b97e5427 |
1 | /* |
2 | * iso2022.c - support for ISO/IEC 2022 (alias ECMA-35). |
3 | * |
4 | * This isn't a complete implementation of ISO/IEC 2022, but it's |
5 | * close. It only handles decoding, because a fully general encoder |
6 | * isn't really useful. It can decode 8-bit and 7-bit versions, with |
7 | * support for single-byte and multi-byte character sets, all four |
8 | * containers (G0, G1, G2, and G3), using both single-shift and |
9 | * locking-shift sequences. |
10 | * |
11 | * The general principle is that any valid ISO/IEC 2022 sequence |
12 | * should either be correctly decoded or should emit an ERROR. The |
13 | * only exception to this is that the C0 and C1 sets are fixed as |
14 | * those of ISO/IEC 6429. Escape sequences for designating control |
15 | * sets are passed through, so a post-processor could fix them up if |
16 | * necessary. |
17 | * |
18 | * DOCS is not currently supported. It will be one day. |
19 | */ |
20 | |
21 | #ifndef ENUM_CHARSETS |
22 | |
23 | #include <assert.h> |
24 | |
25 | #include "charset.h" |
26 | #include "internal.h" |
27 | #include "sbcsdat.h" |
28 | |
29 | #define LS1 (0x0E) |
30 | #define LS0 (0x0F) |
31 | #define ESC (0x1B) |
32 | #define SS2 (0x8E) |
33 | #define SS3 (0x8F) |
34 | |
35 | enum {S4, S6, M4, M6}; |
36 | |
37 | static long int null_dbcs_to_unicode(int, int); |
38 | |
39 | const struct iso2022_subcharset { |
40 | char type, i, f; |
41 | int offset; |
42 | const sbcs_data *sbcs_base; |
43 | long int (*dbcs_fn)(int, int); |
44 | } iso2022_subcharsets[] = { |
45 | { S4, 0, 'B', 0x00, &sbcsdata_CS_ASCII }, |
46 | |
47 | { S4, 0, '<', 0x80, &sbcsdata_CS_DEC_MCS }, |
48 | { S4, 0, 'I', 0x80, &sbcsdata_CS_JISX0201 }, |
49 | { S4, 0, 'J', 0x00, &sbcsdata_CS_JISX0201 }, |
50 | { S4, 0, '~' }, |
51 | { S6, 0, 'A', 0x80, &sbcsdata_CS_ISO8859_1 }, |
52 | { S6, 0, 'B', 0x80, &sbcsdata_CS_ISO8859_2 }, |
53 | { S6, 0, 'C', 0x80, &sbcsdata_CS_ISO8859_3 }, |
54 | { S6, 0, 'D', 0x80, &sbcsdata_CS_ISO8859_4 }, |
55 | { S6, 0, 'F', 0x80, &sbcsdata_CS_ISO8859_7 }, |
56 | { S6, 0, 'G', 0x80, &sbcsdata_CS_ISO8859_6 }, |
57 | { S6, 0, 'H', 0x80, &sbcsdata_CS_ISO8859_8 }, |
58 | { S6, 0, 'L', 0x80, &sbcsdata_CS_ISO8859_5 }, |
59 | { S6, 0, 'M', 0x80, &sbcsdata_CS_ISO8859_9 }, |
60 | { S6, 0, 'T', 0x80, &sbcsdata_CS_ISO8859_11 }, |
61 | { S6, 0, 'V', 0x80, &sbcsdata_CS_ISO8859_10 }, |
62 | { S6, 0, 'Y', 0x80, &sbcsdata_CS_ISO8859_13 }, |
63 | { S6, 0, '_', 0x80, &sbcsdata_CS_ISO8859_14 }, |
64 | { S6, 0, 'b', 0x80, &sbcsdata_CS_ISO8859_15 }, |
65 | { S6, 0, 'f', 0x80, &sbcsdata_CS_ISO8859_16 }, |
66 | { S6, 0, '~' }, /* empty 96-set */ |
67 | #if 0 |
68 | { M4, 0, '@' }, /* JIS C 6226-1978 */ |
69 | #endif |
