Home | History | Annotate | Download | only in browser
      1 // Copyright (c) 2011 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/string_tokenizer.h"
     13 #include "base/string_util.h"
     14 #include "base/utf_string_conversions.h"
     15 #include "chrome/app/chrome_command_ids.h"
     16 #include "grit/generated_resources.h"
     17 #include "ui/base/l10n/l10n_util.h"
     18 #include "ui/base/l10n/l10n_util_collator.h"
     19 #include "unicode/ucnv.h"
     20 
     21 namespace {
     22 
     23 // The maximum length of short list of recently user selected encodings is 3.
     24 const size_t kUserSelectedEncodingsMaxLength = 3;
     25 
     26 typedef struct {
     27   int resource_id;
     28   const char* name;
     29   int category_string_id;
     30 } CanonicalEncodingData;
     31 
     32 // An array of all supported canonical encoding names.
     33 static CanonicalEncodingData canonical_encoding_names[] = {
     34   { IDC_ENCODING_UTF8, "UTF-8", IDS_ENCODING_UNICODE },
     35   { IDC_ENCODING_UTF16LE, "UTF-16LE", IDS_ENCODING_UNICODE },
     36   { IDC_ENCODING_ISO88591, "ISO-8859-1", IDS_ENCODING_WESTERN },
     37   { IDC_ENCODING_WINDOWS1252, "windows-1252", IDS_ENCODING_WESTERN },
     38   { IDC_ENCODING_GBK, "GBK", IDS_ENCODING_SIMP_CHINESE },
     39   { IDC_ENCODING_GB18030, "gb18030", IDS_ENCODING_SIMP_CHINESE },
     40   { IDC_ENCODING_BIG5, "Big5", IDS_ENCODING_TRAD_CHINESE },
     41   { IDC_ENCODING_BIG5HKSCS, "Big5-HKSCS", IDS_ENCODING_TRAD_CHINESE },
     42   { IDC_ENCODING_KOREAN, "windows-949", IDS_ENCODING_KOREAN },
     43   { IDC_ENCODING_SHIFTJIS, "Shift_JIS", IDS_ENCODING_JAPANESE },
     44   { IDC_ENCODING_EUCJP, "EUC-JP", IDS_ENCODING_JAPANESE },
     45   { IDC_ENCODING_ISO2022JP, "ISO-2022-JP", IDS_ENCODING_JAPANESE },
     46   { IDC_ENCODING_THAI, "windows-874", IDS_ENCODING_THAI },
     47   { IDC_ENCODING_ISO885915, "ISO-8859-15", IDS_ENCODING_WESTERN },
     48   { IDC_ENCODING_MACINTOSH, "macintosh", IDS_ENCODING_WESTERN },
     49   { IDC_ENCODING_ISO88592, "ISO-8859-2", IDS_ENCODING_CENTRAL_EUROPEAN },
     50   { IDC_ENCODING_WINDOWS1250, "windows-1250", IDS_ENCODING_CENTRAL_EUROPEAN },
     51   { IDC_ENCODING_ISO88595, "ISO-8859-5", IDS_ENCODING_CYRILLIC },
     52   { IDC_ENCODING_WINDOWS1251, "windows-1251", IDS_ENCODING_CYRILLIC },
     53   { IDC_ENCODING_KOI8R, "KOI8-R", IDS_ENCODING_CYRILLIC },
     54   { IDC_ENCODING_KOI8U, "KOI8-U", IDS_ENCODING_CYRILLIC },
     55   { IDC_ENCODING_ISO88597, "ISO-8859-7", IDS_ENCODING_GREEK },
     56   { IDC_ENCODING_WINDOWS1253, "windows-1253", IDS_ENCODING_GREEK },
     57   { IDC_ENCODING_WINDOWS1254, "windows-1254", IDS_ENCODING_TURKISH },
     58   { IDC_ENCODING_WINDOWS1256, "windows-1256", IDS_ENCODING_ARABIC },
     59   { IDC_ENCODING_ISO88596, "ISO-8859-6", IDS_ENCODING_ARABIC },
     60   { IDC_ENCODING_WINDOWS1255, "windows-1255", IDS_ENCODING_HEBREW },
     61   { IDC_ENCODING_ISO88598I, "ISO-8859-8-I", IDS_ENCODING_HEBREW },
     62   { IDC_ENCODING_ISO88598, "ISO-8859-8", IDS_ENCODING_HEBREW },
     63   { IDC_ENCODING_WINDOWS1258, "windows-1258", IDS_ENCODING_VIETNAMESE },
     64   { IDC_ENCODING_ISO88594, "ISO-8859-4", IDS_ENCODING_BALTIC },
     65   { IDC_ENCODING_ISO885913, "ISO-8859-13", IDS_ENCODING_BALTIC },
     66   { IDC_ENCODING_WINDOWS1257, "windows-1257", IDS_ENCODING_BALTIC },
     67   { IDC_ENCODING_ISO88593, "ISO-8859-3", IDS_ENCODING_SOUTH_EUROPEAN },
     68   { IDC_ENCODING_ISO885910, "ISO-8859-10", IDS_ENCODING_NORDIC },
     69   { IDC_ENCODING_ISO885914, "ISO-8859-14", IDS_ENCODING_CELTIC },
     70   { IDC_ENCODING_ISO885916, "ISO-8859-16", IDS_ENCODING_ROMANIAN },
     71 };
     72 
     73 static const int canonical_encoding_names_length =
     74     arraysize(canonical_encoding_names);
     75 
     76 typedef std::map<int, std::pair<const char*, int> >
     77     IdToCanonicalEncodingNameMapType;
     78 typedef std::map<const std::string, int> CanonicalEncodingNameToIdMapType;
     79 
     80 typedef struct {
     81   const char* canonical_form;
     82   const char* display_form;
     83 } CanonicalEncodingDisplayNamePair;
     84 
     85 static CanonicalEncodingDisplayNamePair canonical_display_name_overrides[] = {
     86   // Only lists the canonical names where we want a different form for display.
