1 /* 2 ******************************************************************************* 3 * 4 * Copyright (C) 2003-2014, International Business Machines 5 * Corporation and others. All Rights Reserved. 6 * 7 ******************************************************************************* 8 * file name: uidna.cpp 9 * encoding: US-ASCII 10 * tab size: 8 (not used) 11 * indentation:4 12 * 13 * created on: 2003feb1 14 * created by: Ram Viswanadha 15 */ 16 17 #include "unicode/utypes.h" 18 19 #if !UCONFIG_NO_IDNA 20 21 #include "unicode/uidna.h" 22 #include "unicode/ustring.h" 23 #include "unicode/usprep.h" 24 #include "punycode.h" 25 #include "ustr_imp.h" 26 #include "cmemory.h" 27 #include "uassert.h" 28 #include "sprpimpl.h" 29 30 /* it is official IDNA ACE Prefix is "xn--" */ 31 static const UChar ACE_PREFIX[] ={ 0x0078,0x006E,0x002d,0x002d } ; 32 #define ACE_PREFIX_LENGTH 4 33 34 #define MAX_LABEL_LENGTH 63 35 /* The Max length of the labels should not be more than MAX_LABEL_LENGTH */ 36 #define MAX_LABEL_BUFFER_SIZE 100 37 38 #define MAX_DOMAIN_NAME_LENGTH 255 39 /* The Max length of the domain names should not be more than MAX_DOMAIN_NAME_LENGTH */ 40 #define MAX_IDN_BUFFER_SIZE MAX_DOMAIN_NAME_LENGTH+1 41 42 #define LOWER_CASE_DELTA 0x0020 43 #define HYPHEN 0x002D 44 #define FULL_STOP 0x002E 45 #define CAPITAL_A 0x0041 46 #define CAPITAL_Z 0x005A 47 48 inline static UChar 49 toASCIILower(UChar ch){ 50 if(CAPITAL_A <= ch && ch <= CAPITAL_Z){ 51 return ch + LOWER_CASE_DELTA; 52 } 53 return ch; 54 } 55 56 inline static UBool 57 startsWithPrefix(const UChar* src , int32_t srcLength){ 58 UBool startsWithPrefix = TRUE; 59 60 if(srcLength < ACE_PREFIX_LENGTH){ 61 return FALSE; 62 } 63 64 for(int8_t i=0; i< ACE_PREFIX_LENGTH; i++){ 65 if(toASCIILower(src[i]) != ACE_PREFIX[i]){ 66 startsWithPrefix = FALSE; 67 } 68 } 69 return startsWithPrefix; 70 } 71 72 73 inline static int32_t 74 compareCaseInsensitiveASCII(const UChar* s1, int32_t s1Len, 75 const UChar* s2, int32_t s2Len){ 76 77 int32_t minLength; 78 int32_t lengthResult; 79 80 // are we comparing different lengths? 81 if(s1Len != s2Len) { 82 if(s1Len < s2Len) { 83 minLength = s1Len; 84 lengthResult = -1; 85 } else { 86 minLength = s2Len; 87 lengthResult = 1; 88 } 89 } else { 90 // ok the lengths are equal 91 minLength = s1Len; 92 lengthResult = 0; 93 } 94 95 UChar c1,c2; 96 int32_t rc; 97 98 for(int32_t i =0;/* no condition */;i++) { 99 100 /* If we reach the ends of both strings then they match */ 101 if(i == minLength) { 102 return lengthResult; 103 } 104 105 c1 = s1[i]; 106 c2 = s2[i]; 107 108 /* Case-insensitive comparison */ 109 if(c1!=c2) { 110 rc=(int32_t)toASCIILower(c1)-(int32_t)toASCIILower(c2); 111 if(rc!=0) { 112 lengthResult=rc; 113 break; 114 } 115 } 116 } 117 return lengthResult; 118 } 119 120 121 /** 122 * Ascertain if the given code point is a label separator as 123 * defined by the IDNA RFC 124 * 125 * @param ch The code point to be ascertained 126 * @return true if the char is a label separator 127 * @stable ICU 2.8 128 */ 129 static inline UBool isLabelSeparator(UChar ch){ 130 switch(ch){ 131 case 0x002e: 132 case 0x3002: 133 case 0xFF0E: 134 case 0xFF61: 135 return TRUE; 136 default: 137 return FALSE; 138 } 139 } 140 141 // returns the length of the label excluding the separator 142 // if *limit == separator then the length returned does not include 143 // the separtor. 