path: root/chromium/third_party/blink/renderer/platform/wtf/text/text_codec_icu.cc

/*
 * Copyright (C) 2004, 2006, 2007, 2008, 2011 Apple Inc. All rights reserved.
 * Copyright (C) 2006 Alexey Proskuryakov <ap@nypop.com>
 *
 * Redistribution and use in source and binary forms, with or without
 * modification, are permitted provided that the following conditions
 * are met:
 * 1. Redistributions of source code must retain the above copyright
 *    notice, this list of conditions and the following disclaimer.
 * 2. Redistributions in binary form must reproduce the above copyright
 *    notice, this list of conditions and the following disclaimer in the
 *    documentation and/or other materials provided with the distribution.
 *
 * THIS SOFTWARE IS PROVIDED BY APPLE COMPUTER, INC. ``AS IS'' AND ANY
 * EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
 * PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL APPLE COMPUTER, INC. OR
 * CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
 * EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
 * PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
 * PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY
 * OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
 * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
 * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
 */

#include "third_party/blink/renderer/platform/wtf/text/text_codec_icu.h"

#include <memory>

#include <unicode/ucnv.h>
#include <unicode/ucnv_cb.h>

#include "base/memory/ptr_util.h"
#include "base/notreached.h"
#include "third_party/blink/renderer/platform/wtf/assertions.h"
#include "third_party/blink/renderer/platform/wtf/text/character_names.h"
#include "third_party/blink/renderer/platform/wtf/text/string_builder.h"
#include "third_party/blink/renderer/platform/wtf/text/string_view.h"
#include "third_party/blink/renderer/platform/wtf/threading.h"

namespace WTF {

const size_t kConversionBufferSize = 16384;

ICUConverterWrapper::~ICUConverterWrapper() {
  if (converter)
    ucnv_close(converter);
}

static UConverter*& CachedConverterICU() {
  return WtfThreading().CachedConverterICU().converter;
}

std::unique_ptr<TextCodec> TextCodecICU::Create(const TextEncoding& encoding,
                                                const void*) {
  return base::WrapUnique(new TextCodecICU(encoding));
}

namespace {
bool IncludeAlias(const char* alias) {
#if !defined(USING_SYSTEM_ICU)
  // Chromium's build of ICU includes *-html aliases to manage the encoding
  // labels defined in the Encoding Standard, but these must not be
  // web-exposed.
  const char* kSuffix = "-html";
  const size_t kSuffixLength = 5;
  size_t alias_length = strlen(alias);
  if ((alias_length >= kSuffixLength) &&
      !strcmp(alias + alias_length - kSuffixLength, kSuffix))
    return false;
#endif
  return true;
}
}  // namespace

void TextCodecICU::RegisterEncodingNames(EncodingNameRegistrar registrar) {
  // We register Hebrew with logical ordering using a separate name.
  // Otherwise, this would share the same canonical name as the
  // visual ordering case, and then TextEncoding could not tell them
  // apart; ICU treats these names as synonyms.
  registrar("ISO-8859-8-I", "ISO-8859-8-I");

  int32_t num_encodings = ucnv_countAvailable();
  for (int32_t i = 0; i < num_encodings; ++i) {
    const char* name = ucnv_getAvailableName(i);
    UErrorCode error = U_ZERO_ERROR;
#if !defined(USING_SYSTEM_ICU)
    const char* primary_standard = "HTML";
    const char* secondary_standard = "MIME";
#else
    const char* primary_standard = "MIME";
    const char* secondary_standard = "IANA";
#endif
    const char* standard_name =
        ucnv_getStandardName(name, primary_standard, &error);
    if (U_FAILURE(error) || !standard_name) {
      error = U_ZERO_ERROR;
      // Try IANA to pick up 'windows-12xx' and other names
      // which are not preferred MIME names but are widely used.
      standard_name = ucnv_getStandardName(name, secondary_standard, &error);
      if (U_FAILURE(error) || !standard_name)
        continue;
    }

#if defined(USING_SYSTEM_ICU)
    // Explicitly do not support UTF-32. https://crbug.com/417850
    // Bundled ICU does not return these names.
    if (!strcmp(standard_name, "UTF-32") ||
        !strcmp(standard_name, "UTF-32LE") ||
        !strcmp(standard_name, "UTF-32BE")) {
      continue;
    }
#endif

// A number of these aliases are handled in Chrome's copy of ICU, but
// Chromium can be compiled with the system ICU.

