1 // Copyright 2014 The Chromium Authors. All rights reserved. 2 // Use of this source code is governed by a BSD-style license that can be 3 // found in the LICENSE file. 4 5 #include "net/base/net_string_util.h" 6 7 #include "base/i18n/i18n_constants.h" 8 #include "base/i18n/icu_string_conversions.h" 9 #include "base/strings/string_util.h" 10 #include "third_party/icu/source/common/unicode/ucnv.h" 11 12 namespace net { 13 14 const char* const kCharsetLatin1 = base::kCodepageLatin1; 15 16 bool ConvertToUtf8(const std::string& text, const char* charset, 17 std::string* output) { 18 output->clear(); 19 20 UErrorCode err = U_ZERO_ERROR; 21 UConverter* converter(ucnv_open(charset, &err)); 22 if (U_FAILURE(err)) 23 return false; 24 25 // A single byte in a legacy encoding can be expanded to 3 bytes in UTF-8. 26 // A 'two-byte character' in a legacy encoding can be expanded to 4 bytes 27 // in UTF-8. Therefore, the expansion ratio is 3 at most. Add one for a 28 // trailing '\0'. 29 size_t output_length = text.length() * 3 + 1; 30 char* buf = WriteInto(output, output_length); 31 output_length = ucnv_toAlgorithmic(UCNV_UTF8, converter, buf, output_length, 32 text.data(), text.length(), &err); 33 ucnv_close(converter); 34 if (U_FAILURE(err)) { 35 output->clear(); 36 return false; 37 } 38 39 output->resize(output_length); 40 return true; 41 } 42 43 bool ConvertToUtf8AndNormalize(const std::string& text, const char* charset, 44 std::string* output) { 45 return base::ConvertToUtf8AndNormalize(text, charset, output); 46 } 47 48 bool ConvertToUTF16(const std::string& text, const char* charset, 49 base::string16* output) { 50 return base::CodepageToUTF16(text, charset, 51 base::OnStringConversionError::FAIL, output); 52 } 53 54 bool ConvertToUTF16WithSubstitutions(const std::string& text, 55 const char* charset, 56 base::string16* output) { 57 return base::CodepageToUTF16(text, charset, 58 base::OnStringConversionError::SUBSTITUTE, 59 output); 60 } 61 62 } // namespace net 63