144 static inline int32_t 145 getNextSeparator(UChar *src, int32_t srcLength, 146 UChar **limit, UBool *done){ 147 if(srcLength == -1){ 148 int32_t i; 149 for(i=0 ; ;i++){ 150 if(src[i] == 0){ 151 *limit = src + i; // point to null 152 *done = TRUE; 153 return i; 154 } 155 if(isLabelSeparator(src[i])){ 156 *limit = src + (i+1); // go past the delimiter 157 return i; 158 159 } 160 } 161 }else{ 162 int32_t i; 163 for(i=0;i<srcLength;i++){ 164 if(isLabelSeparator(src[i])){ 165 *limit = src + (i+1); // go past the delimiter 166 return i; 167 } 168 } 169 // we have not found the delimiter 170 // if(i==srcLength) 171 *limit = src+srcLength; 172 *done = TRUE; 173 174 return i; 175 } 176 } 177 static inline UBool isLDHChar(UChar ch){ 178 // high runner case 179 if(ch>0x007A){ 180 return FALSE; 181 } 182 //[\\u002D \\u0030-\\u0039 \\u0041-\\u005A \\u0061-\\u007A] 183 if( (ch==0x002D) || 184 (0x0030 <= ch && ch <= 0x0039) || 185 (0x0041 <= ch && ch <= 0x005A) || 186 (0x0061 <= ch && ch <= 0x007A) 187 ){ 188 return TRUE; 189 } 190 return FALSE; 191 } 192 193 static int32_t 194 _internal_toASCII(const UChar* src, int32_t srcLength, 195 UChar* dest, int32_t destCapacity, 196 int32_t options, 197 UStringPrepProfile* nameprep, 198 UParseError* parseError, 199 UErrorCode* status) 200 { 201 202 // TODO Revisit buffer handling. The label should not be over 63 ASCII characters. ICU4J may need to be updated too. 203 UChar b1Stack[MAX_LABEL_BUFFER_SIZE], b2Stack[MAX_LABEL_BUFFER_SIZE]; 204 //initialize pointers to stack buffers 205 UChar *b1 = b1Stack, *b2 = b2Stack; 206 int32_t b1Len=0, b2Len, 207 b1Capacity = MAX_LABEL_BUFFER_SIZE, 208 b2Capacity = MAX_LABEL_BUFFER_SIZE , 209 reqLength=0; 210 211 int32_t namePrepOptions = ((options & UIDNA_ALLOW_UNASSIGNED) != 0) ? USPREP_ALLOW_UNASSIGNED: 0; 212 UBool* caseFlags = NULL; 213 214 // the source contains all ascii codepoints 215 UBool srcIsASCII = TRUE; 216 // assume the source contains all LDH codepoints 217 UBool srcIsLDH = TRUE; 218 219 int32_t j=0; 220 221 //get the options 222 UBool useSTD3ASCIIRules = (UBool)((options & UIDNA_USE_STD3_RULES) != 0); 223 224 int32_t failPos = -1; 225 226 if(srcLength == -1){ 227 srcLength = u_strlen(src); 228 } 229 230 if(srcLength > b1Capacity){ 231 b1 = (UChar*) uprv_malloc(srcLength * U_SIZEOF_UCHAR); 232 if(b1==NULL){ 233 *status = U_MEMORY_ALLOCATION_ERROR; 234 goto CLEANUP; 235 } 236 b1Capacity = srcLength; 237 } 238 239 // step 1 240 for( j=0;j<srcLength;j++){ 241 if(src[j] > 0x7F){ 242 srcIsASCII = FALSE; 243 } 244 b1[b1Len++] = src[j]; 245 } 246 247 // step 2 is performed only if the source contains non ASCII 248 if(srcIsASCII == FALSE){ 249 250 // step 2 251 b1Len = usprep_prepare(nameprep, src, srcLength, b1, b1Capacity, namePrepOptions, parseError, status); 252 253 if(*status == U_BUFFER_OVERFLOW_ERROR){ 254 // redo processing of string 255 // we do not have enough room so grow the buffer 256 if(b1 != b1Stack){ 257 uprv_free(b1); 258 } 259 b1 = (UChar*) uprv_malloc(b1Len * U_SIZEOF_UCHAR); 260 if(b1==NULL){ 261 *status = U_MEMORY_ALLOCATION_ERROR; 262 goto CLEANUP; 263 } 264 265 *status = U_ZERO_ERROR; // reset error 266 267 b1Len = usprep_prepare(nameprep, src, srcLength, b1, b1Len, namePrepOptions, parseError, status); 268 } 269 } 270 // error bail out 271 if(U_FAILURE(*status)){ 272 goto CLEANUP; 273 } 274 if(b1Len == 0){ 275 *status = U_IDNA_ZERO_LENGTH_LABEL_ERROR; 276 goto CLEANUP; 277 } 278 279 // for step 3 & 4 280 srcIsASCII = TRUE; 281 for( j=0;j<b1Len;j++){ 282 // check if output of usprep_prepare is all ASCII 283 if(b1[j] > 0x7F){ 284 srcIsASCII = FALSE; 285 }else if(isLDHChar(b1[j])==FALSE){ // if the char is in ASCII range verify that it is an LDH character 286 srcIsLDH = FALSE; 287 failPos = j; 288 } 289 } 290 if(useSTD3ASCIIRules == TRUE){ 291 // verify 3a and 3b 292 // 3(a) Verify the absence of non-LDH ASCII code points; that is, the 293 // absence of 0..2C, 2E..2F, 3A..40, 5B..60, and 7B..7F. 294 // 3(b) Verify the absence of leading and trailing hyphen-minus; that 295 // is, the absence of U+002D at the beginning and end of the 296 // sequence. 297 if( srcIsLDH == FALSE /* source at this point should not contain anyLDH characters */ 298 || b1[0] == HYPHEN || b1[b1Len-1] == HYPHEN){ 299 *status = U_IDNA_STD3_ASCII_RULES_ERROR; 300 301 /* populate the parseError struct */ 302 if(srcIsLDH==FALSE){ 303 // failPos is always set the index of failure 304 uprv_syntaxError(b1,failPos, b1Len,parseError); 305 }else if(b1[0] == HYPHEN){ 306 // fail position is 0 307 uprv_syntaxError(b1,0,b1Len,parseError); 308 }else{ 309 // the last index in the source is always length-1 310 uprv_syntaxError(b1, (b1Len>0) ? b1Len-1 : b1Len, b1Len,parseError); 311 } 312 313 goto CLEANUP; 314 } 315 } 316 // Step 4: if the source is ASCII then proceed to step 8 317 if(srcIsASCII){ 318 if(b1Len <= destCapacity){ 319 uprv_memmove(dest, b1, b1Len * U_SIZEOF_UCHAR); 320 reqLength = b1Len; 321 }else{ 322 reqLength = b1Len; 323 goto CLEANUP; 324 } 325 }else{ 326 // step 5 : verify the sequence does not begin with ACE prefix 327 if(!startsWithPrefix(b1,b1Len)){ 328 329 //step 6: encode the sequence with punycode 330 331 // do not preserve the case flags for now! 332 // TODO: Preserve the case while implementing the RFE 333 // caseFlags = (UBool*) uprv_malloc(b1Len * sizeof(UBool)); 334 // uprv_memset(caseFlags,TRUE,b1Len); 335 336 b2Len = u_strToPunycode(b1,b1Len,b2,b2Capacity,caseFlags, status); 337 338 if(*status == U_BUFFER_OVERFLOW_ERROR){ 339 // redo processing of string 340 /* we do not have enough room so grow the buffer*/ 341 b2 = (UChar*) uprv_malloc(b2Len * U_SIZEOF_UCHAR); 342 if(b2 == NULL){ 343 *status = U_MEMORY_ALLOCATION_ERROR; 344 goto CLEANUP; 345 } 346 347 *status = U_ZERO_ERROR; // reset error 348 349 b2Len = u_strToPunycode(b1,b1Len,b2,b2Len,caseFlags, status); 350 } 351 //error bail out 352 if(U_FAILURE(*status)){ 353 goto