// 1. Treat GB2312 encoding as GBK (its more modern superset), to match other
//    browsers.
// 2. On the Web, GB2312 is encoded as EUC-CN or HZ, while ICU provides a native
//    encoding for encoding GB_2312-80 and several others. So, we need to
//    override this behavior, too.
#if defined(USING_SYSTEM_ICU)
    if (!strcmp(standard_name, "GB2312") ||
        !strcmp(standard_name, "GB_2312-80")) {
      standard_name = "GBK";
    // Similarly, EUC-KR encodings all map to an extended version, but
    // per HTML5, the canonical name still should be EUC-KR.
    } else if (!strcmp(standard_name, "EUC-KR") ||
               !strcmp(standard_name, "KSC_5601") ||
               !strcmp(standard_name, "cp1363")) {
      standard_name = "EUC-KR";
    // And so on.
    } else if (EqualIgnoringASCIICase(standard_name, "iso-8859-9")) {
      // This name is returned in different case by ICU 3.2 and 3.6.
      standard_name = "windows-1254";
    } else if (!strcmp(standard_name, "TIS-620")) {
      standard_name = "windows-874";
    }
#endif

    registrar(standard_name, standard_name);

    uint16_t num_aliases = ucnv_countAliases(name, &error);
    DCHECK(U_SUCCESS(error));
    if (U_SUCCESS(error))
      for (uint16_t j = 0; j < num_aliases; ++j) {
        error = U_ZERO_ERROR;
        const char* alias = ucnv_getAlias(name, j, &error);
        DCHECK(U_SUCCESS(error));
        if (U_SUCCESS(error) && alias != standard_name && IncludeAlias(alias))
          registrar(alias, standard_name);
      }
  }

  // These two entries have to be added here because ICU's converter table
  // cannot have both ISO-8859-8-I and ISO-8859-8.
  registrar("csISO88598I", "ISO-8859-8-I");
  registrar("logical", "ISO-8859-8-I");

#if defined(USING_SYSTEM_ICU)
  // Additional alias for MacCyrillic not present in ICU.
  registrar("maccyrillic", "x-mac-cyrillic");

  // Additional aliases that historically were present in the encoding
  // table in WebKit on Macintosh that don't seem to be present in ICU.
  // Perhaps we can prove these are not used on the web and remove them.
  // Or perhaps we can get them added to ICU.
  registrar("x-mac-roman", "macintosh");
  registrar("x-mac-ukrainian", "x-mac-cyrillic");
  registrar("cn-big5", "Big5");
  registrar("x-x-big5", "Big5");
  registrar("cn-gb", "GBK");
  registrar("csgb231280", "GBK");
  registrar("x-euc-cn", "GBK");
  registrar("x-gbk", "GBK");
  registrar("koi", "KOI8-R");
  registrar("visual", "ISO-8859-8");
  registrar("winarabic", "windows-1256");
  registrar("winbaltic", "windows-1257");
  registrar("wincyrillic", "windows-1251");
  registrar("iso-8859-11", "windows-874");
  registrar("iso8859-11", "windows-874");
  registrar("dos-874", "windows-874");
  registrar("wingreek", "windows-1253");
  registrar("winhebrew", "windows-1255");
  registrar("winlatin2", "windows-1250");
  registrar("winturkish", "windows-1254");
  registrar("winvietnamese", "windows-1258");
  registrar("x-cp1250", "windows-1250");
  registrar("x-cp1251", "windows-1251");
  registrar("x-euc", "EUC-JP");
  registrar("x-windows-949", "EUC-KR");
  registrar("KSC5601", "EUC-KR");
  registrar("x-uhc", "EUC-KR");
  registrar("shift-jis", "Shift_JIS");