70 | { M4, 0, 'A', -0x21, 0, &gb2312_to_unicode }, |
71 | { M4, 0, 'B', -0x21, 0, &jisx0208_to_unicode }, |
72 | { M4, 0, 'C', -0x21, 0, &ksx1001_to_unicode }, |
73 | { M4, 0, 'D', -0x21, 0, &jisx0212_to_unicode }, |
74 | { M4, 0, '~', 0, 0, &null_dbcs_to_unicode }, /* empty 94^n-set */ |
75 | { M6, 0, '~', 0, 0, &null_dbcs_to_unicode }, /* empty 96^n-set */ |
76 | }; |
77 | |
78 | static long int null_dbcs_to_unicode(int r, int c) |
79 | { |
80 | return ERROR; |
81 | } |
82 | |
83 | /* States, or "what we're currently accumulating". */ |
84 | enum { |
85 | IDLE, /* None of the below */ |
86 | SS2CHAR, /* Accumulating a character after SS2 */ |
87 | SS3CHAR, /* Accumulating a character after SS3 */ |
88 | ESCSEQ, /* Accumulating an escape sequence */ |
89 | ESCDROP, /* Discarding an escape sequence */ |
90 | ESCPASS /* Passing through an escape sequence */ |
91 | }; |
92 | |
93 | #if 0 |
94 | #include <stdio.h> |
95 | static void dump_state(charset_state *s) |
96 | { |
97 | unsigned s0 = s->s0, s1 = s->s1; |
98 | char const * const modes[] = { "IDLE", "SS2CHAR", "SS3CHAR", |
99 | "ESCSEQ", "ESCDROP", "ESCPASS" }; |
100 | |
101 | fprintf(stderr, "s0: %s", modes[s0 >> 29]); |
102 | fprintf(stderr, " %02x %02x %02x ", (s0 >> 16) & 0xff, (s0 >> 8) & 0xff, |
103 | s0 & 0xff); |
104 | fprintf(stderr, "s1: LS%d LS%dR", (s1 >> 30) & 3, (s1 >> 28) & 3); |
105 | fprintf(stderr, " %d %d %d %d\n", s1 & 0x7f, (s1 >> 7) & 0x7f, |
106 | (s1 >> 14) & 0x7f, (s1 >> 21) & 0x7f); |
107 | } |
108 | #endif |
109 | |
110 | static void designate(charset_state *state, int container, |
111 | int type, int ibyte, int fbyte) |
112 | { |
113 | unsigned long i; |
114 | |
115 | assert(container >= 0 && container <= 3); |
116 | assert(type == S4 || type == S6 || type == M4 || type == M6); |
117 | |
118 | for (i = 0; i <= lenof(iso2022_subcharsets); i++) { |
119 | if (iso2022_subcharsets[i].type == type && |
120 | iso2022_subcharsets[i].i == ibyte && |
121 | iso2022_subcharsets[i].f == fbyte) { |
122 | state->s1 &= ~(0x7fL << (container * 7)); |
123 | state->s1 |= (i << (container * 7)); |
124 | return; |
125 | } |
126 | } |
127 | /* |
128 | * If we don't find the charset, invoke the empty one, so we |
129 | * output ERROR rather than garbage. |
130 | */ |
131 | designate(state, container, type, 0, '~'); |
132 | } |
133 | |
134 | static void read_iso2022(charset_spec const *charset, long int input_chr, |
135 | charset_state *state, |
136 | void (*emit)(void *ctx, long int output), |
137 | void *emitctx) |
138 | { |
139 | |
140 | /* dump_state(state); */ |
141 | /* |
142 | * We've got 64 bits of state to play with. |
143 | * |
144 | * Locking-shift state: 2 bits each GL/GR |
