Chromium Code Reviews
chromiumcodereview-hr@appspot.gserviceaccount.com (chromiumcodereview-hr) | Please choose your nickname with Settings | Help | Chromium Project | Gerrit Changes | Sign out
(43)

Side by Side Diff: third_party/WebKit/Source/wtf/text/TextCodecICU.cpp

Issue 1709293002: Add support for CSS unicod encoding to the text codec. (Closed) Base URL: https://chromium.googlesource.com/chromium/src.git@master
Patch Set: add a unit test Created 4 years, 7 months ago
Use n/p to move between diff chunks; N/P to move between comments. Draft comments are only viewable by you.
Jump to:
View unified diff | Download patch
OLDNEW
1 /* 1 /*
2 * Copyright (C) 2004, 2006, 2007, 2008, 2011 Apple Inc. All rights reserved. 2 * Copyright (C) 2004, 2006, 2007, 2008, 2011 Apple Inc. All rights reserved.
3 * Copyright (C) 2006 Alexey Proskuryakov <ap@nypop.com> 3 * Copyright (C) 2006 Alexey Proskuryakov <ap@nypop.com>
4 * 4 *
5 * Redistribution and use in source and binary forms, with or without 5 * Redistribution and use in source and binary forms, with or without
6 * modification, are permitted provided that the following conditions 6 * modification, are permitted provided that the following conditions
7 * are met: 7 * are met:
8 * 1. Redistributions of source code must retain the above copyright 8 * 1. Redistributions of source code must retain the above copyright
9 * notice, this list of conditions and the following disclaimer. 9 * notice, this list of conditions and the following disclaimer.
10 * 2. Redistributions in binary form must reproduce the above copyright 10 * 2. Redistributions in binary form must reproduce the above copyright
(...skipping 391 matching lines...) Expand 10 before | Expand all | Expand 10 after
402 switch (character) { 402 switch (character) {
403 case 0x01F9: 403 case 0x01F9:
404 return 0xE7C8; // mapped to xA8xBF by ICU. 404 return 0xE7C8; // mapped to xA8xBF by ICU.
405 case 0x1E3F: 405 case 0x1E3F:
406 return 0xE7C7; // mapped to xA8xBC by ICU. 406 return 0xE7C7; // mapped to xA8xBC by ICU.
407 } 407 }
408 return 0; 408 return 0;
409 } 409 }
410 #endif 410 #endif
411 411
412 // Invalid character handler when writing escaped entities for unrepresentable 412 // Generic helper for writing escaped entities using the specfied UnencodableHan dling.
413 // characters. See the declaration of TextCodec::encode for more. 413 static void formatEscapedEntityCallback(const void* context, UConverterFromUnico deArgs* fromUArgs, const UChar* codeUnits, int32_t length,
414 static void urlEscapedEntityCallback(const void* context, UConverterFromUnicodeA rgs* fromUArgs, const UChar* codeUnits, int32_t length, 414 UChar32 codePoint, UConverterCallbackReason reason, UErrorCode* err, Unencod ableHandling handling)
415 UChar32 codePoint, UConverterCallbackReason reason, UErrorCode* err)
416 { 415 {
417 if (reason == UCNV_UNASSIGNED) { 416 if (reason == UCNV_UNASSIGNED) {
418 *err = U_ZERO_ERROR; 417 *err = U_ZERO_ERROR;
419 418
420 UnencodableReplacementArray entity; 419 UnencodableReplacementArray entity;
421 int entityLen = TextCodec::getUnencodableReplacement(codePoint, URLEncod edEntitiesForUnencodables, entity); 420 int entityLen = TextCodec::getUnencodableReplacement(codePoint, handling , entity);
422 ucnv_cbFromUWriteBytes(fromUArgs, entity, entityLen, 0, err); 421 ucnv_cbFromUWriteBytes(fromUArgs, entity, entityLen, 0, err);
423 } else { 422 } else {
424 UCNV_FROM_U_CALLBACK_ESCAPE(context, fromUArgs, codeUnits, length, codeP oint, reason, err); 423 UCNV_FROM_U_CALLBACK_ESCAPE(context, fromUArgs, codeUnits, length, codeP oint, reason, err);
425 } 424 }
426 } 425 }
427 426
427 // Invalid character handler when writing escaped entities in CSS encoding for
428 // unrepresentable characters. See the declaration of TextCodec::encode for more .