  // Alternative spelling of ISO encoding names.
  registrar("ISO8859-1", "ISO-8859-1");
  registrar("ISO8859-2", "ISO-8859-2");
  registrar("ISO8859-3", "ISO-8859-3");
  registrar("ISO8859-4", "ISO-8859-4");
  registrar("ISO8859-5", "ISO-8859-5");
  registrar("ISO8859-6", "ISO-8859-6");
  registrar("ISO8859-7", "ISO-8859-7");
  registrar("ISO8859-8", "ISO-8859-8");
  registrar("ISO8859-8-I", "ISO-8859-8-I");
  registrar("ISO8859-9", "ISO-8859-9");
  registrar("ISO8859-10", "ISO-8859-10");
  registrar("ISO8859-13", "ISO-8859-13");
  registrar("ISO8859-14", "ISO-8859-14");
  registrar("ISO8859-15", "ISO-8859-15");
  // No need to have an entry for ISO8859-16. ISO-8859-16 has just one label
  // listed in WHATWG Encoding Living Standard, http://encoding.spec.whatwg.org/

  // Additional aliases present in the WHATWG Encoding Standard
  // and Firefox (as of Oct 2014), but not in the upstream ICU.
  // Three entries for windows-1252 need not be listed here because
  // TextCodecLatin1 registers them.
  registrar("csiso58gb231280", "GBK");
  registrar("csiso88596e", "ISO-8859-6");
  registrar("csiso88596i", "ISO-8859-6");
  registrar("csiso88598e", "ISO-8859-8");
  registrar("gb_2312", "GBK");
  registrar("iso88592", "ISO-8859-2");
  registrar("iso88593", "ISO-8859-3");
  registrar("iso88594", "ISO-8859-4");
  registrar("iso88595", "ISO-8859-5");
  registrar("iso88596", "ISO-8859-6");
  registrar("iso88597", "ISO-8859-7");
  registrar("iso88598", "ISO-8859-8");
  registrar("iso88599", "windows-1254");
  registrar("iso885910", "ISO-8859-10");
  registrar("iso885911", "windows-874");
  registrar("iso885913", "ISO-8859-13");
  registrar("iso885914", "ISO-8859-14");
  registrar("iso885915", "ISO-8859-15");
  registrar("iso_8859-2", "ISO-8859-2");
  registrar("iso_8859-3", "ISO-8859-3");
  registrar("iso_8859-4", "ISO-8859-4");
  registrar("iso_8859-5", "ISO-8859-5");
  registrar("iso_8859-6", "ISO-8859-6");
  registrar("iso_8859-7", "ISO-8859-7");
  registrar("iso_8859-8", "ISO-8859-8");
  registrar("iso_8859-9", "windows-1254");
  registrar("iso_8859-15", "ISO-8859-15");
  registrar("koi8_r", "KOI8-R");
  registrar("x-cp1253", "windows-1253");
  registrar("x-cp1254", "windows-1254");
  registrar("x-cp1255", "windows-1255");
  registrar("x-cp1256", "windows-1256");
  registrar("x-cp1257", "windows-1257");
  registrar("x-cp1258", "windows-1258");
#endif
}

void TextCodecICU::RegisterCodecs(TextCodecRegistrar registrar) {
  // See comment above in registerEncodingNames.
  registrar("ISO-8859-8-I", Create, nullptr);

  int32_t num_encodings = ucnv_countAvailable();
  for (int32_t i = 0; i < num_encodings; ++i) {
    const char* name = ucnv_getAvailableName(i);
    UErrorCode error = U_ZERO_ERROR;
    const char* standard_name = ucnv_getStandardName(name, "MIME", &error);
    if (!U_SUCCESS(error) || !standard_name) {
      error = U_ZERO_ERROR;
      standard_name = ucnv_getStandardName(name, "IANA", &error);
      if (!U_SUCCESS(error) || !standard_name)
        continue;
    }
#if defined(USING_SYSTEM_ICU)
    // Explicitly do not support UTF-32. https://crbug.com/417850
    // Bundled ICU does not return these names.
    if (!strcmp(standard_name, "UTF-32") ||
        !strcmp(standard_name, "UTF-32LE") ||
        !strcmp(standard_name, "UTF-32BE")) {
      continue;
    }
#endif
    registrar(standard_name, Create, nullptr);
  }
}

TextCodecICU::TextCodecICU(const TextEncoding& encoding)
    : encoding_(encoding) {}