145 | * Single-shift state: 2 bits |
146 | * Charset designation state: n bits each G0/G1/G2/G3 |
147 | * MBCS/esc seq accumulation: 14 bits (assume max 4-byte sets) |
148 | * MBCS state: 2 bits (off, ESC, GL, GR) |
149 | * For no good reason, put long-term state in s1, short term in s0. |
150 | * |
151 | * s0[31:29] = state enum |
152 | * s0[24:0] = accumulated bytes |
153 | * s1[31:30] = GL locking-shift state |
154 | * s1[29:28] = GR locking-shift state |
155 | * s1[27:21] = G3 charset |
156 | * s1[20:14] = G2 charset |
157 | * s1[13:7] = G1 charset |
158 | * s1[6:0] = G0 charset |
159 | */ |
160 | |
161 | #define LEFT 30 |
162 | #define RIGHT 28 |
163 | #define LOCKING_SHIFT(n,side) \ |
164 | (state->s1 = (state->s1 & ~(3L<<(side))) | ((n ## L)<<(side))) |
165 | #define MODE ((state->s0 & 0xe0000000L) >> 29) |
166 | #define ENTER_MODE(m) (state->s0 = (state->s0 & ~0xe0000000L) | ((m)<<29)) |
167 | #define SINGLE_SHIFT(n) ENTER_MODE(SS2CHAR - 2 + (n)) |
168 | #define ASSERT_IDLE do { \ |
169 | if (state->s0 != 0) emit(emitctx, ERROR); \ |
170 | state->s0 = 0; \ |
171 | } while (0) |
172 | |
173 | if (state->s1 == 0) { |
174 | /* |
175 | * Since there's no LS0R, this means we must just have started. |
176 | * Set up a sane initial state (LS0, LS1R, ASCII in G0/G1/G2/G3). |
177 | */ |
178 | LOCKING_SHIFT(0, LEFT); |
179 | LOCKING_SHIFT(1, RIGHT); |
180 | designate(state, 0, S4, 0, 'B'); |
181 | designate(state, 1, S4, 0, 'B'); |
182 | designate(state, 2, S4, 0, 'B'); |
183 | designate(state, 3, S4, 0, 'B'); |
184 | } |
185 | |
186 | if ((input_chr & 0x60) == 0x00) { |
187 | /* C0 or C1 control */ |
188 | ASSERT_IDLE; |
189 | switch (input_chr) { |
190 | case ESC: |
191 | ENTER_MODE(ESCSEQ); |
192 | break; |
193 | case LS0: |
194 | LOCKING_SHIFT(0, LEFT); |
195 | break; |
196 | case LS1: |
197 | LOCKING_SHIFT(1, LEFT); |
198 | break; |
199 | case SS2: |
200 | SINGLE_SHIFT(2); |
201 | break; |
202 | case SS3: |
203 | SINGLE_SHIFT(3); |
204 | break; |
205 | default: |
206 | emit(emitctx, input_chr); |
207 | break; |
208 | } |
209 | } else if ((input_chr & 0x80) || MODE < ESCSEQ) { |
210 | int is_gl = 0; |
211 | struct iso2022_subcharset const *subcs; |
212 | unsigned container; |
213 | long input_7bit; |
214 | /* |
215 | * Actual data. |
216 | * Force idle state if we're in mid escape sequence, or in a |
217 | * multi-byte character with a different top bit. |
218 | */ |
219 | if (MODE >= ESCSEQ || |
220 | ((state->s0 & 0x00ff0000L) != 0 && |
221 | (((state->s0 >> 16) ^ input_chr) & 0x80))) |
222 | ASSERT_IDLE; |
223 | if (MODE == SS2CHAR || MODE == SS3CHAR) /* Single-shift */ |
224 | container = MODE - SS2CHAR + 2; |
225 | else if (input_chr >= 0x80) /* GR */ |
