chrome/browser/character_encoding.cc

   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
   3 // found in the LICENSE file.
   4
   5 #include "chrome/browser/character_encoding.h"
   6
   7 #include <map>
   8 #include <set>
   9
  10 #include "base/logging.h"
  11 #include "base/memory/scoped_ptr.h"
  12 #include "base/strings/string_tokenizer.h"
  13 #include "base/strings/string_util.h"
  14 #include "base/strings/utf_string_conversions.h"
  15 #include "chrome/app/chrome_command_ids.h"
  16 #include "content/public/browser/browser_thread.h"
  17 #include "grit/generated_resources.h"
  18 #include "third_party/icu/source/common/unicode/ucnv.h"
  19 #include "ui/base/l10n/l10n_util.h"
  20 #include "ui/base/l10n/l10n_util_collator.h"
  21
  22 using content::BrowserThread;
  23
  24 namespace {
  25
  26 // The maximum length of short list of recently user selected encodings is 3.
  27 const size_t kUserSelectedEncodingsMaxLength = 3;
  28
  29 typedef struct {
  30   int resource_id;
  31   const char* name;
  32   int category_string_id;
  33 } CanonicalEncodingData;
  34
  35 // An array of all supported canonical encoding names.
  36 const CanonicalEncodingData kCanonicalEncodingNames[] = {
  37   { IDC_ENCODING_UTF8, "UTF-8", IDS_ENCODING_UNICODE },
  38   { IDC_ENCODING_UTF16LE, "UTF-16LE", IDS_ENCODING_UNICODE },
  39   { IDC_ENCODING_ISO88591, "ISO-8859-1", IDS_ENCODING_WESTERN },
  40   { IDC_ENCODING_WINDOWS1252, "windows-1252", IDS_ENCODING_WESTERN },
  41   { IDC_ENCODING_GBK, "GBK", IDS_ENCODING_SIMP_CHINESE },
  42   { IDC_ENCODING_GB18030, "gb18030", IDS_ENCODING_SIMP_CHINESE },
  43   { IDC_ENCODING_BIG5, "Big5", IDS_ENCODING_TRAD_CHINESE },
  44   { IDC_ENCODING_BIG5HKSCS, "Big5-HKSCS", IDS_ENCODING_TRAD_CHINESE },
  45   { IDC_ENCODING_KOREAN, "EUC-KR", IDS_ENCODING_KOREAN },
  46   { IDC_ENCODING_SHIFTJIS, "Shift_JIS", IDS_ENCODING_JAPANESE },
  47   { IDC_ENCODING_EUCJP, "EUC-JP", IDS_ENCODING_JAPANESE },
  48   { IDC_ENCODING_ISO2022JP, "ISO-2022-JP", IDS_ENCODING_JAPANESE },
  49   { IDC_ENCODING_THAI, "windows-874", IDS_ENCODING_THAI },
  50   { IDC_ENCODING_ISO885915, "ISO-8859-15", IDS_ENCODING_WESTERN },
  51   { IDC_ENCODING_MACINTOSH, "macintosh", IDS_ENCODING_WESTERN },
  52   { IDC_ENCODING_ISO88592, "ISO-8859-2", IDS_ENCODING_CENTRAL_EUROPEAN },
  53   { IDC_ENCODING_WINDOWS1250, "windows-1250", IDS_ENCODING_CENTRAL_EUROPEAN },
  54   { IDC_ENCODING_ISO88595, "ISO-8859-5", IDS_ENCODING_CYRILLIC },
  55   { IDC_ENCODING_WINDOWS1251, "windows-1251", IDS_ENCODING_CYRILLIC },
  56   { IDC_ENCODING_KOI8R, "KOI8-R", IDS_ENCODING_CYRILLIC },
  57   { IDC_ENCODING_KOI8U, "KOI8-U", IDS_ENCODING_CYRILLIC },
  58   { IDC_ENCODING_ISO88597, "ISO-8859-7", IDS_ENCODING_GREEK },
  59   { IDC_ENCODING_WINDOWS1253, "windows-1253", IDS_ENCODING_GREEK },
  60   { IDC_ENCODING_WINDOWS1254, "windows-1254", IDS_ENCODING_TURKISH },
  61   { IDC_ENCODING_WINDOWS1256, "windows-1256", IDS_ENCODING_ARABIC },
  62   { IDC_ENCODING_ISO88596, "ISO-8859-6", IDS_ENCODING_ARABIC },
  63   { IDC_ENCODING_WINDOWS1255, "windows-1255", IDS_ENCODING_HEBREW },
  64   { IDC_ENCODING_ISO88598I, "ISO-8859-8-I", IDS_ENCODING_HEBREW },
  65   { IDC_ENCODING_ISO88598, "ISO-8859-8", IDS_ENCODING_HEBREW },