TextCodecICU::~TextCodecICU() {
  ReleaseICUConverter();
}

void TextCodecICU::ReleaseICUConverter() const {
  if (converter_icu_) {
    UConverter*& cached_converter = CachedConverterICU();
    if (cached_converter)
      ucnv_close(cached_converter);
    cached_converter = converter_icu_;
    converter_icu_ = nullptr;
  }
}

void TextCodecICU::CreateICUConverter() const {
  DCHECK(!converter_icu_);

#if defined(USING_SYSTEM_ICU)
  const char* name = encoding_.GetName();
  needs_gbk_fallbacks_ =
      name[0] == 'G' && name[1] == 'B' && name[2] == 'K' && !name[3];
#endif

  UErrorCode err;

  UConverter*& cached_converter = CachedConverterICU();
  if (cached_converter) {
    err = U_ZERO_ERROR;
    const char* cached_name = ucnv_getName(cached_converter, &err);
    if (U_SUCCESS(err) && encoding_ == TextEncoding(cached_name)) {
      converter_icu_ = cached_converter;
      cached_converter = nullptr;
      return;
    }
  }

  err = U_ZERO_ERROR;
  converter_icu_ = ucnv_open(encoding_.GetName(), &err);
  DLOG_IF(ERROR, err == U_AMBIGUOUS_ALIAS_WARNING)
      << "ICU ambiguous alias warning for encoding: " << encoding_.GetName();
  if (converter_icu_)
    ucnv_setFallback(converter_icu_, TRUE);
}

int TextCodecICU::DecodeToBuffer(UChar* target,
                                 UChar* target_limit,
                                 const char*& source,
                                 const char* source_limit,
                                 int32_t* offsets,
                                 bool flush,
                                 UErrorCode& err) {
  UChar* target_start = target;
  err = U_ZERO_ERROR;
  ucnv_toUnicode(converter_icu_, &target, target_limit, &source, source_limit,
                 offsets, flush, &err);
  return static_cast<int>(target - target_start);
}

class ErrorCallbackSetter final {
  STACK_ALLOCATED();

 public:
  ErrorCallbackSetter(UConverter* converter, bool stop_on_error)
      : converter_(converter), should_stop_on_encoding_errors_(stop_on_error) {
    if (should_stop_on_encoding_errors_) {
      UErrorCode err = U_ZERO_ERROR;
      ucnv_setToUCallBack(converter_, UCNV_TO_U_CALLBACK_STOP, nullptr,
                          &saved_action_, &saved_context_, &err);
      DCHECK_EQ(err, U_ZERO_ERROR);
    }
  }
  ~ErrorCallbackSetter() {
    if (should_stop_on_encoding_errors_) {
      UErrorCode err = U_ZERO_ERROR;
      const void* old_context;
      UConverterToUCallback old_action;
      ucnv_setToUCallBack(converter_, saved_action_, saved_context_,
                          &old_action, &old_context, &err);
      DCHECK_EQ(old_action, UCNV_TO_U_CALLBACK_STOP);
      DCHECK(!old_context);
      DCHECK_EQ(err, U_ZERO_ERROR);
    }
  }

 private:
  UConverter* converter_;
  bool should_stop_on_encoding_errors_;
  const void* saved_context_;
  UConverterToUCallback saved_action_;
};

String TextCodecICU::Decode(const char* bytes,
                            wtf_size_t length,
                            FlushBehavior flush,
                            bool stop_on_error,
                            bool& saw_error) {
  // Get a converter for the passed-in encoding.
  if (!converter_icu_) {
    CreateICUConverter();
    DCHECK(converter_icu_);
    if (!converter_icu_) {
      DLOG(ERROR)
          << "error creating ICU encoder even though encoding was in table";
      return String();
    }
  }

  ErrorCallbackSetter callback_setter(converter_icu_, stop_on_error);

  StringBuilder result;

  UChar buffer[kConversionBufferSize];
  UChar* buffer_limit = buffer + kConversionBufferSize;
  const char* source = reinterpret_cast<const char*>(bytes);
  const char* source_limit = source + length;
  int32_t* offsets = nullptr;
  UErrorCode err = U_ZERO_ERROR;

  do {
    int uchars_decoded =
        DecodeToBuffer(buffer, buffer_limit, source, source_limit, offsets,
                       flush != FlushBehavior::kDoNotFlush, err);
    result.Append(buffer, uchars_decoded);
  } while (err == U_BUFFER_OVERFLOW_ERROR);

  if (U_FAILURE(err)) {
    // flush the converter so it can be reused, and not be bothered by this
    // error.
    do {
      DecodeToBuffer(buffer, buffer_limit, source, source_limit, offsets, true,
                     err);
    } while (source < source_limit);
    saw_error = true;
  }