226 | container = (state->s1 >> 28) & 3; |
227 | else { /* GL */ |
228 | container = state->s1 >> 30; |
229 | is_gl = 1; |
230 | } |
231 | input_7bit = input_chr & ~0x80; |
232 | subcs = &iso2022_subcharsets[(state->s1 >> (container * 7)) & 0x7f]; |
233 | if ((subcs->type == S4 || subcs->type == M4) && |
234 | (input_7bit == 0x20 || input_7bit == 0x7f)) { |
235 | /* characters not in 94-char set */ |
236 | if (is_gl) emit(emitctx, input_7bit); |
237 | else emit(emitctx, ERROR); |
238 | } else if (subcs->type == M4 || subcs->type == M6) { |
239 | if ((state->s0 & 0x00ff0000L) == 0) { |
240 | state->s0 |= input_chr << 16; |
241 | return; |
242 | } else { |
243 | emit(emitctx, |
244 | subcs->dbcs_fn(((state->s0 >> 16) & 0x7f) + subcs->offset, |
245 | input_7bit + subcs->offset)); |
246 | } |
247 | } else { |
248 | if ((state->s0 & 0x00ff0000L) != 0) |
249 | emit(emitctx, ERROR); |
250 | emit(emitctx, subcs->sbcs_base ? |
251 | sbcs_to_unicode(subcs->sbcs_base, input_7bit + subcs->offset): |
252 | ERROR); |
253 | } |
254 | state->s0 = 0; |
255 | } else { |
256 | unsigned i1, i2; |
257 | if (MODE == ESCPASS) { |
258 | emit(emitctx, input_chr); |
259 | if ((input_chr & 0xf0) != 0x20) |
260 | ENTER_MODE(IDLE); |
261 | return; |
262 | } |
263 | |
264 | /* |
265 | * Intermediate bytes shall be any of the 16 positions of |
266 | * column 02 of the code table; they are denoted by the symbol |
267 | * I. |
268 | */ |
269 | if ((input_chr & 0xf0) == 0x20) { |
270 | if (((state->s0 >> 16) & 0xff) == 0) |
271 | state->s0 |= input_chr << 16; |
272 | else if (((state->s0 >> 8) & 0xff) == 0) |
273 | state->s0 |= input_chr << 8; |
274 | else { |
275 | /* Long escape sequence. Switch to ESCPASS or ESCDROP. */ |
276 | i1 = (state->s0 >> 16) & 0xff; |
277 | i2 = (state->s0 >> 8) & 0xff; |
278 | switch (i1) { |
279 | case '(': case ')': case '*': case '+': |
280 | case '-': case '.': case '/': |
281 | case '$': |
282 | ENTER_MODE(ESCDROP); |
283 | break; |
284 | default: |
285 | emit(emitctx, ESC); |
286 | emit(emitctx, i1); |
287 | emit(emitctx, i2); |
288 | emit(emitctx, input_chr); |
289 | state->s0 = 0; |
290 | ENTER_MODE(ESCPASS); |
291 | break; |
292 | } |
293 | } |
294 | return; |
295 | } |
296 | |
297 | /* |
298 | * Final bytes shall be any of the 79 positions of columns 03 |
299 | * to 07 of the code table excluding position 07/15; they are |
300 | * denoted by the symbol F. |
301 | */ |
302 | i1 = (state->s0 >> 16) & 0xff; |
303 | i2 = (state->s0 >> 8) & 0xff; |
304 | if (MODE == ESCDROP) |
305 | input_chr = 0; /* Make sure it won't match. */ |
306 | state->s0 = 0; |
307 | switch (i1) { |
308 | case 0: /* No intermediate bytes */ |
309 | switch (input_chr) { |
