Differences From Artifact [2e4dae0bbe]:
- File
src/OFString_UTF8.m
— part of check-in
[1cb8fee5c3]
at
2012-10-13 21:06:59
on branch trunk
— of_bswap* -> OF_BSWAP*.
This makes it clear that multiple evaluation of parameters is possible. (user: js, size: 29614) [annotate] [blame] [check-ins using]
To Artifact [dd6ca8a27b]:
- File src/OFString_UTF8.m — part of check-in [9817eab3be] at 2012-10-13 21:29:30 on branch trunk — Rename s->UTF8 to s->isUTF8. (user: js, size: 29674) [annotate] [blame] [check-ins using]
︙ | ︙ | |||
252 253 254 255 256 257 258 | s->cString = storage; s->cStringLength = UTF8StringLength; switch (of_string_utf8_check(UTF8String, UTF8StringLength, &s->length)) { case 1: | | | 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 | s->cString = storage; s->cStringLength = UTF8StringLength; switch (of_string_utf8_check(UTF8String, UTF8StringLength, &s->length)) { case 1: s->isUTF8 = YES; break; case -1: @throw [OFInvalidEncodingException exceptionWithClass: [self class]]; } memcpy(s->cString, UTF8String, UTF8StringLength); |
︙ | ︙ | |||
299 300 301 302 303 304 305 | switch (of_string_utf8_check(cString, cStringLength, &s->length)) { case 1: if (encoding == OF_STRING_ENCODING_ASCII) @throw [OFInvalidEncodingException exceptionWithClass: [self class]]; | | | 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 | switch (of_string_utf8_check(cString, cStringLength, &s->length)) { case 1: if (encoding == OF_STRING_ENCODING_ASCII) @throw [OFInvalidEncodingException exceptionWithClass: [self class]]; s->isUTF8 = YES; break; case -1: @throw [OFInvalidEncodingException exceptionWithClass: [self class]]; } memcpy(s->cString, cString, cStringLength); |
︙ | ︙ | |||
325 326 327 328 329 330 331 | size_t bytes; if (!(cString[i] & 0x80)) { s->cString[j++] = cString[i]; continue; } | | | 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 | size_t bytes; if (!(cString[i] & 0x80)) { s->cString[j++] = cString[i]; continue; } s->isUTF8 = YES; bytes = of_string_utf8_encode( (uint8_t)cString[i], buffer); if (bytes == 0) @throw [OFInvalidEncodingException exceptionWithClass: [self class]]; |
︙ | ︙ | |||
375 376 377 378 379 380 381 | character = table[(uint8_t)cString[i]]; if (character == 0xFFFD) @throw [OFInvalidEncodingException exceptionWithClass: [self class]]; | | | 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 | character = table[(uint8_t)cString[i]]; if (character == 0xFFFD) @throw [OFInvalidEncodingException exceptionWithClass: [self class]]; s->isUTF8 = YES; characterBytes = of_string_utf8_encode(character, buffer); if (characterBytes == 0) @throw [OFInvalidEncodingException exceptionWithClass: [self class]]; |
︙ | ︙ | |||
425 426 427 428 429 430 431 | s->cString = (char*)UTF8String; s->cStringLength = UTF8StringLength; switch (of_string_utf8_check(UTF8String, UTF8StringLength, &s->length)) { case 1: | | | 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 | s->cString = (char*)UTF8String; s->cStringLength = UTF8StringLength; switch (of_string_utf8_check(UTF8String, UTF8StringLength, &s->length)) { case 1: s->isUTF8 = YES; break; case -1: @throw [OFInvalidEncodingException exceptionWithClass: [self class]]; } } @catch (id e) { [self release]; |
︙ | ︙ | |||
450 451 452 453 454 455 456 | @try { s = &s_store; s->cStringLength = [string UTF8StringLength]; if ([string isKindOfClass: [OFString_UTF8 class]] || [string isKindOfClass: [OFMutableString_UTF8 class]]) | | | | 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 | @try { s = &s_store; s->cStringLength = [string UTF8StringLength]; if ([string isKindOfClass: [OFString_UTF8 class]] || [string isKindOfClass: [OFMutableString_UTF8 class]]) s->isUTF8 = ((OFString_UTF8*)string)->s->isUTF8; else s->isUTF8 = YES; s->length = [string length]; s->cString = [self allocMemoryWithSize: s->cStringLength + 1]; memcpy(s->cString, [string UTF8String], s->cStringLength + 1); } @catch (id e) { [self release]; |
︙ | ︙ | |||
503 504 505 506 507 508 509 | buffer); switch (characterLen) { case 1: s->cString[j++] = buffer[0]; break; case 2: | | | | | 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 | buffer); switch (characterLen) { case 1: s->cString[j++] = buffer[0]; break; case 2: s->isUTF8 = YES; s->cStringLength++; memcpy(s->cString + j, buffer, 2); j += 2; break; case 3: s->isUTF8 = YES; s->cStringLength += 2; memcpy(s->cString + j, buffer, 3); j += 3; break; case 4: s->isUTF8 = YES; s->cStringLength += 3; memcpy(s->cString + j, buffer, 4); j += 4; break; default: |
︙ | ︙ | |||
610 611 612 613 614 615 616 | characterLen = of_string_utf8_encode(character, buffer); switch (characterLen) { case 1: s->cString[j++] = buffer[0]; break; case 2: | | | | | 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 | characterLen = of_string_utf8_encode(character, buffer); switch (characterLen) { case 1: s->cString[j++] = buffer[0]; break; case 2: s->isUTF8 = YES; s->cStringLength++; memcpy(s->cString + j, buffer, 2); j += 2; break; case 3: s->isUTF8 = YES; s->cStringLength += 2; memcpy(s->cString + j, buffer, 3); j += 3; break; case 4: s->isUTF8 = YES; s->cStringLength += 3; memcpy(s->cString + j, buffer, 4); j += 4; break; default: |
