, /* <control> (overlong form of 00) */ |
685 | 0, -1); |
686 | utf8_read_test(TESTSTR("\xED\xA0\x80"), |
687 | ERROR, /* <Non Private Use High Surrogate, First> (surrogate) */ |
688 | 0, -1); |
689 | utf8_read_test(TESTSTR("\xED\xAD\xBF"), |
690 | ERROR, /* <Non Private Use High Surrogate, Last> (surrogate) */ |
691 | 0, -1); |
692 | utf8_read_test(TESTSTR("\xED\xAE\x80"), |
693 | ERROR, /* <Private Use High Surrogate, First> (surrogate) */ |
694 | 0, -1); |
695 | utf8_read_test(TESTSTR("\xED\xAF\xBF"), |
696 | ERROR, /* <Private Use High Surrogate, Last> (surrogate) */ |
697 | 0, -1); |
698 | utf8_read_test(TESTSTR("\xED\xB0\x80"), |
699 | ERROR, /* <Low Surrogate, First> (surrogate) */ |
700 | 0, -1); |
701 | utf8_read_test(TESTSTR("\xED\xBE\x80"), |
702 | ERROR, /* <no name available> (surrogate) */ |
703 | 0, -1); |
704 | utf8_read_test(TESTSTR("\xED\xBF\xBF"), |
705 | ERROR, /* <Low Surrogate, Last> (surrogate) */ |
706 | 0, -1); |
707 | utf8_read_test(TESTSTR("\xED\xA0\x80\xED\xB0\x80"), |
708 | ERROR, /* <Non Private Use High Surrogate, First> (surrogate) */ |
709 | ERROR, /* <Low Surrogate, First> (surrogate) */ |
710 | 0, -1); |
711 | utf8_read_test(TESTSTR("\xED\xA0\x80\xED\xBF\xBF"), |
712 | ERROR, /* <Non Private Use High Surrogate, First> (surrogate) */ |
713 | ERROR, /* <Low Surrogate, Last> (surrogate) */ |
714 | 0, -1); |
715 | utf8_read_test(TESTSTR("\xED\xAD\xBF\xED\xB0\x80"), |
716 | ERROR, /* <Non Private Use High Surrogate, Last> (surrogate) */ |
717 | ERROR, /* <Low Surrogate, First> (surrogate) */ |
718 | 0, -1); |
719 | utf8_read_test(TESTSTR("\xED\xAD\xBF\xED\xBF\xBF"), |
720 | ERROR, /* <Non Private Use High Surrogate, Last> (surrogate) */ |
721 | ERROR, /* <Low Surrogate, Last> (surrogate) */ |
722 | 0, -1); |
723 | utf8_read_test(TESTSTR("\xED\xAE\x80\xED\xB0\x80"), |
724 | ERROR, /* <Private Use High Surrogate, First> (surrogate) */ |
725 | ERROR, /* <Low Surrogate, First> (surrogate) */ |
726 | 0, -1); |
727 | utf8_read_test(TESTSTR("\xED\xAE\x80\xED\xBF\xBF"), |
728 | ERROR, /* <Private Use High Surrogate, First> (surrogate) */ |
729 | ERROR, /* <Low Surrogate, Last> (surrogate) */ |
730 | 0, -1); |
731 | utf8_read_test(TESTSTR("\xED\xAF\xBF\xED\xB0\x80"), |
732 | ERROR, /* <Private Use High Surrogate, Last> (surrogate) */ |
733 | ERROR, /* <Low Surrogate, First> (surrogate) */ |
734 | 0, -1); |
735 | utf8_read_test(TESTSTR("\xED\xAF\xBF\xED\xBF\xBF"), |
736 | ERROR, /* <Private Use High Surrogate, Last> (surrogate) */ |
737 | ERROR, /* <Low Surrogate, Last> (surrogate) */ |
738 | 0, -1); |
739 | utf8_read_test(TESTSTR("\xEF\xBF\xBE"), |
740 | ERROR, /* <no name available> (invalid char) */ |
741 | 0, -1); |
742 | utf8_read_test(TESTSTR("\xEF\xBF\xBF"), |
743 | ERROR, /* <no name available> (invalid char) */ |
744 | 0, -1); |
745 | printf("read tests completed\n"); |
746 | printf("write tests beginning\n"); |
747 | { |
748 | const static long str[] = |
749 | {0x03BAL, 0x1F79L, 0x03C3L, 0x03BCL, 0x03B5L, 0}; |
750 | utf8_write_test(TESTSTR(str), |
751 | 0xCE, 0xBA, |
752 | 0xE1, 0xBD, 0xB9, |
753 | 0xCF, 0x83, |
754 | 0xCE, 0xBC, |
755 | 0xCE, 0xB5, |