310 | case 'N': /* SS2 */ |
311 | SINGLE_SHIFT(2); |
312 | break; |
313 | case 'O': /* SS3 */ |
314 | SINGLE_SHIFT(3); |
315 | break; |
316 | case 'n': /* LS2 */ |
317 | LOCKING_SHIFT(2, LEFT); |
318 | break; |
319 | case 'o': /* LS3 */ |
320 | LOCKING_SHIFT(3, LEFT); |
321 | break; |
322 | case '|': /* LS3R */ |
323 | LOCKING_SHIFT(3, RIGHT); |
324 | break; |
325 | case '}': /* LS2R */ |
326 | LOCKING_SHIFT(2, RIGHT); |
327 | break; |
328 | case '~': /* LS1R */ |
329 | LOCKING_SHIFT(1, RIGHT); |
330 | break; |
331 | default: |
332 | /* Unsupported escape sequence. Spit it back out. */ |
333 | emit(emitctx, ESC); |
334 | emit(emitctx, input_chr); |
335 | } |
336 | break; |
337 | case ' ': /* ACS */ |
338 | /* |
339 | * Various coding structure facilities specify that designating |
340 | * a code element also invokes it. As far as I can see, invoking |
341 | * it now will have the same practical effect, since those |
342 | * facilities also ban the use of locking shifts. |
343 | */ |
344 | switch (input_chr) { |
345 | case 'A': /* G0 element used and invoked into GL */ |
346 | LOCKING_SHIFT(0, LEFT); |
347 | break; |
348 | case 'C': /* G0 in GL, G1 in GR */ |
349 | case 'D': /* Ditto, at least for 8-bit codes */ |
350 | case 'L': /* ISO 4873 (ECMA-43) level 1 */ |
351 | case 'M': /* ISO 4873 (ECMA-43) level 2 */ |
352 | LOCKING_SHIFT(0, LEFT); |
353 | LOCKING_SHIFT(1, RIGHT); |
354 | break; |
355 | } |
356 | break; |
357 | case '&': /* IRR */ |
358 | /* |
359 | * IRR (Identify Revised Registration) is ignored here, |
360 | * since any revised registration must be |
361 | * upward-compatible with the old one, so either we'll |
362 | * support the new one or we'll emit ERROR when we run |
363 | * into a new character. In either case, there's nothing |
364 | * to be done here. |
365 | */ |
366 | break; |
367 | case '(': /* GZD4 */ case ')': /* G1D4 */ |
368 | case '*': /* G2D4 */ case '+': /* G3D4 */ |
369 | designate(state, i1 - '(', S4, i2, input_chr); |
370 | break; |
371 | case '-': /* G1D6 */ case '.': /* G2D6 */ case '/': /* G3D6 */ |
372 | designate(state, i1 - ',', S6, i2, input_chr); |
373 | break; |
374 | case '$': /* G?DM? */ |
375 | switch (i2) { |
376 | case 0: /* Obsolete version of GZDM4 */ |
377 | i2 = '('; |
378 | case '(': /* GZDM4 */ case ')': /* G1DM4 */ |
379 | case '*': /* G2DM4 */ case '+': /* G3DM4 */ |
380 | designate(state, i2 - '(', M4, 0, input_chr); |
381 | break; |
382 | case '-': /* G1DM6 */ |
383 | case '.': /* G2DM6 */ case '/': /* G3DM6 */ |
384 | designate(state, i2 - ',', M6, 0, input_chr); |
385 | break; |
386 | default: |
387 | emit(emitctx, ERROR); |
388 | break; |
389 | } |
390 | case '%': /* DOCS */ |
391 | /* FIXME */ |
