1 /* $NetBSD: citrus_utf7.c,v 1.7 2022/04/19 20:32:14 rillig Exp $ */ 2 3 /*- 4 * Copyright (c)2004, 2005 Citrus Project, 5 * All rights reserved. 6 * 7 * Redistribution and use in source and binary forms, with or without 8 * modification, are permitted provided that the following conditions 9 * are met: 10 * 1. Redistributions of source code must retain the above copyright 11 * notice, this list of conditions and the following disclaimer. 12 * 2. Redistributions in binary form must reproduce the above copyright 13 * notice, this list of conditions and the following disclaimer in the 14 * documentation and/or other materials provided with the distribution. 15 * 16 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND 17 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 18 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 19 * ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE 20 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 21 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 22 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 23 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 24 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 25 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 26 * SUCH DAMAGE. 27 * 28 */ 29 30 #include <sys/cdefs.h> 31 #if defined(LIB_SCCS) && !defined(lint) 32 __RCSID("$NetBSD: citrus_utf7.c,v 1.7 2022/04/19 20:32:14 rillig Exp $"); 33 #endif /* LIB_SCCS and not lint */ 34 35 #include <assert.h> 36 #include <errno.h> 37 #include <string.h> 38 #include <stdio.h> 39 #include <stdint.h> 40 #include <stdlib.h> 41 #include <limits.h> 42 #include <wchar.h> 43 44 #include "citrus_namespace.h" 45 #include "citrus_types.h" 46 #include "citrus_module.h" 47 #include "citrus_ctype.h" 48 #include "citrus_stdenc.h" 49 #include "citrus_utf7.h" 50 51 /* ---------------------------------------------------------------------- 52 * private stuffs used by templates 53 */ 54 55 typedef struct { 56 uint16_t cell[0x80]; 57 #define EI_MASK UINT16_C(0xff) 58 #define EI_DIRECT UINT16_C(0x100) 59 #define EI_OPTION UINT16_C(0x200) 60 #define EI_SPACE UINT16_C(0x400) 61 } _UTF7EncodingInfo; 62 63 typedef struct { 64 unsigned int 65 mode: 1, /* whether base64 mode */ 66 bits: 4, /* need to hold 0 - 15 */ 67 cache: 22, /* 22 = BASE64_BIT + UTF16_BIT */ 68 surrogate: 1; /* whether surrogate pair or not */ 69 int chlen; 70 char ch[4]; /* BASE64_IN, 3 * 6 = 18, most closed to UTF16_BIT */ 71 } _UTF7State; 72 73 typedef struct { 74 _UTF7EncodingInfo ei; 75 struct { 76 /* for future multi-locale facility */ 77 _UTF7State s_mblen; 78 _UTF7State s_mbrlen; 79 _UTF7State s_mbrtowc; 80 _UTF7State s_mbtowc; 81 _UTF7State s_mbsrtowcs; 82 _UTF7State s_mbsnrtowcs; 83 _UTF7State s_wcrtomb; 84 _UTF7State s_wcsrtombs; 85 _UTF7State s_wcsnrtombs; 86 _UTF7State s_wctomb; 87 } states; 88 } _UTF7CTypeInfo; 89 90 #define _CEI_TO_EI(_cei_) (&(_cei_)->ei) 91 #define _CEI_TO_STATE(_cei_, _func_) (_cei_)->states.s_##_func_ 