429 static void cssEscapedEntityCallback(const void* context, UConverterFromUnicodeA rgs* fromUArgs, const UChar* codeUnits, int32_t length,
430 UChar32 codePoint, UConverterCallbackReason reason, UErrorCode* err)
431 {
432 formatEscapedEntityCallback(context, fromUArgs, codeUnits, length, codePoint , reason, err, CSSEncodedEntitiesForUnencodables);
433 }
434
435 // Invalid character handler when writing escaped entities in HTML/XML encoding for
436 // unrepresentable characters. See the declaration of TextCodec::encode for more .
437 static void urlEscapedEntityCallback(const void* context, UConverterFromUnicodeA rgs* fromUArgs, const UChar* codeUnits, int32_t length,
438 UChar32 codePoint, UConverterCallbackReason reason, UErrorCode* err)
439 {
440 formatEscapedEntityCallback(context, fromUArgs, codeUnits, length, codePoint , reason, err, URLEncodedEntitiesForUnencodables);
441 }
442
428 #if defined(USING_SYSTEM_ICU) 443 #if defined(USING_SYSTEM_ICU)
429 // Substitutes special GBK characters, escaping all other unassigned entities. 444 // Substitutes special GBK characters, escaping all other unassigned entities.
430 static void gbkCallbackEscape(const void* context, UConverterFromUnicodeArgs* fr omUArgs, const UChar* codeUnits, int32_t length, 445 static void gbkCallbackEscape(const void* context, UConverterFromUnicodeArgs* fr omUArgs, const UChar* codeUnits, int32_t length,
431 UChar32 codePoint, UConverterCallbackReason reason, UErrorCode* err) 446 UChar32 codePoint, UConverterCallbackReason reason, UErrorCode* err)
432 { 447 {
433 UChar outChar; 448 UChar outChar;
434 if (reason == UCNV_UNASSIGNED && (outChar = fallbackForGBK(codePoint))) { 449 if (reason == UCNV_UNASSIGNED && (outChar = fallbackForGBK(codePoint))) {
435 const UChar* source = &outChar; 450 const UChar* source = &outChar;
436 *err = U_ZERO_ERROR; 451 *err = U_ZERO_ERROR;
437 ucnv_cbFromUWriteUChars(fromUArgs, &source, source + 1, 0, err); 452 ucnv_cbFromUWriteUChars(fromUArgs, &source, source + 1, 0, err);
438 return; 453 return;
439 } 454 }
440 UCNV_FROM_U_CALLBACK_ESCAPE(context, fromUArgs, codeUnits, length, codePoint , reason, err); 455 UCNV_FROM_U_CALLBACK_ESCAPE(context, fromUArgs, codeUnits, length, codePoint , reason, err);
441 } 456 }
442 457
458 // Combines both gbkCssEscapedEntityCallback and GBK character substitution.
459 static void gbkCssEscapedEntityCallack(const void* context, UConverterFromUnicod eArgs* fromUArgs, const UChar* codeUnits, int32_t length,
460 UChar32 codePoint, UConverterCallbackReason reason, UErrorCode* err)
461 {
462 if (reason == UCNV_UNASSIGNED) {
463 if (UChar outChar = fallbackForGBK(codePoint)) {
464 const UChar* source = &outChar;
465 *err = U_ZERO_ERROR;
466 ucnv_cbFromUWriteUChars(fromUArgs, &source, source + 1, 0, err);
467 return;
468 }
469 cssEscapedEntityCallback(context, fromUArgs, codeUnits, length, codePoin t, reason, err);
470 return;
471 }
472 UCNV_FROM_U_CALLBACK_ESCAPE(context, fromUArgs, codeUnits, length, codePoint , reason, err);
473 }
474
443 // Combines both gbkUrlEscapedEntityCallback and GBK character substitution. 475 // Combines both gbkUrlEscapedEntityCallback and GBK character substitution.