756 | 0, -1); |
757 | } |
758 | { |
759 | const static long str[] = {0x0000L, 0}; |
760 | utf8_write_test(TESTSTR(str), |
761 | 0x00, |
762 | 0, -1); |
763 | } |
764 | { |
765 | const static long str[] = {0x0080L, 0}; |
766 | utf8_write_test(TESTSTR(str), |
767 | 0xC2, 0x80, |
768 | 0, -1); |
769 | } |
770 | { |
771 | const static long str[] = {0x0800L, 0}; |
772 | utf8_write_test(TESTSTR(str), |
773 | 0xE0, 0xA0, 0x80, |
774 | 0, -1); |
775 | } |
776 | { |
777 | const static long str[] = {0x00010000L, 0}; |
778 | utf8_write_test(TESTSTR(str), |
779 | 0xF0, 0x90, 0x80, 0x80, |
780 | 0, -1); |
781 | } |
782 | { |
783 | const static long str[] = {0x00200000L, 0}; |
784 | utf8_write_test(TESTSTR(str), |
785 | 0xF8, 0x88, 0x80, 0x80, 0x80, |
786 | 0, -1); |
787 | } |
788 | { |
789 | const static long str[] = {0x04000000L, 0}; |
790 | utf8_write_test(TESTSTR(str), |
791 | 0xFC, 0x84, 0x80, 0x80, 0x80, 0x80, |
792 | 0, -1); |
793 | } |
794 | { |
795 | const static long str[] = {0x007FL, 0}; |
796 | utf8_write_test(TESTSTR(str), |
797 | 0x7F, |
798 | 0, -1); |
799 | } |
800 | { |
801 | const static long str[] = {0x07FFL, 0}; |
802 | utf8_write_test(TESTSTR(str), |
803 | 0xDF, 0xBF, |
804 | 0, -1); |
805 | } |
806 | { |
807 | const static long str[] = {0xFFFDL, 0}; |
808 | utf8_write_test(TESTSTR(str), |
809 | 0xEF, 0xBF, 0xBD, |
810 | 0, -1); |
811 | } |
812 | { |
813 | const static long str[] = {0xFFFFL, 0}; |
814 | utf8_write_test(TESTSTR(str), |
815 | ERROR, |
816 | 0, -1); |
817 | } |
818 | { |
819 | const static long str[] = {0x001FFFFFL, 0}; |
820 | utf8_write_test(TESTSTR(str), |
821 | 0xF7, 0xBF, 0xBF, 0xBF, |
822 | 0, -1); |
823 | } |
824 | { |
825 | const static long str[] = {0x03FFFFFFL, 0}; |
826 | utf8_write_test(TESTSTR(str), |
827 | 0xFB, 0xBF, 0xBF, 0xBF, 0xBF, |
828 | 0, -1); |
829 | } |
830 | { |
831 | const static long str[] = {0x7FFFFFFFL, 0}; |
832 | utf8_write_test(TESTSTR(str), |
833 | 0xFD, 0xBF, 0xBF, 0xBF, 0xBF, 0xBF, |
834 | 0, -1); |
835 | } |
836 | { |
837 | const static long str[] = {0xD7FFL, 0}; |
838 | utf8_write_test(TESTSTR(str), |
839 | 0xED, 0x9F, 0xBF, |
840 | 0, -1); |
841 | } |
842 | { |
843 | const static long str[] = {0xD800L, 0}; |
844 | utf8_write_test(TESTSTR(str), |
845 | ERROR, |
846 | 0, -1); |
847 | } |
848 | { |
849 | const static long str[] = {0xD800L, 0xDC00L, 0}; |
850 | utf8_write_test(TESTSTR(str), |
851 | ERROR, |
852 | ERROR, |
853 | 0, -1); |
854 | } |
855 | { |
856 | const static long str[] = {0xDFFFL, 0}; |
857 | utf8_write_test(TESTSTR(str), |
858 | ERROR, |
859 | 0, -1); |
860 | } |
861 | { |
862 | const static long str[] = {0xE000L, 0}; |
863 | utf8_write_test(TESTSTR(str), |
864 | 0xEE, 0x80, 0x80, |
865 | 0, -1); |
866 | } |
867 | printf("write tests completed\n"); |
868 | |
869 | printf("total: %d errors\n", total_errs); |
870 | return (total_errs != 0); |
871 | } |
872 | #endif /* TESTMODE */ |
873 | |
874 | const charset_spec charset_CS_UTF8 = { |
875 | CS_UTF8, read_utf8, write_utf8, NULL |
876 | }; |
877 | |
878 | #else /* ENUM_CHARSETS */ |
879 | |
880 | ENUM_CHARSET(CS_UTF8) |
881 | |
882 | #endif /* ENUM_CHARSETS */ |