392 | break; |
393 | default: |
394 | /* Unsupported nF escape sequence. Re-emit it. */ |
395 | emit(emitctx, ESC); |
396 | emit(emitctx, i1); |
397 | if (i2) emit(emitctx, i2); |
398 | emit(emitctx, input_chr); |
399 | break; |
400 | } |
401 | } |
402 | } |
403 | |
404 | const charset_spec charset_CS_ISO2022 = { |
405 | CS_ISO2022, read_iso2022, NULL, NULL |
406 | }; |
407 | |
408 | #ifdef TESTMODE |
409 | |
410 | #include <stdio.h> |
411 | #include <stdarg.h> |
412 | #include <string.h> |
413 | |
414 | int total_errs = 0; |
415 | |
416 | void iso2022_emit(void *ctx, long output) |
417 | { |
418 | wchar_t **p = (wchar_t **)ctx; |
419 | *(*p)++ = output; |
420 | } |
421 | |
422 | void iso2022_read_test(int line, char *input, int inlen, ...) |
423 | { |
424 | va_list ap; |
425 | wchar_t *p, str[512]; |
426 | int i; |
427 | charset_state state; |
428 | unsigned long l; |
429 | |
430 | state.s0 = state.s1 = 0; |
431 | p = str; |
432 | |
433 | for (i = 0; i < inlen; i++) |
434 | read_iso2022(NULL, input[i] & 0xFF, &state, iso2022_emit, &p); |
435 | |
436 | va_start(ap, inlen); |
437 | l = 0; |
438 | for (i = 0; i < p - str; i++) { |
439 | l = va_arg(ap, long int); |
440 | if (l == -1) { |
441 | printf("%d: correct string shorter than output\n", line); |
442 | total_errs++; |
443 | break; |
444 | } |
445 | if (l != str[i]) { |
446 | printf("%d: char %d came out as %08x, should be %08lx\n", |
447 | line, i, str[i], l); |
448 | total_errs++; |
449 | } |
450 | } |
451 | if (l != -1) { |
452 | l = va_arg(ap, long int); |
453 | if (l != -1) { |
454 | printf("%d: correct string longer than output\n", line); |
455 | total_errs++; |
456 | } |
457 | } |
458 | va_end(ap); |
459 | } |
460 | |
461 | /* Macro to concoct the first three parameters of iso2022_read_test. */ |
462 | #define TESTSTR(x) __LINE__, x, lenof(x) |
463 | |
464 | int main(void) |
465 | { |
466 | printf("read tests beginning\n"); |
467 | /* Simple test (Emacs sample text for Japanese, in ISO-2022-JP) */ |
468 | iso2022_read_test(TESTSTR("Japanese (\x1b$BF|K\\8l\x1b(B)\t" |
469 | "\x1b$B$3$s$K$A$O\x1b(B, " |
470 | "\x1b$B%3%s%K%A%O\x1b(B\n"), |
471 | 'J','a','p','a','n','e','s','e',' ','(', |
472 | 0x65E5, 0x672C, 0x8A9E, ')', '\t', |
473 | 0x3053, 0x3093, 0x306b, 0x3061, 0x306f, ',', ' ', |
474 | 0x30b3, 0x30f3, 0x30cb, 0x30c1, 0x30cf, '\n', 0, -1); |
475 | /* Same thing in EUC-JP (with designations, and half-width katakana) */ |
476 | iso2022_read_test(TESTSTR("\x1b$)B\x1b*I\x1b$+D" |
477 | "Japanese (\xc6\xfc\xcb\xdc\xb8\xec)\t" |
478 | "\xa4\xb3\xa4\xf3\xa4\xcb\xa4\xc1\xa4\xcf, " |
479 | "\x8e\xba\x8e\xdd\x8e\xc6\x8e\xc1\x8e\xca\n"), |
480 | 'J','a','p','a','n','e','s','e',' ','(', |