92 93 #define _FUNCNAME(m) _citrus_UTF7_##m 94 #define _ENCODING_INFO _UTF7EncodingInfo 95 #define _CTYPE_INFO _UTF7CTypeInfo 96 #define _ENCODING_STATE _UTF7State 97 #define _ENCODING_MB_CUR_MAX(_ei_) 4 98 #define _ENCODING_IS_STATE_DEPENDENT 1 99 #define _STATE_NEEDS_EXPLICIT_INIT(_ps_) 0 100 101 static __inline void 102 /*ARGSUSED*/ 103 _citrus_UTF7_init_state(_UTF7EncodingInfo * __restrict ei, 104 _UTF7State * __restrict s) 105 { 106 /* ei appears to be unused */ 107 _DIAGASSERT(s != NULL); 108 109 memset((void *)s, 0, sizeof(*s)); 110 } 111 112 static __inline void 113 /*ARGSUSED*/ 114 _citrus_UTF7_pack_state(_UTF7EncodingInfo * __restrict ei, 115 void *__restrict pspriv, const _UTF7State * __restrict s) 116 { 117 /* ei seem to be unused */ 118 _DIAGASSERT(pspriv != NULL); 119 _DIAGASSERT(s != NULL); 120 121 memcpy(pspriv, (const void *)s, sizeof(*s)); 122 } 123 124 static __inline void 125 /*ARGSUSED*/ 126 _citrus_UTF7_unpack_state(_UTF7EncodingInfo * __restrict ei, 127 _UTF7State * __restrict s, const void * __restrict pspriv) 128 { 129 /* ei seem to be unused */ 130 _DIAGASSERT(s != NULL); 131 _DIAGASSERT(pspriv != NULL); 132 133 memcpy((void *)s, pspriv, sizeof(*s)); 134 } 135 136 static const char base64[] = 137 "ABCDEFGHIJKLMNOPQRSTUVWXYZ" 138 "abcdefghijklmnopqrstuvwxyz" 139 "0123456789+/"; 140 141 static const char direct[] = 142 "ABCDEFGHIJKLMNOPQRSTUVWXYZ" 143 "abcdefghijklmnopqrstuvwxyz" 144 "0123456789(),-./:?"; 145 146 static const char option[] = "!\"#$%&';<=>@[]^_`{|}"; 147 static const char spaces[] = " \t\r\n"; 148 149 #define BASE64_BIT 6 150 #define UTF16_BIT 16 151 152 #define BASE64_MAX 0x3f 153 #define UTF16_MAX UINT16_C(0xffff) 154 #define UTF32_MAX UINT32_C(0x10ffff) 155 156 #define BASE64_IN '+' 157 #define BASE64_OUT '-' 158 159 #define SHIFT7BIT(c) ((c) >> 7) 160 #define ISSPECIAL(c) ((c) == '\0' || (c) == BASE64_IN) 161 162 #define FINDLEN(ei, c) \ 163 (SHIFT7BIT((c)) ? -1 : (((ei)->cell[(c)] & EI_MASK) - 1)) 164 165 #define ISDIRECT(ei, c) (!SHIFT7BIT((c)) && (ISSPECIAL((c)) || \ 166 ei->cell[(c)] & (EI_DIRECT | EI_OPTION | EI_SPACE))) 167 168 #define ISSAFE(ei, c) (!SHIFT7BIT((c)) && (ISSPECIAL((c)) || \ 169 (c < 0x80 && ei->cell[(c)] & (EI_DIRECT | EI_SPACE)))) 170 171 /* surrogate pair */ 172 #define SRG_BASE UINT32_C(0x10000) 173 #define HISRG_MIN UINT16_C(0xd800) 174 #define HISRG_MAX UINT16_C(0xdbff) 175 #define LOSRG_MIN UINT16_C(0xdc00) 176 #define LOSRG_MAX UINT16_C(0xdfff) 177 178 static int 179 _citrus_UTF7_mbtoutf16(_UTF7EncodingInfo * __restrict ei, 180 uint16_t * __restrict u16, const char ** __restrict s, size_t n, 181 _UTF7State * __restrict psenc, size_t * __restrict nresult) 182 { 183 _UTF7State sv; 184 const char *s0; 185 int i, done, len; 186 187 _DIAGASSERT(ei != NULL); 188 _DIAGASSERT(s != NULL && *s != NULL); 189 _DIAGASSERT(psenc != NULL); 190 191 s0 = *s; 192 sv = *psenc; 193 194 for (i = 0, done = 0; done == 0; i++) { 195 _DIAGASSERT(i <= psenc->chlen); 196 if (i == psenc->chlen) { 197 if (n-- < 1) { 198 *nresult = (size_t)-2; 199 *s = s0; 200 sv.chlen = psenc->chlen; 201 *psenc = sv; 202 return 0; 203 } 204 psenc->ch[psenc->chlen++] = *s0++; 205 } 206 if (SHIFT7BIT((int)psenc->ch[i])) 207 goto ilseq; 208 if (!psenc->mode) { 209 if (psenc->bits > 0 || psenc->cache > 0) 210 return EINVAL; 211 if (psenc->ch[i] == BASE64_IN) { 212 psenc->mode = 1; 213 } else { 214 if (!ISDIRECT(ei, (int)psenc->ch[i])) 215 goto ilseq; 216 *u16 = (uint16_t)psenc->ch[i]; 217 done = 1; 218 continue; 219 } 220 } else { 221 if (psenc->ch[i] == BASE64_OUT && psenc->cache == 0) { 222 psenc->mode = 0; 223 *u16 = (uint16_t)BASE64_IN; 224 done = 1; 225 continue; 226 } 227 len = FINDLEN(ei, (int)psenc->ch[i]); 228 if (len < 0) { 229 if (psenc->bits >= BASE64_BIT) 230 return EINVAL; 231 psenc->mode = 0; 232 psenc->bits = psenc->cache = 0; 233 if (psenc->ch[i] != BASE64_OUT) { 234 if (!ISDIRECT(ei, (int)psenc->ch[i])) 235 goto ilseq; 236 *u16 = (uint16_t)psenc->ch[i]; 237 done = 1; 238 } 239 } else { 240 psenc->cache = 241 (psenc->cache << BASE64_BIT) | len; 242 switch (psenc->bits) { 243 case 0: case 2: case 4: case 6: case 8: 244 psenc->bits += BASE64_BIT; 245 break; 246 case 10: case 12: case 14: 247 psenc->bits -= (UTF16_BIT - BASE64_BIT); 248 *u16 = (psenc->cache >> psenc->bits) 249 & UTF16_MAX; 250 done = 1; 251 break; 252 default: 253 return EINVAL; 254 } 255 } 256 } 257 } 258 259 if (psenc->chlen > i) 260 return EINVAL; 261 psenc->chlen = 0; 262 *nresult = (size_t)((*u16 == 0) ? 0 : s0 - *s); 263 *s = s0; 264 265 return 0; 266 267 ilseq: 268 *nresult = (size_t)-1; 269 return EILSEQ; 270 } 271 272 static int 273 _citrus_UTF7_mbrtowc_priv(_UTF7EncodingInfo * __restrict ei, 274 wchar_t * __restrict pwc, const char ** __restrict s, size_t n, 275 _UTF7State * __restrict psenc, size_t * __restrict nresult) 276 { 277 const char *s0; 278 uint32_t u32; 279 uint16_t hi, lo; 280 size_t siz, nr; 281 int err; 282 283 _DIAGASSERT(ei != NULL); 284 /* pwc may be null */ 285 _DIAGASSERT(s != NULL); 286 _DIAGASSERT(psenc != NULL); 287 288 if (*s == NULL) { 289 _citrus_UTF7_init_state(ei, psenc); 290 *nresult = (size_t)_ENCODING_IS_STATE_DEPENDENT; 291 return 0; 292 } 293 s0 = *s; 294 if (psenc->surrogate) { 295 hi = (psenc->cache >> 2) & UTF16_MAX; 296 if (hi < HISRG_MIN || hi > HISRG_MAX) 297 return EINVAL; 298 siz = 0; 299 } else { 300 err = _citrus_UTF7_mbtoutf16(ei, &hi, &s0, n, psenc, &nr); 301 if (nr == (size_t)-1 || nr == (size_t)-2) { 302 *nresult = nr; 303 return err; 304 } 305 if (err != 0) 306 return err; 307 n -= nr; 308 siz = nr; 309 if (hi < HISRG_MIN || hi > HISRG_MAX) { 310 u32 = (uint32_t)hi; 311 goto done; 312 } 313 psenc->surrogate = 1; 314 } 315 err = _citrus_UTF7_mbtoutf16(ei, &lo, &s0, n, psenc, &nr); 316 if (nr == (size_t)-1 || nr == (size_t)-2) { 317 *nresult = nr; 318 return err; 319 } 320 if (err != 0) 321 return err; 322 hi -= HISRG_MIN; 323 lo -= LOSRG_MIN; 324 u32 = (hi << 10 | lo) + SRG_BASE; 325 