#if !defined(USING_SYSTEM_ICU)
  // Chrome's copy of ICU does not have the issue described below.
  return result.ToString();
#else
  String resultString = result.ToString();

  // <http://bugs.webkit.org/show_bug.cgi?id=17014>
  // Simplified Chinese pages use the code A3A0 to mean "full-width space", but
  // ICU decodes it as U+E5E5.
  if (!strcmp(encoding_.GetName(), "GBK")) {
    if (EqualIgnoringASCIICase(encoding_.GetName(), "gb18030"))
      resultString.Replace(0xE5E5, ideographicSpaceCharacter);
    // Make GBK compliant to the encoding spec and align with GB18030
    resultString.Replace(0x01F9, 0xE7C8);
    // FIXME: Once https://www.w3.org/Bugs/Public/show_bug.cgi?id=28740#c3
    // is resolved, add U+1E3F => 0xE7C7.
  }

  return resultString;
#endif
}

#if defined(USING_SYSTEM_ICU)
// U+01F9 and U+1E3F have to be mapped to xA8xBF and xA8xBC per the encoding
// spec, but ICU converter does not have them.
static UChar FallbackForGBK(UChar32 character) {
  switch (character) {
    case 0x01F9:
      return 0xE7C8;  // mapped to xA8xBF by ICU.
    case 0x1E3F:
      return 0xE7C7;  // mapped to xA8xBC by ICU.
  }
  return 0;
}
#endif

// Generic helper for writing escaped entities using the specified
// UnencodableHandling.
static void FormatEscapedEntityCallback(const void* context,
                                        UConverterFromUnicodeArgs* from_u_args,
                                        const UChar* code_units,
                                        int32_t length,
                                        UChar32 code_point,
                                        UConverterCallbackReason reason,
                                        UErrorCode* err,
                                        UnencodableHandling handling) {
  if (reason == UCNV_UNASSIGNED) {
    *err = U_ZERO_ERROR;

    UnencodableReplacementArray entity;
    uint32_t entity_len =
        TextCodec::GetUnencodableReplacement(code_point, handling, entity);
    String entity_u(entity, entity_len);
    entity_u.Ensure16Bit();
    const UChar* entity_u_pointers[2] = {
        entity_u.Characters16(), entity_u.Characters16() + entity_u.length(),
    };
    ucnv_cbFromUWriteUChars(from_u_args, entity_u_pointers,
                            entity_u_pointers[1], 0, err);
  } else {
    UCNV_FROM_U_CALLBACK_ESCAPE(context, from_u_args, code_units, length,
                                code_point, reason, err);
  }
}

static void NumericEntityCallback(const void* context,
                                  UConverterFromUnicodeArgs* from_u_args,
                                  const UChar* code_units,
                                  int32_t length,
                                  UChar32 code_point,
                                  UConverterCallbackReason reason,
                                  UErrorCode* err) {
  FormatEscapedEntityCallback(context, from_u_args, code_units, length,
                              code_point, reason, err,
                              kEntitiesForUnencodables);
}

// Invalid character handler when writing escaped entities in CSS encoding for
// unrepresentable characters. See the declaration of TextCodec::encode for
// more.
static void CssEscapedEntityCallback(const void* context,
                                     UConverterFromUnicodeArgs* from_u_args,
                                     const UChar* code_units,
                                     int32_t length,
                                     UChar32 code_point,
                                     UConverterCallbackReason reason,
                                     UErrorCode* err) {
  FormatEscapedEntityCallback(context, from_u_args, code_units, length,
                              code_point, reason, err,
                              kCSSEncodedEntitiesForUnencodables);
}

// Invalid character handler when writing escaped entities in HTML/XML encoding
// for unrepresentable characters. See the declaration of TextCodec::encode for
// more.
static void UrlEscapedEntityCallback(const void* context,
                                     UConverterFromUnicodeArgs* from_u_args,
                                     const UChar* code_units,
                                     int32_t length,
                                     UChar32 code_point,
                                     UConverterCallbackReason reason,
                                     UErrorCode* err) {
  FormatEscapedEntityCallback(context, from_u_args, code_units, length,
                              code_point, reason, err,
                              kURLEncodedEntitiesForUnencodables);
}