     87   { "macintosh", "Macintosh" },
     88   { "windows-874", "Windows-874" },
     89   { "windows-949", "Windows-949" },
     90   { "windows-1250", "Windows-1250" },
     91   { "windows-1251", "Windows-1251" },
     92   { "windows-1252", "Windows-1252" },
     93   { "windows-1253", "Windows-1253" },
     94   { "windows-1254", "Windows-1254" },
     95   { "windows-1255", "Windows-1255" },
     96   { "windows-1256", "Windows-1256" },
     97   { "windows-1257", "Windows-1257" },
     98   { "windows-1258", "Windows-1258" },
     99 };
    100 
    101 static const int canonical_display_name_overrides_length =
    102     arraysize(canonical_display_name_overrides);
    103 
    104 typedef std::map<std::string, const char*>
    105     CanonicalNameDisplayNameMapType;
    106 
    107 class CanonicalEncodingMap {
    108  public:
    109   CanonicalEncodingMap()
    110       : id_to_encoding_name_map_(NULL),
    111         encoding_name_to_id_map_(NULL),
    112         encoding_name_to_display_name_map_(NULL) { }
    113   const IdToCanonicalEncodingNameMapType* GetIdToCanonicalEncodingNameMapData();
    114   const CanonicalEncodingNameToIdMapType* GetCanonicalEncodingNameToIdMapData();
    115   const CanonicalNameDisplayNameMapType* GetCanonicalNameDisplayNameMapData();
    116   std::vector<int>* locale_dependent_encoding_ids() {
    117     return &locale_dependent_encoding_ids_;
    118   }
    119 
    120   std::vector<CharacterEncoding::EncodingInfo>* current_display_encodings() {
    121     return &current_display_encodings_;
    122   }
    123 
    124  private:
    125   scoped_ptr<IdToCanonicalEncodingNameMapType> id_to_encoding_name_map_;
    126   scoped_ptr<CanonicalEncodingNameToIdMapType> encoding_name_to_id_map_;
    127   scoped_ptr<CanonicalNameDisplayNameMapType>
    128       encoding_name_to_display_name_map_;
    129   std::vector<int> locale_dependent_encoding_ids_;
    130   std::vector<CharacterEncoding::EncodingInfo> current_display_encodings_;
    131 
    132   DISALLOW_COPY_AND_ASSIGN(CanonicalEncodingMap);
    133 };
    134 
    135 const IdToCanonicalEncodingNameMapType*
    136     CanonicalEncodingMap::GetIdToCanonicalEncodingNameMapData() {
    137   // Testing and building map is not thread safe, this function is supposed to
    138   // only run in UI thread. Myabe I should add a lock in here for making it as
    139   // thread safe.
    140   if (!id_to_encoding_name_map_.get()) {
    141     id_to_encoding_name_map_.reset(new IdToCanonicalEncodingNameMapType);
    142     for (int i = 0; i < canonical_encoding_names_length; ++i) {
    143       int resource_id = canonical_encoding_names[i].resource_id;
    144       (*id_to_encoding_name_map_)[resource_id] =
    145         std::make_pair(canonical_encoding_names[i].name,
    146                        canonical_encoding_names[i].category_string_id);
    147     }
    148   }
    149   return id_to_encoding_name_map_.get();
    150 }
    151 
    152 const CanonicalEncodingNameToIdMapType*
    153     CanonicalEncodingMap::GetCanonicalEncodingNameToIdMapData() {
    154   if (!encoding_name_to_id_map_.get()) {
    155     encoding_name_to_id_map_.reset(new CanonicalEncodingNameToIdMapType);
    156     for (int i = 0; i < canonical_encoding_names_length; ++i) {
    157       (*encoding_name_to_id_map_)[canonical_encoding_names[i].name] =
    158           canonical_encoding_names[i].resource_id;
    159     }
    160   }
    161   return encoding_name_to_id_map_.get();
    162 }
    163 
    164 const CanonicalNameDisplayNameMapType*
    165     CanonicalEncodingMap::GetCanonicalNameDisplayNameMapData() {
    166   if (!encoding_name_to_display_name_map_.get()) {
    167     encoding_name_to_display_name_map_.reset(
    168         new CanonicalNameDisplayNameMapType);
    169     // First store the names in the canonical_encoding_names list.