siz += nr; 326 done: 327 *s = s0; 328 if (pwc != NULL) 329 *pwc = (wchar_t)u32; 330 if (u32 == (uint32_t)0) { 331 *nresult = (size_t)0; 332 _citrus_UTF7_init_state(ei, psenc); 333 } else { 334 *nresult = siz; 335 psenc->surrogate = 0; 336 } 337 return err; 338 } 339 340 static int 341 _citrus_UTF7_utf16tomb(_UTF7EncodingInfo * __restrict ei, 342 char * __restrict s, size_t n, uint16_t u16, 343 _UTF7State * __restrict psenc, size_t * __restrict nresult) 344 { 345 int bits, i; 346 347 _DIAGASSERT(ei != NULL); 348 _DIAGASSERT(psenc != NULL); 349 350 if (psenc->chlen != 0 || psenc->bits > BASE64_BIT) 351 return EINVAL; 352 353 if (ISSAFE(ei, u16)) { 354 if (psenc->mode) { 355 if (psenc->bits > 0) { 356 bits = BASE64_BIT - psenc->bits; 357 i = (psenc->cache << bits) & BASE64_MAX; 358 psenc->ch[psenc->chlen++] = base64[i]; 359 psenc->bits = psenc->cache = 0; 360 } 361 if (u16 == BASE64_OUT || FINDLEN(ei, u16) >= 0) 362 psenc->ch[psenc->chlen++] = BASE64_OUT; 363 psenc->mode = 0; 364 } 365 if (psenc->bits != 0) 366 return EINVAL; 367 psenc->ch[psenc->chlen++] = (char)u16; 368 if (u16 == BASE64_IN) 369 psenc->ch[psenc->chlen++] = BASE64_OUT; 370 } else { 371 if (!psenc->mode) { 372 if (psenc->bits > 0) 373 return EINVAL; 374 psenc->ch[psenc->chlen++] = BASE64_IN; 375 psenc->mode = 1; 376 } 377 psenc->cache = (psenc->cache << UTF16_BIT) | u16; 378 bits = UTF16_BIT + psenc->bits; 379 psenc->bits = bits % BASE64_BIT; 380 while ((bits -= BASE64_BIT) >= 0) { 381 i = (psenc->cache >> bits) & BASE64_MAX; 382 psenc->ch[psenc->chlen++] = base64[i]; 383 } 384 } 385 memcpy(s, psenc->ch, psenc->chlen); 386 *nresult = psenc->chlen; 387 psenc->chlen = 0; 388 389 return 0; 390 } 391 392 static int 393 _citrus_UTF7_wcrtomb_priv(_UTF7EncodingInfo * __restrict ei, 394 char * __restrict s, size_t n, wchar_t wchar, 395 _UTF7State * __restrict psenc, size_t * __restrict nresult) 396 { 397 uint32_t u32; 398 uint16_t u16[2]; 399 int err, len, i; 400 size_t siz, nr; 401 402 _DIAGASSERT(ei != NULL); 403 _DIAGASSERT(s != NULL); 404 _DIAGASSERT(psenc != NULL); 405 _DIAGASSERT(nresult != NULL); 406 407 u32 = (uint32_t)wchar; 408 if (u32 <= UTF16_MAX) { 409 u16[0] = (uint16_t)u32; 410 len = 1; 411 } else if (u32 <= UTF32_MAX) { 412 u32 -= SRG_BASE; 413 u16[0] = (u32 >> 10) + HISRG_MIN; 414 u16[1] = ((uint16_t)(u32 & UINT32_C(0x3ff))) + LOSRG_MIN; 415 len = 2; 416 } else { 417 *nresult = (size_t)-1; 418 return EILSEQ; 419 } 420 siz = 0; 421 for (i = 0; i < len; ++i) { 422 err = _citrus_UTF7_utf16tomb(ei, s, n, u16[i], psenc, &nr); 423 if (err != 0) 424 return err; /* XXX: state has been modified */ 425 s += nr; 426 n -= nr; 427 siz += nr; 428 } 429 *nresult = siz; 430 431 return 0; 432 } 433 434 static int 435 /* ARGSUSED */ 436 _citrus_UTF7_put_state_reset(_UTF7EncodingInfo * __restrict ei, 437 char * __restrict s, size_t n, _UTF7State * __restrict psenc, 438 size_t * __restrict nresult) 439 { 440 int bits, pos; 441 442 _DIAGASSERT(ei != NULL); 443 _DIAGASSERT(s != NULL); 444 _DIAGASSERT(psenc != NULL); 445 _DIAGASSERT(nresult != NULL); 446 447 if (psenc->chlen != 0 || psenc->bits > BASE64_BIT || psenc->surrogate) 448 return EINVAL; 449 450 if (psenc->mode) { 451 if (psenc->bits > 0) { 452 if (n-- < 1) 453 return E2BIG; 454 bits = BASE64_BIT - psenc->bits; 455 pos = (psenc->cache << bits) & BASE64_MAX; 456 psenc->ch[psenc->chlen++] = base64[pos]; 457 psenc->ch[psenc->chlen++] = BASE64_OUT; 458 psenc->bits = psenc->cache = 0; 459 } 460 psenc->mode = 0; 461 } 462 if (psenc->bits != 0) 463 return EINVAL; 464 if (n-- < 1) 465 return E2BIG; 466 467 _DIAGASSERT(n >= psenc->chlen); 468 *nresult = (size_t)psenc->chlen; 469 if (psenc->chlen > 0) { 470 memcpy(s, psenc->ch, psenc->chlen); 471 psenc->chlen = 0; 472 } 473 474 return 0; 475 } 476 477 static __inline int 478 /*ARGSUSED*/ 479 _citrus_UTF7_stdenc_wctocs(_UTF7EncodingInfo * __restrict ei, 480 _csid_t * __restrict csid, 481 _index_t * __restrict idx, wchar_t wc) 482 { 483 /* ei seem to be unused */ 484 _DIAGASSERT(csid != NULL); 485 _DIAGASSERT(idx != NULL); 486 487 *csid = 0; 488 *idx = (_index_t)wc; 489 490 return 0; 491 } 492 493 static __inline int 494 /*ARGSUSED*/ 495 _citrus_UTF7_stdenc_cstowc(_UTF7EncodingInfo * __restrict ei, 496 wchar_t * __restrict wc, 497 _csid_t csid, _index_t idx) 498 { 499 /* ei seem to be unused */ 500 _DIAGASSERT(wc != NULL); 501 502 if (csid != 0) 503 return EILSEQ; 504 *wc = (wchar_t)idx; 505 506 return 0; 507 } 508 509 static __inline int 510 /*ARGSUSED*/ 511 _citrus_UTF7_stdenc_get_state_desc_generic(_UTF7EncodingInfo * __restrict ei, 512 _UTF7State * __restrict psenc, 513 int * __restrict rstate) 514 { 515 516 if (psenc->chlen == 0) 517 *rstate = _STDENC_SDGEN_INITIAL; 518 else 519 *rstate = _STDENC_SDGEN_INCOMPLETE_CHAR; 520 521 return 0; 522 } 523 524 static void 525 /*ARGSUSED*/ 526 _citrus_UTF7_encoding_module_uninit(_UTF7EncodingInfo *ei) 527 { 528 /* ei seems to be unused */ 529 } 530 531 static int 532 /*ARGSUSED*/ 533 _citrus_UTF7_encoding_module_init(_UTF7EncodingInfo * __restrict ei, 534 const void * __restrict var, size_t lenvar) 535 { 536 const char *s; 537 538 _DIAGASSERT(ei != NULL); 539 /* var may be null */ 540 541 memset(ei, 0, sizeof(*ei)); 542 543 #define FILL(str, flag) \ 544 do { \ 545 for (s = str; *s != '\0'; s++) \ 546 ei->cell[*s & 0x7f] |= flag; \ 547 } while (0) 548 549 FILL(base64, (s - base64) + 1); 550 FILL(direct, EI_DIRECT); 551 FILL(option, EI_OPTION); 552 FILL(spaces, EI_SPACE); 553 554 return 0; 555 } 556 557 /* ---------------------------------------------------------------------- 558 * public interface for ctype 559 */ 560 561 _CITRUS_CTYPE_DECLS(UTF7); 562 _CITRUS_CTYPE_DEF_OPS(UTF7); 563 564 #include "citrus_ctype_template.h" 565 566 /* ---------------------------------------------------------------------- 567 * public interface for stdenc 568 */ 569 570 _CITRUS_STDENC_DECLS(UTF7); 571 _CITRUS_STDENC_DEF_OPS(UTF7); 572 573 #include "citrus_stdenc_template.h" 574