#if defined(USING_SYSTEM_ICU)
// Substitutes special GBK characters, escaping all other unassigned entities.
static void GbkCallbackEscape(const void* context,
                              UConverterFromUnicodeArgs* from_unicode_args,
                              const UChar* code_units,
                              int32_t length,
                              UChar32 code_point,
                              UConverterCallbackReason reason,
                              UErrorCode* err) {
  UChar out_char;
  if (reason == UCNV_UNASSIGNED && (out_char = FallbackForGBK(code_point))) {
    const UChar* source = &out_char;
    *err = U_ZERO_ERROR;
    ucnv_cbFromUWriteUChars(from_unicode_args, &source, source + 1, 0, err);
    return;
  }
  NumericEntityCallback(context, from_unicode_args, code_units, length,
                        code_point, reason, err);
}

// Combines both gbkCssEscapedEntityCallback and GBK character substitution.
static void GbkCssEscapedEntityCallack(
    const void* context,
    UConverterFromUnicodeArgs* from_unicode_args,
    const UChar* code_units,
    int32_t length,
    UChar32 code_point,
    UConverterCallbackReason reason,
    UErrorCode* err) {
  if (reason == UCNV_UNASSIGNED) {
    if (UChar out_char = FallbackForGBK(code_point)) {
      const UChar* source = &out_char;
      *err = U_ZERO_ERROR;
      ucnv_cbFromUWriteUChars(from_unicode_args, &source, source + 1, 0, err);
      return;
    }
    CssEscapedEntityCallback(context, from_unicode_args, code_units, length,
                             code_point, reason, err);
    return;
  }
  UCNV_FROM_U_CALLBACK_ESCAPE(context, from_unicode_args, code_units, length,
                              code_point, reason, err);
}

// Combines both gbkUrlEscapedEntityCallback and GBK character substitution.
static void GbkUrlEscapedEntityCallack(
    const void* context,
    UConverterFromUnicodeArgs* from_unicode_args,
    const UChar* code_units,
    int32_t length,
    UChar32 code_point,
    UConverterCallbackReason reason,
    UErrorCode* err) {
  if (reason == UCNV_UNASSIGNED) {
    if (UChar out_char = FallbackForGBK(code_point)) {
      const UChar* source = &out_char;
      *err = U_ZERO_ERROR;
      ucnv_cbFromUWriteUChars(from_unicode_args, &source, source + 1, 0, err);
      return;
    }
    UrlEscapedEntityCallback(context, from_unicode_args, code_units, length,
                             code_point, reason, err);
    return;
  }
  UCNV_FROM_U_CALLBACK_ESCAPE(context, from_unicode_args, code_units, length,
                              code_point, reason, err);
}

static void GbkCallbackSubstitute(const void* context,
                                  UConverterFromUnicodeArgs* from_unicode_args,
                                  const UChar* code_units,
                                  int32_t length,
                                  UChar32 code_point,
                                  UConverterCallbackReason reason,
                                  UErrorCode* err) {
  UChar out_char;
  if (reason == UCNV_UNASSIGNED && (out_char = FallbackForGBK(code_point))) {
    const UChar* source = &out_char;
    *err = U_ZERO_ERROR;
    ucnv_cbFromUWriteUChars(from_unicode_args, &source, source + 1, 0, err);
    return;
  }
  UCNV_FROM_U_CALLBACK_SUBSTITUTE(context, from_unicode_args, code_units,
                                  length, code_point, reason, err);
}
#endif  // USING_SYSTEM_ICU

static void NotReachedEntityCallback(const void* context,
                                     UConverterFromUnicodeArgs* from_u_args,
                                     const UChar* code_units,
                                     int32_t length,
                                     UChar32 code_point,
                                     UConverterCallbackReason reason,
                                     UErrorCode* err) {
  NOTREACHED();
}

class TextCodecInput final {
  STACK_ALLOCATED();

 public:
  TextCodecInput(const TextEncoding& encoding,
                 const UChar* characters,
                 wtf_size_t length)
      : begin_(characters), end_(characters + length) {}