481 | 0x65E5, 0x672C, 0x8A9E, ')', '\t', |
482 | 0x3053, 0x3093, 0x306b, 0x3061, 0x306f, ',', ' ', |
483 | 0xff7a, 0xff9d, 0xff86, 0xff81, 0xff8a, '\n', 0, -1); |
484 | /* Multibyte single-shift */ |
485 | iso2022_read_test(TESTSTR("\x1b$)B\x1b*I\x1b$+D\x8f\"/!"), |
486 | 0x02D8, '!', 0, -1); |
487 | /* Non-existent SBCS */ |
488 | iso2022_read_test(TESTSTR("\x1b(!Zfnord\n"), |
489 | ERROR, ERROR, ERROR, ERROR, ERROR, '\n', 0, -1); |
490 | /* Pass-through of ordinary escape sequences, including a long one */ |
491 | iso2022_read_test(TESTSTR("\x1b""b\x1b#5\x1b#!!!5"), |
492 | 0x1B, 'b', 0x1B, '#', '5', |
493 | 0x1B, '#', '!', '!', '!', '5', 0, -1); |
494 | /* Non-existent DBCS (also 5-byte escape sequence) */ |
495 | iso2022_read_test(TESTSTR("\x1b$(!Bfnord!"), |
496 | ERROR, ERROR, ERROR, 0, -1); |
497 | /* Incomplete DB characters */ |
498 | iso2022_read_test(TESTSTR("\x1b$B(,(\x1b(BHi\x1b$B(,(\n"), |
499 | 0x2501, ERROR, 'H', 'i', 0x2501, ERROR, '\n', 0, -1); |
500 | iso2022_read_test(TESTSTR("\x1b$)B\x1b*I\x1b$+D\xa4""B"), |
501 | ERROR, 'B', 0, -1); |
502 | iso2022_read_test(TESTSTR("\x1b$)B\x1b*I\x1b$+D\x0e\x1b|$\xa2\xaf"), |
503 | ERROR, 0x02D8, 0, -1); |
504 | /* Incomplete escape sequence */ |
505 | iso2022_read_test(TESTSTR("\x1b\n"), ERROR, '\n', 0, -1); |
506 | iso2022_read_test(TESTSTR("\x1b-A\x1b~\x1b\xa1"), ERROR, 0xa1, 0, -1); |
507 | /* Incomplete single-shift */ |
508 | iso2022_read_test(TESTSTR("\x8e\n"), ERROR, '\n', 0, -1); |
509 | iso2022_read_test(TESTSTR("\x1b$*B\x8e(\n"), ERROR, '\n', 0, -1); |
510 | /* Corner cases (02/00 and 07/15) */ |
511 | iso2022_read_test(TESTSTR("\x1b(B\x20\x7f"), 0x20, 0x7f, 0, -1); |
512 | iso2022_read_test(TESTSTR("\x1b(I\x20\x7f"), 0x20, 0x7f, 0, -1); |
513 | iso2022_read_test(TESTSTR("\x1b$B\x20\x7f"), 0x20, 0x7f, 0, -1); |
514 | iso2022_read_test(TESTSTR("\x1b-A\x0e\x20\x7f"), 0xa0, 0xff, 0, -1); |
515 | iso2022_read_test(TESTSTR("\x1b$-~\x0e\x20\x7f"), ERROR, 0, -1); |
516 | iso2022_read_test(TESTSTR("\x1b)B\xa0\xff"), ERROR, ERROR, 0, -1); |
517 | iso2022_read_test(TESTSTR("\x1b)I\xa0\xff"), ERROR, ERROR, 0, -1); |
518 | iso2022_read_test(TESTSTR("\x1b$)B\xa0\xff"), ERROR, ERROR, 0, -1); |
519 | iso2022_read_test(TESTSTR("\x1b-A\x1b~\xa0\xff"), 0xa0, 0xff, 0, -1); |
520 | iso2022_read_test(TESTSTR("\x1b$-~\x1b~\xa0\xff"), ERROR, 0, -1); |
521 | /* Designate control sets */ |
522 | iso2022_read_test(TESTSTR("\x1b!@"), 0x1b, '!', '@', 0, -1); |
523 | printf("read tests completed\n"); |
524 | printf("total: %d errors\n", total_errs); |
525 | return (total_errs != 0); |
526 | } |
527 | |
528 | #endif /* TESTMODE */ |
529 | |
530 | #else /* ENUM_CHARSETS */ |
531 | |
532 | ENUM_CHARSET(CS_ISO2022) |
533 | |
534 | #endif |