  66   { IDC_ENCODING_WINDOWS1258, "windows-1258", IDS_ENCODING_VIETNAMESE },
  67   { IDC_ENCODING_ISO88594, "ISO-8859-4", IDS_ENCODING_BALTIC },
  68   { IDC_ENCODING_ISO885913, "ISO-8859-13", IDS_ENCODING_BALTIC },
  69   { IDC_ENCODING_WINDOWS1257, "windows-1257", IDS_ENCODING_BALTIC },
  70   { IDC_ENCODING_ISO88593, "ISO-8859-3", IDS_ENCODING_SOUTH_EUROPEAN },
  71   { IDC_ENCODING_ISO885910, "ISO-8859-10", IDS_ENCODING_NORDIC },
  72   { IDC_ENCODING_ISO885914, "ISO-8859-14", IDS_ENCODING_CELTIC },
  73   { IDC_ENCODING_ISO885916, "ISO-8859-16", IDS_ENCODING_ROMANIAN },
  74 };
  75
  76 const int kCanonicalEncodingNamesLength = arraysize(kCanonicalEncodingNames);
  77
  78 typedef std::map<int, std::pair<const char*, int> >
  79     IdToCanonicalEncodingNameMapType;
  80 typedef std::map<const std::string, int> CanonicalEncodingNameToIdMapType;
  81
  82 typedef struct {
  83   const char* canonical_form;
  84   const char* display_form;
  85 } CanonicalEncodingDisplayNamePair;
  86
  87 const CanonicalEncodingDisplayNamePair kCanonicalDisplayNameOverrides[] = {
  88   // Only lists the canonical names where we want a different form for display.
  89   { "macintosh", "Macintosh" },
  90   { "windows-874", "Windows-874" },
  91   { "windows-1250", "Windows-1250" },
  92   { "windows-1251", "Windows-1251" },
  93   { "windows-1252", "Windows-1252" },
  94   { "windows-1253", "Windows-1253" },
  95   { "windows-1254", "Windows-1254" },
  96   { "windows-1255", "Windows-1255" },
  97   { "windows-1256", "Windows-1256" },
  98   { "windows-1257", "Windows-1257" },
  99   { "windows-1258", "Windows-1258" },
 100 };
 101
 102 const int kCanonicalDisplayNameOverridesLength =
 103     arraysize(kCanonicalDisplayNameOverrides);
 104
 105 typedef std::map<std::string, const char*> CanonicalNameDisplayNameMapType;
 106
 107 class CanonicalEncodingMap {
 108  public:
 109   CanonicalEncodingMap() {}
 110   const IdToCanonicalEncodingNameMapType* GetIdToCanonicalEncodingNameMapData();
 111   const CanonicalEncodingNameToIdMapType* GetCanonicalEncodingNameToIdMapData();
 112   const CanonicalNameDisplayNameMapType* GetCanonicalNameDisplayNameMapData();
 113   std::vector<int>* locale_dependent_encoding_ids() {
 114     return &locale_dependent_encoding_ids_;
 115   }
 116
 117   std::vector<CharacterEncoding::EncodingInfo>* current_display_encodings() {
 118     return &current_display_encodings_;
 119   }
 120
 121  private:
 122   scoped_ptr<IdToCanonicalEncodingNameMapType> id_to_encoding_name_map_;
 123   scoped_ptr<CanonicalEncodingNameToIdMapType> encoding_name_to_id_map_;
 124   scoped_ptr<CanonicalNameDisplayNameMapType>
 125       encoding_name_to_display_name_map_;
 126   std::vector<int> locale_dependent_encoding_ids_;
 127   std::vector<CharacterEncoding::EncodingInfo> current_display_encodings_;
 128
 129   DISALLOW_COPY_AND_ASSIGN(CanonicalEncodingMap);
 130 };
 131
 132 const IdToCanonicalEncodingNameMapType*
 133     CanonicalEncodingMap::GetIdToCanonicalEncodingNameMapData() {
 134   // Testing and building map is not thread safe, this function is supposed to
 135   // only run in UI thread. Myabe I should add a lock in here for making it as
 136   // thread safe.