CLEANUP; 354 } 355 // TODO : Reconsider while implementing the case preserve RFE 356 // convert all codepoints to lower case ASCII 357 // toASCIILower(b2,b2Len); 358 reqLength = b2Len+ACE_PREFIX_LENGTH; 359 360 if(reqLength > destCapacity){ 361 *status = U_BUFFER_OVERFLOW_ERROR; 362 goto CLEANUP; 363 } 364 //Step 7: prepend the ACE prefix 365 uprv_memcpy(dest,ACE_PREFIX,ACE_PREFIX_LENGTH * U_SIZEOF_UCHAR); 366 //Step 6: copy the contents in b2 into dest 367 uprv_memcpy(dest+ACE_PREFIX_LENGTH, b2, b2Len * U_SIZEOF_UCHAR); 368 369 }else{ 370 *status = U_IDNA_ACE_PREFIX_ERROR; 371 //position of failure is 0 372 uprv_syntaxError(b1,0,b1Len,parseError); 373 goto CLEANUP; 374 } 375 } 376 // step 8: verify the length of label 377 if(reqLength > MAX_LABEL_LENGTH){ 378 *status = U_IDNA_LABEL_TOO_LONG_ERROR; 379 } 380 381 CLEANUP: 382 if(b1 != b1Stack){ 383 uprv_free(b1); 384 } 385 if(b2 != b2Stack){ 386 uprv_free(b2); 387 } 388 uprv_free(caseFlags); 389 390 return u_terminateUChars(dest, destCapacity, reqLength, status); 391 } 392 393 static int32_t 394 _internal_toUnicode(const UChar* src, int32_t srcLength, 395 UChar* dest, int32_t destCapacity, 396 int32_t options, 397 UStringPrepProfile* nameprep, 398 UParseError* parseError, 399 UErrorCode* status) 400 { 401 402 //get the options 403 //UBool useSTD3ASCIIRules = (UBool)((options & UIDNA_USE_STD3_RULES) != 0); 404 int32_t namePrepOptions = ((options & UIDNA_ALLOW_UNASSIGNED) != 0) ? USPREP_ALLOW_UNASSIGNED: 0; 405 406 // TODO Revisit buffer handling. The label should not be over 63 ASCII characters. ICU4J may need to be updated too. 407 UChar b1Stack[MAX_LABEL_BUFFER_SIZE], b2Stack[MAX_LABEL_BUFFER_SIZE], b3Stack[MAX_LABEL_BUFFER_SIZE]; 408 409 //initialize pointers to stack buffers 410 UChar *b1 = b1Stack, *b2 = b2Stack, *b1Prime=NULL, *b3=b3Stack; 411 int32_t b1Len = 0, b2Len, b1PrimeLen, b3Len, 412 b1Capacity = MAX_LABEL_BUFFER_SIZE, 413 b2Capacity = MAX_LABEL_BUFFER_SIZE, 414 b3Capacity = MAX_LABEL_BUFFER_SIZE, 415 reqLength=0; 416 417 UBool* caseFlags = NULL; 418 419 UBool srcIsASCII = TRUE; 420 /*UBool srcIsLDH = TRUE; 421 int32_t failPos =0;*/ 422 423 // step 1: find out if all the codepoints in src are ASCII 424 if(srcLength==-1){ 425 srcLength = 0; 426 for(;src[srcLength]!=0;){ 427 if(src[srcLength]> 0x7f){ 428 srcIsASCII = FALSE; 429 }/*else if(isLDHChar(src[srcLength])==FALSE){ 430 // here we do not assemble surrogates 431 // since we know that LDH code points 432 // are in the ASCII range only 433 srcIsLDH = FALSE; 434 failPos = srcLength; 435 }*/ 436 srcLength++; 437 } 438 }else if(srcLength > 0){ 439 for(int32_t j=0; j<srcLength; j++){ 440 if(src[j]> 0x7f){ 441 srcIsASCII = FALSE; 442 }/*else if(isLDHChar(src[j])==FALSE){ 443 // here we do not assemble surrogates 444 // since we know that LDH code points 445 // are in the ASCII range only 446 srcIsLDH = FALSE; 447 failPos = j; 448 }*/ 449 } 450 }else{ 451 return 0; 452 } 453 454 if(srcIsASCII == FALSE){ 455 // step 2: process the string 456 b1Len = usprep_prepare(nameprep, src, srcLength, b1, b1Capacity, namePrepOptions, parseError, status); 457 