OLD | NEW |
1 // Copyright (c) 2012 The Chromium Authors. All rights reserved. | 1 // Copyright (c) 2012 The Chromium Authors. All rights reserved. |
2 // Use of this source code is governed by a BSD-style license that can be | 2 // Use of this source code is governed by a BSD-style license that can be |
3 // found in the LICENSE file. | 3 // found in the LICENSE file. |
4 | 4 |
5 #include "base/i18n/icu_string_conversions.h" | 5 #include "base/i18n/icu_string_conversions.h" |
6 | 6 |
7 #include <vector> | 7 #include <vector> |
8 | 8 |
9 #include "base/basictypes.h" | 9 #include "base/basictypes.h" |
10 #include "base/logging.h" | 10 #include "base/logging.h" |
(...skipping 116 matching lines...) Expand 10 before | Expand all | Expand 10 after Loading... |
127 break; | 127 break; |
128 case OnStringConversionError::SUBSTITUTE: | 128 case OnStringConversionError::SUBSTITUTE: |
129 ucnv_setToUCallBack(converter, ToUnicodeCallbackSubstitute, 0, | 129 ucnv_setToUCallBack(converter, ToUnicodeCallbackSubstitute, 0, |
130 NULL, NULL, status); | 130 NULL, NULL, status); |
131 break; | 131 break; |
132 default: | 132 default: |
133 NOTREACHED(); | 133 NOTREACHED(); |
134 } | 134 } |
135 } | 135 } |
136 | 136 |
137 inline UConverterType utf32_platform_endian() { | |
138 #if U_IS_BIG_ENDIAN | |
139 return UCNV_UTF32_BigEndian; | |
140 #else | |
141 return UCNV_UTF32_LittleEndian; | |
142 #endif | |
143 } | |
144 | |
145 } // namespace | 137 } // namespace |
146 | 138 |
147 // Codepage <-> Wide/UTF-16 --------------------------------------------------- | 139 // Codepage <-> Wide/UTF-16 --------------------------------------------------- |
148 | 140 |
149 bool UTF16ToCodepage(const string16& utf16, | 141 bool UTF16ToCodepage(const string16& utf16, |
150 const char* codepage_name, | 142 const char* codepage_name, |
151 OnStringConversionError::Type on_error, | 143 OnStringConversionError::Type on_error, |
152 std::string* encoded) { | 144 std::string* encoded) { |
153 encoded->clear(); | 145 encoded->clear(); |
154 | 146 |
(...skipping 35 matching lines...) Expand 10 before | Expand all | Expand 10 after Loading... |
190 ucnv_close(converter); | 182 ucnv_close(converter); |
191 if (!U_SUCCESS(status)) { | 183 if (!U_SUCCESS(status)) { |
192 utf16->clear(); // Make sure the output is empty on error. | 184 utf16->clear(); // Make sure the output is empty on error. |
193 return false; | 185 return false; |
194 } | 186 } |
195 | 187 |
196 utf16->assign(buffer.get(), actual_size); | 188 utf16->assign(buffer.get(), actual_size); |
197 return true; | 189 return true; |
198 } | 190 } |
199 | 191 |
200 bool WideToCodepage(const std::wstring& wide, | |
201 const char* codepage_name, | |
202 OnStringConversionError::Type on_error, | |
203 std::string* encoded) { | |
204 #if defined(WCHAR_T_IS_UTF16) | |
205 return UTF16ToCodepage(wide, codepage_name, on_error, encoded); | |
206 #elif defined(WCHAR_T_IS_UTF32) | |
207 encoded->clear(); | |
208 | |
209 UErrorCode status = U_ZERO_ERROR; | |
210 UConverter* converter = ucnv_open(codepage_name, &status); | |
211 if (!U_SUCCESS(status)) | |
212 return false; | |
213 | |
214 int utf16_len; | |
215 // When wchar_t is wider than UChar (16 bits), transform |wide| into a | |
216 // UChar* string. Size the UChar* buffer to be large enough to hold twice | |
217 // as many UTF-16 code units (UChar's) as there are Unicode code points, | |
218 // in case each code points translates to a UTF-16 surrogate pair, | |
219 // and leave room for a NUL terminator. | |
220 std::vector<UChar> utf16(wide.length() * 2 + 1); | |
221 u_strFromUTF32(&utf16[0], utf16.size(), &utf16_len, | |