 137   if (!id_to_encoding_name_map_.get()) {
 138     id_to_encoding_name_map_.reset(new IdToCanonicalEncodingNameMapType);
 139     for (int i = 0; i < kCanonicalEncodingNamesLength; ++i) {
 140       int resource_id = kCanonicalEncodingNames[i].resource_id;
 141       (*id_to_encoding_name_map_)[resource_id] =
 142         std::make_pair(kCanonicalEncodingNames[i].name,
 143                        kCanonicalEncodingNames[i].category_string_id);
 144     }
 145   }
 146   return id_to_encoding_name_map_.get();
 147 }
 148
 149 const CanonicalEncodingNameToIdMapType*
 150     CanonicalEncodingMap::GetCanonicalEncodingNameToIdMapData() {
 151   if (!encoding_name_to_id_map_.get()) {
 152     encoding_name_to_id_map_.reset(new CanonicalEncodingNameToIdMapType);
 153     for (int i = 0; i < kCanonicalEncodingNamesLength; ++i) {
 154       (*encoding_name_to_id_map_)[kCanonicalEncodingNames[i].name] =
 155           kCanonicalEncodingNames[i].resource_id;
 156     }
 157   }
 158   return encoding_name_to_id_map_.get();
 159 }
 160
 161 const CanonicalNameDisplayNameMapType*
 162     CanonicalEncodingMap::GetCanonicalNameDisplayNameMapData() {
 163   if (!encoding_name_to_display_name_map_.get()) {
 164     encoding_name_to_display_name_map_.reset(
 165         new CanonicalNameDisplayNameMapType);
 166     // First store the names in the kCanonicalEncodingNames list.
 167     for (int i = 0; i < kCanonicalEncodingNamesLength; ++i) {
 168       (*encoding_name_to_display_name_map_)[kCanonicalEncodingNames[i].name] =
 169           kCanonicalEncodingNames[i].name;
 170     }
 171     // Then save in the overrides.
 172     for (int i = 0; i < kCanonicalDisplayNameOverridesLength; ++i) {
 173       (*encoding_name_to_display_name_map_)
 174           [kCanonicalDisplayNameOverrides[i].canonical_form] =
 175           kCanonicalDisplayNameOverrides[i].display_form;
 176     }
 177     DCHECK(static_cast<int>(encoding_name_to_display_name_map_->size()) ==
 178            kCanonicalEncodingNamesLength)
 179         << "Got an override that wasn't in the encoding list";
 180   }
 181   return encoding_name_to_display_name_map_.get();
 182 }
 183
 184 // A static map object which contains all resourceid-nonsequenced canonical
 185 // encoding names.