if(*status == U_BUFFER_OVERFLOW_ERROR){ 458 // redo processing of string 459 /* we do not have enough room so grow the buffer*/ 460 b1 = (UChar*) uprv_malloc(b1Len * U_SIZEOF_UCHAR); 461 if(b1==NULL){ 462 *status = U_MEMORY_ALLOCATION_ERROR; 463 goto CLEANUP; 464 } 465 466 *status = U_ZERO_ERROR; // reset error 467 468 b1Len = usprep_prepare(nameprep, src, srcLength, b1, b1Len, namePrepOptions, parseError, status); 469 } 470 //bail out on error 471 if(U_FAILURE(*status)){ 472 goto CLEANUP; 473 } 474 }else{ 475 476 //just point src to b1 477 b1 = (UChar*) src; 478 b1Len = srcLength; 479 } 480 481 // The RFC states that 482 // <quote> 483 // ToUnicode never fails. If any step fails, then the original input 484 // is returned immediately in that step. 485 // </quote> 486 487 //step 3: verify ACE Prefix 488 if(startsWithPrefix(b1,b1Len)){ 489 490 //step 4: Remove the ACE Prefix 491 b1Prime = b1 + ACE_PREFIX_LENGTH; 492 b1PrimeLen = b1Len - ACE_PREFIX_LENGTH; 493 494 //step 5: Decode using punycode 495 b2Len = u_strFromPunycode(b1Prime, b1PrimeLen, b2, b2Capacity, caseFlags,status); 496 497 if(*status == U_BUFFER_OVERFLOW_ERROR){ 498 // redo processing of string 499 /* we do not have enough room so grow the buffer*/ 500 b2 = (UChar*) uprv_malloc(b2Len * U_SIZEOF_UCHAR); 501 if(b2==NULL){ 502 *status = U_MEMORY_ALLOCATION_ERROR; 503 goto CLEANUP; 504 } 505 506 *status = U_ZERO_ERROR; // reset error 507 508 b2Len = u_strFromPunycode(b1Prime, b1PrimeLen, b2, b2Len, caseFlags, status); 509 } 510 511 512 //step 6:Apply toASCII 513 b3Len = uidna_toASCII(b2, b2Len, b3, b3Capacity, options, parseError, status); 514 515 if(*status == U_BUFFER_OVERFLOW_ERROR){ 516 // redo processing of string 517 /* we do not have enough room so grow the buffer*/ 518 b3 = (UChar*) uprv_malloc(b3Len * U_SIZEOF_UCHAR); 519 if(b3==NULL){ 520 *status = U_MEMORY_ALLOCATION_ERROR; 521 goto CLEANUP; 522 } 523 524 *status = U_ZERO_ERROR; // reset error 525 526 b3Len = uidna_toASCII(b2,b2Len,b3,b3Len,options,parseError, status); 527 528 } 529 //bail out on error 530 if(U_FAILURE(*status)){ 531 goto CLEANUP; 532 } 533 534 //step 7: verify 535 if(compareCaseInsensitiveASCII(b1, b1Len, b3, b3Len) !=0){ 536 // Cause the original to be returned. 537 *status = U_IDNA_VERIFICATION_ERROR; 538 goto CLEANUP; 539 } 540 541 //step 8: return output of step 5 542 reqLength = b2Len; 543 if(b2Len <= destCapacity) { 544 uprv_memmove(dest, b2, b2Len * U_SIZEOF_UCHAR); 545 } 546 } 547 else{ 548 // See the start of this if statement for why this is commented out. 549 // verify that STD3 ASCII rules are satisfied 550 /*if(useSTD3ASCIIRules == TRUE){ 551 if( srcIsLDH == FALSE // source contains some non-LDH characters 552 || src[0] == HYPHEN || src[srcLength-1] == HYPHEN){ 553 *status = U_IDNA_STD3_ASCII_RULES_ERROR; 554 555 // populate the parseError struct 556 if(srcIsLDH==FALSE){ 557 // failPos is always set the index of failure 558 uprv_syntaxError(src,failPos, srcLength,parseError); 559 }else if(src[0] == HYPHEN){ 560 // fail position is 0 561 uprv_syntaxError(src,0,srcLength,parseError); 562 }else{ 563 // the last index in the source is always length-1 564 uprv_syntaxError(src, (srcLength>0) ? srcLength-1 : srcLength, srcLength,parseError); 565 } 566 567 goto CLEANUP; 568 } 569 }*/ 570 // just return the source 571 //copy the source to destination 572 if(srcLength <= destCapacity){ 573 uprv_memmove(dest,src,srcLength * U_SIZEOF_UCHAR); 574 } 575 reqLength = srcLength; 576 } 577 578 579 CLEANUP: 580 581 if(b1 != b1Stack && b1!