︙ | ︙ | |||
682 683 684 685 686 687 688 | s->cStringLength = cStringLength; @try { switch (of_string_utf8_check(tmp, cStringLength, &s->length)) { case 1: | | | 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 | s->cStringLength = cStringLength; @try { switch (of_string_utf8_check(tmp, cStringLength, &s->length)) { case 1: s->isUTF8 = YES; break; case -1: @throw [OFInvalidEncodingException exceptionWithClass: [self class]]; } s->cString = [self |
︙ | ︙ | |||
720 721 722 723 724 725 726 | s = &s_store; s->cStringLength = [firstComponent UTF8StringLength]; if ([firstComponent isKindOfClass: [OFString_UTF8 class]] || [firstComponent isKindOfClass: [OFMutableString_UTF8 class]]) | | | > | | | 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 | s = &s_store; s->cStringLength = [firstComponent UTF8StringLength]; if ([firstComponent isKindOfClass: [OFString_UTF8 class]] || [firstComponent isKindOfClass: [OFMutableString_UTF8 class]]) s->isUTF8 = ((OFString_UTF8*)firstComponent)->s->isUTF8; else s->isUTF8 = YES; s->length = [firstComponent length]; /* Calculate length and see if we need UTF-8 */ va_copy(argumentsCopy, arguments); while ((component = va_arg(argumentsCopy, OFString*)) != nil) { s->cStringLength += 1 + [component UTF8StringLength]; s->length += 1 + [component length]; if ([component isKindOfClass: [OFString_UTF8 class]] || [component isKindOfClass: [OFMutableString_UTF8 class]]) s->isUTF8 = ((OFString_UTF8*)component)->s->isUTF8; else s->isUTF8 = YES; } s->cString = [self allocMemoryWithSize: s->cStringLength + 1]; cStringLength = [firstComponent UTF8StringLength]; memcpy(s->cString, [firstComponent UTF8String], cStringLength); i = cStringLength; |
︙ | ︙ | |||
784 785 786 787 788 789 790 | - (const char*)cStringWithEncoding: (of_string_encoding_t)encoding { switch (encoding) { case OF_STRING_ENCODING_UTF_8: return s->cString; case OF_STRING_ENCODING_ASCII: | | | 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 | - (const char*)cStringWithEncoding: (of_string_encoding_t)encoding { switch (encoding) { case OF_STRING_ENCODING_UTF_8: return s->cString; case OF_STRING_ENCODING_ASCII: if (s->isUTF8) @throw [OFInvalidEncodingException exceptionWithClass: [self class]]; return s->cString; default: @throw [OFNotImplementedException exceptionWithClass: [self class] |
︙ | ︙ | |||
812 813 814 815 816 817 818 | - (size_t)cStringLengthWithEncoding: (of_string_encoding_t)encoding { switch (encoding) { case OF_STRING_ENCODING_UTF_8: return s->cStringLength; case OF_STRING_ENCODING_ASCII: | | | 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 | - (size_t)cStringLengthWithEncoding: (of_string_encoding_t)encoding { switch (encoding) { case OF_STRING_ENCODING_UTF_8: return s->cStringLength; case OF_STRING_ENCODING_ASCII: if (s->isUTF8) @throw [OFInvalidEncodingException exceptionWithClass: [self class]]; return s->cStringLength; default: @throw [OFNotImplementedException exceptionWithClass: [self class] |
︙ | ︙ | |||
897 898 899 900 901 902 903 | @throw [OFInvalidArgumentException exceptionWithClass: [self class] selector: _cmd]; otherCString = [otherString UTF8String]; otherCStringLength = [otherString UTF8StringLength]; | | | 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 | @throw [OFInvalidArgumentException exceptionWithClass: [self class] selector: _cmd]; otherCString = [otherString UTF8String]; otherCStringLength = [otherString UTF8StringLength]; if (!s->isUTF8) { minimumCStringLength = (s->cStringLength > otherCStringLength ? otherCStringLength : s->cStringLength); if ((compare = memcasecmp(s->cString, otherCString, minimumCStringLength)) == 0) { if (s->cStringLength > otherCStringLength) return OF_ORDERED_DESCENDING; |
︙ | ︙ | |||
1005 1006 1007 1008 1009 1010 1011 | - (of_unichar_t)characterAtIndex: (size_t)index { of_unichar_t character; if (index >= s->length) @throw [OFOutOfRangeException exceptionWithClass: [self class]]; | | | 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 | - (of_unichar_t)characterAtIndex: (size_t)index { of_unichar_t character; if (index >= s->length) @throw [OFOutOfRangeException exceptionWithClass: [self class]]; if (!s->isUTF8) return s->cString[index]; index = of_string_utf8_get_position(s->cString, index, s->cStringLength); if (!of_string_utf8_decode(s->cString + index, s->cStringLength - index, &character)) |
︙ | ︙ | |||
1092 1093 1094 1095 1096 1097 1098 | { size_t start = range.start; size_t end = range.start + range.length; if (end > s->length) @throw [OFOutOfRangeException exceptionWithClass: [self class]]; | | | 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 | { size_t start = range.start; size_t end = range.start + range.length; if (end > s->length) @throw [OFOutOfRangeException exceptionWithClass: [self class]]; if (s->isUTF8) { start = of_string_utf8_get_position(s->cString, start, s->cStringLength); end = of_string_utf8_get_position(s->cString, end, s->cStringLength); } return [OFString stringWithUTF8String: s->cString + start |
︙ | ︙ |