  TextCodecInput(const TextEncoding& encoding,
                 const LChar* characters,
                 wtf_size_t length) {
    buffer_.ReserveInitialCapacity(length);
    for (wtf_size_t i = 0; i < length; ++i)
      buffer_.push_back(characters[i]);
    begin_ = buffer_.data();
    end_ = begin_ + buffer_.size();
  }

  const UChar* begin() const { return begin_; }
  const UChar* end() const { return end_; }

 private:
  const UChar* begin_;
  const UChar* end_;
  Vector<UChar> buffer_;
};

std::string TextCodecICU::EncodeInternal(const TextCodecInput& input,
                                         UnencodableHandling handling) {
  const UChar* source = input.begin();
  const UChar* end = input.end();

  UErrorCode err = U_ZERO_ERROR;

  switch (handling) {
    case kEntitiesForUnencodables:
#if !defined(USING_SYSTEM_ICU)
      ucnv_setFromUCallBack(converter_icu_, NumericEntityCallback, nullptr,
                            nullptr, nullptr, &err);
#else
      ucnv_setFromUCallBack(
          converter_icu_,
          needs_gbk_fallbacks_ ? GbkCallbackEscape : NumericEntityCallback, 0,
          0, 0, &err);
#endif
      break;
    case kURLEncodedEntitiesForUnencodables:
#if !defined(USING_SYSTEM_ICU)
      ucnv_setFromUCallBack(converter_icu_, UrlEscapedEntityCallback, nullptr,
                            nullptr, nullptr, &err);
#else
      ucnv_setFromUCallBack(converter_icu_,
                            needs_gbk_fallbacks_ ? GbkUrlEscapedEntityCallack
                                                 : UrlEscapedEntityCallback,
                            0, 0, 0, &err);
#endif
      break;
    case kCSSEncodedEntitiesForUnencodables:
#if !defined(USING_SYSTEM_ICU)
      ucnv_setFromUCallBack(converter_icu_, CssEscapedEntityCallback, nullptr,
                            nullptr, nullptr, &err);
#else
      ucnv_setFromUCallBack(converter_icu_,
                            needs_gbk_fallbacks_ ? GbkCssEscapedEntityCallack
                                                 : CssEscapedEntityCallback,
                            0, 0, 0, &err);
#endif
      break;
    case kNoUnencodables:
      DCHECK(encoding_ == UTF16BigEndianEncoding() ||
             encoding_ == UTF16LittleEndianEncoding() ||
             encoding_ == UTF8Encoding());
      ucnv_setFromUCallBack(converter_icu_, NotReachedEntityCallback, nullptr,
                            nullptr, nullptr, &err);
      break;
  }

  DCHECK(U_SUCCESS(err));
  if (U_FAILURE(err))
    return std::string();

  Vector<char> result;
  wtf_size_t size = 0;
  do {
    char buffer[kConversionBufferSize];
    char* target = buffer;
    char* target_limit = target + kConversionBufferSize;
    err = U_ZERO_ERROR;
    ucnv_fromUnicode(converter_icu_, &target, target_limit, &source, end,
                     nullptr, true, &err);
    wtf_size_t count = static_cast<wtf_size_t>(target - buffer);
    result.Grow(size + count);
    memcpy(result.data() + size, buffer, count);
    size += count;
  } while (err == U_BUFFER_OVERFLOW_ERROR);

  return std::string(result.data(), size);
}

template <typename CharType>
std::string TextCodecICU::EncodeCommon(const CharType* characters,
                                       wtf_size_t length,
                                       UnencodableHandling handling) {
  if (!length)
    return "";

  if (!converter_icu_)
    CreateICUConverter();
  if (!converter_icu_)
    return std::string();

  TextCodecInput input(encoding_, characters, length);
  return EncodeInternal(input, handling);
}

std::string TextCodecICU::Encode(const UChar* characters,
                                 wtf_size_t length,
                                 UnencodableHandling handling) {
  return EncodeCommon(characters, length, handling);
}

std::string TextCodecICU::Encode(const LChar* characters,
                                 wtf_size_t length,
                                 UnencodableHandling handling) {
  return EncodeCommon(characters, length, handling);
}

}  // namespace WTF