=src){ 582 uprv_free(b1); 583 } 584 if(b2 != b2Stack){ 585 uprv_free(b2); 586 } 587 uprv_free(caseFlags); 588 589 // The RFC states that 590 // <quote> 591 // ToUnicode never fails. If any step fails, then the original input 592 // is returned immediately in that step. 593 // </quote> 594 // So if any step fails lets copy source to destination 595 if(U_FAILURE(*status)){ 596 //copy the source to destination 597 if(dest && srcLength <= destCapacity){ 598 // srcLength should have already been set earlier. 599 U_ASSERT(srcLength >= 0); 600 uprv_memmove(dest,src,srcLength * U_SIZEOF_UCHAR); 601 } 602 reqLength = srcLength; 603 *status = U_ZERO_ERROR; 604 } 605 606 return u_terminateUChars(dest, destCapacity, reqLength, status); 607 } 608 609 U_CAPI int32_t U_EXPORT2 610 uidna_toASCII(const UChar* src, int32_t srcLength, 611 UChar* dest, int32_t destCapacity, 612 int32_t options, 613 UParseError* parseError, 614 UErrorCode* status){ 615 616 if(status == NULL || U_FAILURE(*status)){ 617 return 0; 618 } 619 if((src==NULL) || (srcLength < -1) || (destCapacity<0) || (!dest && destCapacity > 0)){ 620 *status = U_ILLEGAL_ARGUMENT_ERROR; 621 return 0; 622 } 623 624 UStringPrepProfile* nameprep = usprep_openByType(USPREP_RFC3491_NAMEPREP, status); 625 626 if(U_FAILURE(*status)){ 627 return -1; 628 } 629 630 int32_t retLen = _internal_toASCII(src, srcLength, dest, destCapacity, options, nameprep, parseError, status); 631 632 /* close the profile*/ 633 usprep_close(nameprep); 634 635 return retLen; 636 } 637 638 U_CAPI int32_t U_EXPORT2 639 uidna_toUnicode(const UChar* src, int32_t srcLength, 640 UChar* dest, int32_t destCapacity, 641 int32_t options, 642 UParseError* parseError, 643 UErrorCode* status){ 644 645 if(status == NULL || U_FAILURE(*status)){ 646 return 0; 647 } 648 if( (src==NULL) || (srcLength < -1) || (destCapacity<0) || (!dest && destCapacity > 0)){ 649 *status = U_ILLEGAL_ARGUMENT_ERROR; 650 return 0; 651 } 652 653 UStringPrepProfile* nameprep = usprep_openByType(USPREP_RFC3491_NAMEPREP, status); 654 655 if(U_FAILURE(*status)){ 656 return -1; 657 } 658 659 int32_t retLen = _internal_toUnicode(src, srcLength, dest, destCapacity, options, nameprep, parseError, status); 660 661 usprep_close(nameprep); 662 663 return retLen; 664 } 665 666 667 U_CAPI int32_t U_EXPORT2 668 uidna_IDNToASCII( const UChar *src, int32_t srcLength, 669 UChar* dest, int32_t destCapacity, 670 int32_t options, 671 UParseError *parseError, 672 UErrorCode *status){ 673 674 if(status == NULL || U_FAILURE(*status)){ 675 return 0; 676 } 677 if((src==NULL) || (srcLength < -1) || (destCapacity<0) || (!dest && destCapacity > 0)){ 678 *status = U_ILLEGAL_ARGUMENT_ERROR; 