 186 CanonicalEncodingMap* CanonicalEncodingMapSingleton() {
 187   DCHECK(BrowserThread::CurrentlyOn(BrowserThread::UI));
 188   static CanonicalEncodingMap* singleton = new CanonicalEncodingMap;
 189   return singleton;
 190 }
 191
 192 const int kDefaultEncodingMenus[] = {
 193   IDC_ENCODING_UTF16LE,
 194   IDC_ENCODING_ISO88591,
 195   IDC_ENCODING_WINDOWS1252,
 196   IDC_ENCODING_GBK,
 197   IDC_ENCODING_GB18030,
 198   IDC_ENCODING_BIG5,
 199   IDC_ENCODING_BIG5HKSCS,
 200   IDC_ENCODING_KOREAN,
 201   IDC_ENCODING_SHIFTJIS,
 202   IDC_ENCODING_EUCJP,
 203   IDC_ENCODING_ISO2022JP,
 204   IDC_ENCODING_THAI,
 205   IDC_ENCODING_ISO885915,
 206   IDC_ENCODING_MACINTOSH,
 207   IDC_ENCODING_ISO88592,
 208   IDC_ENCODING_WINDOWS1250,
 209   IDC_ENCODING_ISO88595,
 210   IDC_ENCODING_WINDOWS1251,
 211   IDC_ENCODING_KOI8R,
 212   IDC_ENCODING_KOI8U,
 213   IDC_ENCODING_ISO88597,
 214   IDC_ENCODING_WINDOWS1253,
 215   IDC_ENCODING_WINDOWS1254,
 216   IDC_ENCODING_WINDOWS1256,
 217   IDC_ENCODING_ISO88596,
 218   IDC_ENCODING_WINDOWS1255,
 219   IDC_ENCODING_ISO88598I,
 220   IDC_ENCODING_ISO88598,
 221   IDC_ENCODING_WINDOWS1258,
 222   IDC_ENCODING_ISO88594,
 223   IDC_ENCODING_ISO885913,
 224   IDC_ENCODING_WINDOWS1257,
 225   IDC_ENCODING_ISO88593,
 226   IDC_ENCODING_ISO885910,
 227   IDC_ENCODING_ISO885914,
 228   IDC_ENCODING_ISO885916,
 229 };
 230
 231 const int kDefaultEncodingMenusLength = arraysize(kDefaultEncodingMenus);
 232
 233 // Parse the input |encoding_list| which is a encoding list separated with
 234 // comma, get available encoding ids and save them to |available_list|.
 235 // The parameter |maximum_size| indicates maximum size of encoding items we
 236 // want to get from the |encoding_list|.
 237 void ParseEncodingListSeparatedWithComma(
 238     const std::string& encoding_list, std::vector<int>* const available_list,
 239     size_t maximum_size) {
 240   base::StringTokenizer tokenizer(encoding_list, ",");
 241   while (tokenizer.GetNext()) {
 242     int id = CharacterEncoding::GetCommandIdByCanonicalEncodingName(
 243         tokenizer.token());
 244     // Ignore invalid encoding.
 245     if (!id)
 246       continue;
 247     available_list->push_back(id);
 248     if (available_list->size() == maximum_size)
 249       return;
 250   }
 251 }
 252
 253 base::string16 GetEncodingDisplayName(const std::string& encoding_name,
 254                                       int category_string_id) {
 255   base::string16 category_name = l10n_util::GetStringUTF16(category_string_id);
 256   if (category_string_id != IDS_ENCODING_KOREAN &&
 257       category_string_id != IDS_ENCODING_THAI &&
 258       category_string_id != IDS_ENCODING_TURKISH) {
 259     const CanonicalNameDisplayNameMapType* map =
 260         CanonicalEncodingMapSingleton()->GetCanonicalNameDisplayNameMapData();
 261     DCHECK(map);
 262
 263     CanonicalNameDisplayNameMapType::const_iterator found_name =
 264         map->find(encoding_name);
 265     DCHECK(found_name != map->end());
 266     return l10n_util::GetStringFUTF16(IDS_ENCODING_DISPLAY_TEMPLATE,
 267                                       category_name,
 268                                       base::ASCIIToUTF16(found_name->second));
 269   }
 270   return category_name;
 271 }
 272
 273 int GetEncodingCategoryStringIdByCommandId(int id) {
 274   const IdToCanonicalEncodingNameMapType* map =
 275       CanonicalEncodingMapSingleton()->GetIdToCanonicalEncodingNameMapData();
 276   DCHECK(map);
 277
 278   IdToCanonicalEncodingNameMapType::const_iterator found_name = map->find(id);
 279   if (found_name != map->end())
 280     return found_name->second.second;
 281   return 0;
 282 }
 283
 284 std::string GetEncodingCategoryStringByCommandId(int id) {
 285   int category_id = GetEncodingCategoryStringIdByCommandId(id);
 286   if (category_id)
 287     return l10n_util::GetStringUTF8(category_id);
 288   return std::string();
 289 }
 290
 291 }  // namespace
 292
 293 CharacterEncoding::EncodingInfo::EncodingInfo(int id)
 294     : encoding_id(id) {
 295   encoding_category_name =
 296       base::UTF8ToUTF16(GetEncodingCategoryStringByCommandId(id));
 297   encoding_display_name = GetCanonicalEncodingDisplayNameByCommandId(id);
 298 }
 299
 300 // Static.