444 static void gbkUrlEscapedEntityCallack(const void* context, UConverterFromUnicod eArgs* fromUArgs, const UChar* codeUnits, int32_t length, 476 static void gbkUrlEscapedEntityCallack(const void* context, UConverterFromUnicod eArgs* fromUArgs, const UChar* codeUnits, int32_t length,
445 UChar32 codePoint, UConverterCallbackReason reason, UErrorCode* err) 477 UChar32 codePoint, UConverterCallbackReason reason, UErrorCode* err)
446 { 478 {
447 if (reason == UCNV_UNASSIGNED) { 479 if (reason == UCNV_UNASSIGNED) {
448 if (UChar outChar = fallbackForGBK(codePoint)) { 480 if (UChar outChar = fallbackForGBK(codePoint)) {
449 const UChar* source = &outChar; 481 const UChar* source = &outChar;
450 *err = U_ZERO_ERROR; 482 *err = U_ZERO_ERROR;
451 ucnv_cbFromUWriteUChars(fromUArgs, &source, source + 1, 0, err); 483 ucnv_cbFromUWriteUChars(fromUArgs, &source, source + 1, 0, err);
452 return; 484 return;
(...skipping 67 matching lines...) Expand 10 before | Expand all | Expand 10 after
520 ucnv_setFromUCallBack(m_converterICU, m_needsGBKFallbacks ? gbkCallbackE scape : UCNV_FROM_U_CALLBACK_ESCAPE, UCNV_ESCAPE_XML_DEC, 0, 0, &err); 552 ucnv_setFromUCallBack(m_converterICU, m_needsGBKFallbacks ? gbkCallbackE scape : UCNV_FROM_U_CALLBACK_ESCAPE, UCNV_ESCAPE_XML_DEC, 0, 0, &err);
521 #endif 553 #endif
522 break; 554 break;
523 case URLEncodedEntitiesForUnencodables: 555 case URLEncodedEntitiesForUnencodables:
524 #if !defined(USING_SYSTEM_ICU) 556 #if !defined(USING_SYSTEM_ICU)
525 ucnv_setFromUCallBack(m_converterICU, urlEscapedEntityCallback, 0, 0, 0, &err); 557 ucnv_setFromUCallBack(m_converterICU, urlEscapedEntityCallback, 0, 0, 0, &err);
526 #else 558 #else
527 ucnv_setFromUCallBack(m_converterICU, m_needsGBKFallbacks ? gbkUrlEscape dEntityCallack : urlEscapedEntityCallback, 0, 0, 0, &err); 559 ucnv_setFromUCallBack(m_converterICU, m_needsGBKFallbacks ? gbkUrlEscape dEntityCallack : urlEscapedEntityCallback, 0, 0, 0, &err);
528 #endif 560 #endif
529 break; 561 break;
562 case CSSEncodedEntitiesForUnencodables:
563 #if !defined(USING_SYSTEM_ICU)
564 ucnv_setFromUCallBack(m_converterICU, cssEscapedEntityCallback, 0, 0, 0, &err);
565 #else
566 ucnv_setFromUCallBack(m_converterICU, m_needsGBKFallbacks ? gbkCssEscape dEntityCallack : cssEscapedEntityCallback, 0, 0, 0, &err);
567 #endif
568 break;
530 } 569 }
531 570
532 ASSERT(U_SUCCESS(err)); 571 ASSERT(U_SUCCESS(err));
533 if (U_FAILURE(err)) 572 if (U_FAILURE(err))
534 return CString(); 573 return CString();
535 574
536 Vector<char> result; 575 Vector<char> result;
537 size_t size = 0; 576 size_t size = 0;
538 do { 577 do {
539 char buffer[ConversionBufferSize]; 578 char buffer[ConversionBufferSize];
(...skipping 29 matching lines...) Expand all
569 { 608 {
570 return encodeCommon(characters, length, handling); 609 return encodeCommon(characters, length, handling);
571 } 610 }
572 611
573 CString TextCodecICU::encode(const LChar* characters, size_t length, Unencodable Handling handling) 612 CString TextCodecICU::encode(const LChar* characters, size_t length, Unencodable Handling handling)
574 { 613 {
575 return encodeCommon(characters, length, handling); 614 return encodeCommon(characters, length, handling);
576 } 615 }
577 616
578 } // namespace WTF 617 } // namespace WTF
OLDNEW
« no previous file with comments | « third_party/WebKit/Source/wtf/text/TextCodec.cpp ('k') | third_party/WebKit/Source/wtf/text/TextCodecTest.cpp » ('j') | no next file with comments »

Powered by Google App Engine
This is Rietveld 408576698