679 return 0; 680 } 681 682 int32_t reqLength = 0; 683 684 UStringPrepProfile* nameprep = usprep_openByType(USPREP_RFC3491_NAMEPREP, status); 685 686 if(U_FAILURE(*status)){ 687 return 0; 688 } 689 690 //initialize pointers 691 UChar *delimiter = (UChar*)src; 692 UChar *labelStart = (UChar*)src; 693 UChar *currentDest = (UChar*) dest; 694 int32_t remainingLen = srcLength; 695 int32_t remainingDestCapacity = destCapacity; 696 int32_t labelLen = 0, labelReqLength = 0; 697 UBool done = FALSE; 698 699 700 for(;;){ 701 702 labelLen = getNextSeparator(labelStart,remainingLen, &delimiter,&done); 703 labelReqLength = 0; 704 if(!(labelLen==0 && done)){// make sure this is not a root label separator. 705 706 labelReqLength = _internal_toASCII( labelStart, labelLen, 707 currentDest, remainingDestCapacity, 708 options, nameprep, 709 parseError, status); 710 711 if(*status == U_BUFFER_OVERFLOW_ERROR){ 712 713 *status = U_ZERO_ERROR; // reset error 714 remainingDestCapacity = 0; 715 } 716 } 717 718 719 if(U_FAILURE(*status)){ 720 break; 721 } 722 723 reqLength +=labelReqLength; 724 // adjust the destination pointer 725 if(labelReqLength < remainingDestCapacity){ 726 currentDest = currentDest + labelReqLength; 727 remainingDestCapacity -= labelReqLength; 728 }else{ 729 // should never occur 730 remainingDestCapacity = 0; 731 } 732 733 if(done == TRUE){ 734 break; 735 } 736 737 // add the label separator 738 if(remainingDestCapacity > 0){ 739 *currentDest++ = FULL_STOP; 740 remainingDestCapacity--; 741 } 742 reqLength++; 743 744 labelStart = delimiter; 745 if(remainingLen >0 ){ 746 remainingLen = (int32_t)(srcLength - (delimiter - src)); 747 } 748 749 } 750 751 if(reqLength > MAX_DOMAIN_NAME_LENGTH){ 752 *status = U_IDNA_DOMAIN_NAME_TOO_LONG_ERROR; 753 } 754 755 usprep_close(nameprep); 756 757 return u_terminateUChars(dest, destCapacity, reqLength, status); 758 } 759 760 U_CAPI int32_t U_EXPORT2 761 uidna_IDNToUnicode( const UChar* src, int32_t srcLength, 762 UChar* dest, int32_t destCapacity, 763 int32_t options, 764 UParseError* parseError, 765 UErrorCode* status){ 766 767 if(status == NULL || U_FAILURE(*status)){ 768 return 0; 769 } 770 if((src==NULL) || (srcLength < -1) || (destCapacity<0) || (!dest && destCapacity > 0)){ 771 *status = U_ILLEGAL_ARGUMENT_ERROR; 772 return 0; 773 } 774 775 int32_t reqLength = 0; 776 777 UStringPrepProfile* nameprep = usprep_openByType(USPREP_RFC3491_NAMEPREP, status); 778 779 if(U_FAILURE(*status)){ 780 return 0; 781 } 782 783 //initialize pointers 784 UChar *delimiter = (UChar*)src; 785 UChar *labelStart = (UChar*)src; 786 UChar *currentDest = (UChar*) dest; 787 int32_t remainingLen = srcLength; 788 int32_t remainingDestCapacity = destCapacity; 789 int32_t labelLen = 0, labelReqLength = 0; 790 UBool done = FALSE; 791 792 for(;;){ 793 794 labelLen = getNextSeparator(labelStart,remainingLen, &delimiter,&done); 795 796 // The RFC states that 797 // <quote> 798 // ToUnicode never fails. If any step fails, then the original input 799 // is returned immediately in that step. 800 // </quote> 801 // _internal_toUnicode will copy the label. 