 301 int CharacterEncoding::GetCommandIdByCanonicalEncodingName(
 302     const std::string& encoding_name) {
 303   const CanonicalEncodingNameToIdMapType* map =
 304       CanonicalEncodingMapSingleton()->GetCanonicalEncodingNameToIdMapData();
 305   DCHECK(map);
 306
 307   CanonicalEncodingNameToIdMapType::const_iterator found_id =
 308       map->find(encoding_name);
 309   if (found_id != map->end())
 310     return found_id->second;
 311   return 0;
 312 }
 313
 314 // Static.
 315 std::string CharacterEncoding::GetCanonicalEncodingNameByCommandId(int id) {
 316   const IdToCanonicalEncodingNameMapType* map =
 317       CanonicalEncodingMapSingleton()->GetIdToCanonicalEncodingNameMapData();
 318   DCHECK(map);
 319
 320   IdToCanonicalEncodingNameMapType::const_iterator found_name = map->find(id);
 321   if (found_name != map->end())
 322     return found_name->second.first;
 323   return std::string();
 324 }
 325
 326 // Static.
 327 base::string16 CharacterEncoding::GetCanonicalEncodingDisplayNameByCommandId(
 328     int id) {
 329   const IdToCanonicalEncodingNameMapType* map =
 330       CanonicalEncodingMapSingleton()->GetIdToCanonicalEncodingNameMapData();
 331   DCHECK(map);
 332
 333   IdToCanonicalEncodingNameMapType::const_iterator found_name = map->find(id);
 334   if (found_name != map->end())
 335     return GetEncodingDisplayName(found_name->second.first,
 336                                   found_name->second.second);
 337   return base::string16();
 338 }
 339
 340 // Static.
 341 // Return count number of all supported canonical encoding.
 342 int CharacterEncoding::GetSupportCanonicalEncodingCount() {
 343   return kCanonicalEncodingNamesLength;
 344 }
 345
 346 // Static.
 347 std::string CharacterEncoding::GetCanonicalEncodingNameByIndex(int index) {
 348   if (index < kCanonicalEncodingNamesLength)
 349     return kCanonicalEncodingNames[index].name;
 350   return std::string();
 351 }
 352
 353 // Static.
 354 base::string16 CharacterEncoding::GetCanonicalEncodingDisplayNameByIndex(
 355     int index) {
 356   if (index < kCanonicalEncodingNamesLength)
 357     return GetEncodingDisplayName(kCanonicalEncodingNames[index].name,
 358         kCanonicalEncodingNames[index].category_string_id);
 359   return base::string16();
 360 }
 361
 362 // Static.
 363 int CharacterEncoding::GetEncodingCommandIdByIndex(int index) {
 364   if (index < kCanonicalEncodingNamesLength)
 365     return kCanonicalEncodingNames[index].resource_id;
 366   return 0;
 367 }
 368
 369 // Static.
 370 std::string CharacterEncoding::GetCanonicalEncodingNameByAliasName(
 371     const std::string& alias_name) {
 372   // If the input alias_name is already canonical encoding name, just return it.
 373   const CanonicalEncodingNameToIdMapType* map =
 374       CanonicalEncodingMapSingleton()->GetCanonicalEncodingNameToIdMapData();
 375   DCHECK(map);
 376
 377   CanonicalEncodingNameToIdMapType::const_iterator found_id =
 378       map->find(alias_name);
 379   if (found_id != map->end())
 380     return alias_name;
 381
 382   UErrorCode error_code = U_ZERO_ERROR;
 383
 384   const char* canonical_name = ucnv_getCanonicalName(
 385       alias_name.c_str(), "MIME", &error_code);
 386   // If failed,  then try IANA next.