222 reinterpret_cast<const UChar32*>(wide.c_str()), | |
223 wide.length(), &status); | |
224 DCHECK(U_SUCCESS(status)) << "failed to convert wstring to UChar*"; | |
225 | |
226 return ConvertFromUTF16(converter, &utf16[0], utf16_len, on_error, encoded); | |
227 #endif // defined(WCHAR_T_IS_UTF32) | |
228 } | |
229 | |
230 bool CodepageToWide(const std::string& encoded, | |
231 const char* codepage_name, | |
232 OnStringConversionError::Type on_error, | |
233 std::wstring* wide) { | |
234 #if defined(WCHAR_T_IS_UTF16) | |
235 return CodepageToUTF16(encoded, codepage_name, on_error, wide); | |
236 #elif defined(WCHAR_T_IS_UTF32) | |
237 wide->clear(); | |
238 | |
239 UErrorCode status = U_ZERO_ERROR; | |
240 UConverter* converter = ucnv_open(codepage_name, &status); | |
241 if (!U_SUCCESS(status)) | |
242 return false; | |
243 | |
244 // The maximum length in 4 byte unit of UTF-32 output would be | |
245 // at most the same as the number of bytes in input. In the worst | |
246 // case of GB18030 (excluding escaped-based encodings like ISO-2022-JP), | |
247 // this can be 4 times larger than actually needed. | |
248 size_t wchar_max_length = encoded.length() + 1; | |
249 | |
250 SetUpErrorHandlerForToUChars(on_error, converter, &status); | |
251 scoped_ptr<wchar_t[]> buffer(new wchar_t[wchar_max_length]); | |
252 int actual_size = ucnv_toAlgorithmic(utf32_platform_endian(), converter, | |
253 reinterpret_cast<char*>(buffer.get()), | |
254 static_cast<int>(wchar_max_length) * sizeof(wchar_t), encoded.data(), | |
255 static_cast<int>(encoded.length()), &status); | |
256 ucnv_close(converter); | |
257 if (!U_SUCCESS(status)) { | |
258 wide->clear(); // Make sure the output is empty on error. | |
259 return false; | |
260 } | |
261 | |
262 // actual_size is # of bytes. | |
263 wide->assign(buffer.get(), actual_size / sizeof(wchar_t)); | |
264 return true; | |
265 #endif // defined(WCHAR_T_IS_UTF32) | |
266 } | |
267 | |
268 bool ConvertToUtf8AndNormalize(const std::string& text, | 192 bool ConvertToUtf8AndNormalize(const std::string& text, |
269 const std::string& charset, | 193 const std::string& charset, |
270 std::string* result) { | 194 std::string* result) { |
271 result->clear(); | 195 result->clear(); |
272 string16 utf16; | 196 string16 utf16; |
273 if (!CodepageToUTF16( | 197 if (!CodepageToUTF16( |
274 text, charset.c_str(), OnStringConversionError::FAIL, &utf16)) | 198 text, charset.c_str(), OnStringConversionError::FAIL, &utf16)) |
275 return false; | 199 return false; |
276 | 200 |
277 UErrorCode status = U_ZERO_ERROR; | 201 UErrorCode status = U_ZERO_ERROR; |
278 size_t max_length = utf16.length() + 1; | 202 size_t max_length = utf16.length() + 1; |
279 string16 normalized_utf16; | 203 string16 normalized_utf16; |
280 scoped_ptr<char16[]> buffer(new char16[max_length]); | 204 scoped_ptr<char16[]> buffer(new char16[max_length]); |
281 int actual_length = unorm_normalize( | 205 int actual_length = unorm_normalize( |
282 utf16.c_str(), utf16.length(), UNORM_NFC, 0, | 206 utf16.c_str(), utf16.length(), UNORM_NFC, 0, |
283 buffer.get(), static_cast<int>(max_length), &status); | 207 buffer.get(), static_cast<int>(max_length), &status); |
284 if (!U_SUCCESS(status)) | 208 if (!U_SUCCESS(status)) |
285 return false; | 209 return false; |
286 normalized_utf16.assign(buffer.get(), actual_length); | 210 normalized_utf16.assign(buffer.get(), actual_length); |
287 | 211 |
288 return UTF16ToUTF8(normalized_utf16.data(), | 212 return UTF16ToUTF8(normalized_utf16.data(), |
289 normalized_utf16.length(), result); | 213 normalized_utf16.length(), result); |
290 } | 214 } |
291 | 215 |
292 } // namespace base | 216 } // namespace base |
OLD | NEW |