802 /*if(labelLen==0 && done==FALSE){ 803 *status = U_IDNA_ZERO_LENGTH_LABEL_ERROR; 804 break; 805 }*/ 806 807 labelReqLength = _internal_toUnicode(labelStart, labelLen, 808 currentDest, remainingDestCapacity, 809 options, nameprep, 810 parseError, status); 811 812 if(*status == U_BUFFER_OVERFLOW_ERROR){ 813 *status = U_ZERO_ERROR; // reset error 814 remainingDestCapacity = 0; 815 } 816 817 if(U_FAILURE(*status)){ 818 break; 819 } 820 821 reqLength +=labelReqLength; 822 // adjust the destination pointer 823 if(labelReqLength < remainingDestCapacity){ 824 currentDest = currentDest + labelReqLength; 825 remainingDestCapacity -= labelReqLength; 826 }else{ 827 // should never occur 828 remainingDestCapacity = 0; 829 } 830 831 if(done == TRUE){ 832 break; 833 } 834 835 // add the label separator 836 // Unlike the ToASCII operation we don't normalize the label separators 837 if(remainingDestCapacity > 0){ 838 *currentDest++ = *(labelStart + labelLen); 839 remainingDestCapacity--; 840 } 841 reqLength++; 842 843 labelStart = delimiter; 844 if(remainingLen >0 ){ 845 remainingLen = (int32_t)(srcLength - (delimiter - src)); 846 } 847 848 } 849 850 if(reqLength > MAX_DOMAIN_NAME_LENGTH){ 851 *status = U_IDNA_DOMAIN_NAME_TOO_LONG_ERROR; 852 } 853 854 usprep_close(nameprep); 855 856 return u_terminateUChars(dest, destCapacity, reqLength, status); 857 } 858 859 U_CAPI int32_t U_EXPORT2 860 uidna_compare( const UChar *s1, int32_t length1, 861 const UChar *s2, int32_t length2, 862 int32_t options, 863 UErrorCode* status){ 864 865 if(status == NULL || U_FAILURE(*status)){ 866 return -1; 867 } 868 869 UChar b1Stack[MAX_IDN_BUFFER_SIZE], b2Stack[MAX_IDN_BUFFER_SIZE]; 870 UChar *b1 = b1Stack, *b2 = b2Stack; 871 int32_t b1Len, b2Len, b1Capacity = MAX_IDN_BUFFER_SIZE, b2Capacity = MAX_IDN_BUFFER_SIZE; 872 int32_t result=-1; 873 874 UParseError parseError; 875 876 b1Len = uidna_IDNToASCII(s1, length1, b1, b1Capacity, options, &parseError, status); 877 if(*status == U_BUFFER_OVERFLOW_ERROR){ 878 // redo processing of string 879 b1 = (UChar*) uprv_malloc(b1Len * U_SIZEOF_UCHAR); 880 if(b1==NULL){ 881 *status = U_MEMORY_ALLOCATION_ERROR; 882 goto CLEANUP; 883 } 884 885 *status = U_ZERO_ERROR; // reset error 886 887 b1Len = uidna_IDNToASCII(s1,length1,b1,b1Len, options, &parseError, status); 888 889 } 890 891 b2Len = uidna_IDNToASCII(s2,length2, b2,b2Capacity, options, &parseError, status); 892 if(*status == U_BUFFER_OVERFLOW_ERROR){ 893 // redo processing of string 894 b2 = (UChar*) uprv_malloc(b2Len * U_SIZEOF_UCHAR); 895 if(b2==NULL){ 896 *status = U_MEMORY_ALLOCATION_ERROR; 897 goto CLEANUP; 898 } 899 900 *status = U_ZERO_ERROR; // reset error 901 902 b2Len = uidna_IDNToASCII(s2, length2, b2, b2Len, options, &parseError, status); 903 904 } 905 // when toASCII is applied all label separators are replaced with FULL_STOP 906 result = compareCaseInsensitiveASCII(b1,b1Len,b2,b2Len); 907 908 CLEANUP: 909 if(b1 != b1Stack){ 910 uprv_free(b1); 911 } 912 913 if(b2 != b2Stack){ 914 uprv_free(b2); 915 } 916 917 return result; 918 } 919 920 #endif /* #if !UCONFIG_NO_IDNA */ 921