 387   if (U_FAILURE(error_code) || !canonical_name) {
 388     error_code = U_ZERO_ERROR;
 389     canonical_name = ucnv_getCanonicalName(
 390       alias_name.c_str(), "IANA", &error_code);
 391   }
 392
 393   if (canonical_name) {
 394     // TODO(jnd) use a map to handle all customized {alias, canonical}
 395     // encoding mappings if we have more than one pair.
 396     // We don't want to add an unnecessary charset to the encoding menu, so we
 397     // alias 'US-ASCII' to 'ISO-8859-1' in our UI without touching WebKit.
 398     // http://crbug.com/15801.
 399     if (alias_name == "US-ASCII")
 400       return GetCanonicalEncodingNameByCommandId(IDC_ENCODING_ISO88591);
 401     return canonical_name;
 402   } else {
 403     return std::string();
 404   }
 405 }
 406
 407 // Static
 408 // According to the behavior of user recently selected encoding short list in
 409 // Firefox, we always put UTF-8 as top position, after then put user
 410 // recent selected encodings, then put local dependent encoding items.
 411 // At last, we put all remaining encoding items.
 412 const std::vector<CharacterEncoding::EncodingInfo>*
 413     CharacterEncoding::GetCurrentDisplayEncodings(
 414     const std::string& locale,
 415     const std::string& locale_encodings,
 416     const std::string& recently_select_encodings) {
 417   std::vector<int>* const locale_dependent_encoding_list =
 418       CanonicalEncodingMapSingleton()->locale_dependent_encoding_ids();
 419   std::vector<CharacterEncoding::EncodingInfo>* const encoding_list =
 420       CanonicalEncodingMapSingleton()->current_display_encodings();
 421
 422   // Initialize locale dependent static encoding list.
 423   if (locale_dependent_encoding_list->empty() && !locale_encodings.empty())
 424     ParseEncodingListSeparatedWithComma(locale_encodings,
 425                                         locale_dependent_encoding_list,
 426                                         kUserSelectedEncodingsMaxLength);
 427
 428   CR_DEFINE_STATIC_LOCAL(std::string, cached_user_selected_encodings, ());
 429   // Build current display encoding list.
 430   if (encoding_list->empty() ||
 431       cached_user_selected_encodings != recently_select_encodings) {
 432     // Update user recently selected encodings.
 433     cached_user_selected_encodings = recently_select_encodings;
 434     // Clear old encoding list since user recently selected encodings changed.
 435     encoding_list->clear();
 436     // Always add UTF-8 to first encoding position.
 437     encoding_list->push_back(EncodingInfo(IDC_ENCODING_UTF8));
 438     std::set<int> inserted_encoding;
 439     inserted_encoding.insert(IDC_ENCODING_UTF8);
 440
 441     // Parse user recently selected encodings and get list
 442     std::vector<int> recently_select_encoding_list;
 443     ParseEncodingListSeparatedWithComma(recently_select_encodings,
 444                                         &recently_select_encoding_list,
 445                                         kUserSelectedEncodingsMaxLength);
 446
 447     // Put 'cached encodings' (dynamic encoding list) after 'local dependent
 448     // encoding list'.
 449     recently_select_encoding_list.insert(recently_select_encoding_list.begin(),
 450         locale_dependent_encoding_list->begin(),
 451         locale_dependent_encoding_list->end());
 452     for (std::vector<int>::iterator it = recently_select_encoding_list.begin();
 453          it != recently_select_encoding_list.end(); ++it) {
 454         // Test whether we have met this encoding id.
 455         bool ok = inserted_encoding.insert(*it).second;
 456         // Duplicated encoding, ignore it. Ideally, this situation should not
 457         // happened, but just in case some one manually edit preference file.
 458         if (!ok)
 459           continue;
 460         encoding_list->push_back(EncodingInfo(*it));
 461     }
 462     // Append a separator;
 463     encoding_list->push_back(EncodingInfo(0));
 464
 465     // We need to keep "Unicode (UTF-16LE)" always at the top (among the rest
 466     // of encodings) instead of being sorted along with other encodings. So if
 467     // "Unicode (UTF-16LE)" is already in previous encodings, sort the rest
 468     // of encodings. Otherwise Put "Unicode (UTF-16LE)" on the first of the
 469     // rest of encodings, skip "Unicode (UTF-16LE)" and sort all left encodings.
 470     int start_sorted_index = encoding_list->size();
 471     if (inserted_encoding.find(IDC_ENCODING_UTF16LE) ==
 472         inserted_encoding.end()) {
 473       encoding_list->push_back(EncodingInfo(IDC_ENCODING_UTF16LE));
 474       inserted_encoding.insert(IDC_ENCODING_UTF16LE);
 475       start_sorted_index++;
 476     }
 477
 478     // Add the rest of encodings that are neither in the static encoding list
 479     // nor in the list of recently selected encodings.
 480     // Build the encoding list sorted in the current locale sorting order.
 481     for (int i = 0; i < kDefaultEncodingMenusLength; ++i) {
 482       int id = kDefaultEncodingMenus[i];
 483       // We have inserted this encoding, skip it.
 484       if (inserted_encoding.find(id) != inserted_encoding.end())
 485         continue;
 486       encoding_list->push_back(EncodingInfo(id));
 487     }
 488     // Sort the encoding list.
 489     l10n_util::SortVectorWithStringKey(locale,
 490                                        encoding_list,
 491                                        start_sorted_index,
 492                                        encoding_list->size(),
 493                                        true);
 494   }
 495   DCHECK(!encoding_list->empty());
 496   return encoding_list;
 497 }
 498
 499 // Static
 500 bool CharacterEncoding::UpdateRecentlySelectedEncoding(
 501     const std::string& original_selected_encodings,
 502     int new_selected_encoding_id,
 503     std::string* selected_encodings) {
 504   // Get encoding name.
 505   std::string encoding_name =
 506       GetCanonicalEncodingNameByCommandId(new_selected_encoding_id);
 507   DCHECK(!encoding_name.empty());
 508   // Check whether the new encoding is in local dependent encodings or original
 509   // recently selected encodings. If yes, do not add it.
 510   std::vector<int>* locale_dependent_encoding_list =
 511       CanonicalEncodingMapSingleton()->locale_dependent_encoding_ids();
 512   DCHECK(locale_dependent_encoding_list);
 513   std::vector<int> selected_encoding_list;
 514   ParseEncodingListSeparatedWithComma(original_selected_encodings,
 515                                       &selected_encoding_list,
 516                                       kUserSelectedEncodingsMaxLength);
 517   // Put 'cached encodings' (dynamic encoding list) after 'local dependent
 518   // encoding list' for check.
 519   std::vector<int> top_encoding_list(*locale_dependent_encoding_list);
 520   // UTF8 is always in our optimized encoding list.
 521   top_encoding_list.insert(top_encoding_list.begin(), IDC_ENCODING_UTF8);
 522   top_encoding_list.insert(top_encoding_list.end(),
 523                            selected_encoding_list.begin(),
 524                            selected_encoding_list.end());
 525   for (std::vector<int>::const_iterator it = top_encoding_list.begin();
 526        it != top_encoding_list.end(); ++it)
 527     if (*it == new_selected_encoding_id)
 528       return false;
 529   // Need to add the encoding id to recently selected encoding list.
 530   // Remove the last encoding in original list.
 531   if (selected_encoding_list.size() == kUserSelectedEncodingsMaxLength)
 532     selected_encoding_list.pop_back();
 533   // Insert new encoding to head of selected encoding list.
 534   *selected_encodings = encoding_name;
 535   // Generate the string for rest selected encoding list.
 536   for (std::vector<int>::const_iterator it = selected_encoding_list.begin();
 537        it != selected_encoding_list.end(); ++it) {
 538     selected_encodings->append(1, L',');
 539     selected_encodings->append(GetCanonicalEncodingNameByCommandId(*it));
 540   }
 541   return true;
 542 }