    170     for (int i = 0; i < canonical_encoding_names_length; ++i) {
    171       (*encoding_name_to_display_name_map_)[canonical_encoding_names[i].name] =
    172           canonical_encoding_names[i].name;
    173     }
    174     // Then save in the overrides.
    175     for (int i = 0; i < canonical_display_name_overrides_length; ++i) {
    176       (*encoding_name_to_display_name_map_)
    177           [canonical_display_name_overrides[i].canonical_form] =
    178           canonical_display_name_overrides[i].display_form;
    179     }
    180     DCHECK(static_cast<int>(encoding_name_to_display_name_map_->size()) ==
    181            canonical_encoding_names_length)
    182         << "Got an override that wasn't in the encoding list";
    183   }
    184   return encoding_name_to_display_name_map_.get();
    185 }
    186 
    187 // A static map object which contains all resourceid-nonsequenced canonical
    188 // encoding names.
    189 static CanonicalEncodingMap canonical_encoding_name_map_singleton;
    190 
    191 const int default_encoding_menus[] = {
    192   IDC_ENCODING_UTF16LE,
    193   IDC_ENCODING_ISO88591,
    194   IDC_ENCODING_WINDOWS1252,
    195   IDC_ENCODING_GBK,
    196   IDC_ENCODING_GB18030,
    197   IDC_ENCODING_BIG5,
    198   IDC_ENCODING_BIG5HKSCS,
    199   IDC_ENCODING_KOREAN,
    200   IDC_ENCODING_SHIFTJIS,
    201   IDC_ENCODING_EUCJP,
    202   IDC_ENCODING_ISO2022JP,
    203   IDC_ENCODING_THAI,
    204   IDC_ENCODING_ISO885915,
    205   IDC_ENCODING_MACINTOSH,
    206   IDC_ENCODING_ISO88592,
    207   IDC_ENCODING_WINDOWS1250,
    208   IDC_ENCODING_ISO88595,
    209   IDC_ENCODING_WINDOWS1251,
    210   IDC_ENCODING_KOI8R,
    211   IDC_ENCODING_KOI8U,
    212   IDC_ENCODING_ISO88597,
    213   IDC_ENCODING_WINDOWS1253,
    214   IDC_ENCODING_WINDOWS1254,
    215   IDC_ENCODING_WINDOWS1256,
    216   IDC_ENCODING_ISO88596,
    217   IDC_ENCODING_WINDOWS1255,
    218   IDC_ENCODING_ISO88598I,
    219   IDC_ENCODING_ISO88598,
    220   IDC_ENCODING_WINDOWS1258,
    221   IDC_ENCODING_ISO88594,
    222   IDC_ENCODING_ISO885913,
    223   IDC_ENCODING_WINDOWS1257,
    224   IDC_ENCODING_ISO88593,
    225   IDC_ENCODING_ISO885910,
    226   IDC_ENCODING_ISO885914,
    227   IDC_ENCODING_ISO885916,
    228 };
    229 
    230 const int default_encoding_menus_length = arraysize(default_encoding_menus);
    231 
    232 // Parse the input |encoding_list| which is a encoding list separated with
    233 // comma, get available encoding ids and save them to |available_list|.
    234 // The parameter |maximum_size| indicates maximum size of encoding items we
    235 // want to get from the |encoding_list|.
    236 void ParseEncodingListSeparatedWithComma(
    237     const std::string& encoding_list, std::vector<int>* const available_list,
    238     size_t maximum_size) {
    239   StringTokenizer tokenizer(encoding_list, ",");
    240   while (tokenizer.GetNext()) {
    241     int id = CharacterEncoding::GetCommandIdByCanonicalEncodingName(
    242         tokenizer.token());
    243     // Ignore invalid encoding.
    244     if (!id)
    245       continue;
    246     available_list->push_back(id);
    247     if (available_list->size() == maximum_size)
    248       return;
    249   }
    250 }
    251 
    252 string16 GetEncodingDisplayName(std::string encoding_name,
    253                                 int category_string_id) {
    254   string16 category_name = l10n_util::GetStringUTF16(category_string_id);
    255   if (category_string_id != IDS_ENCODING_KOREAN &&
    256       category_string_id != IDS_ENCODING_THAI &&
    257       category_string_id != IDS_ENCODING_TURKISH) {
    258     const CanonicalNameDisplayNameMapType* map =
    259         canonical_encoding_name_map_singleton.
    260             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                                       ASCIIToUTF16(found_name->second));
    269   }
    270   return category_name;
    271 }
    272 
    273 int GetEncodingCategoryStringIdByCommandId(int id) {
    274   const IdToCanonicalEncodingNameMapType* map =
    275       canonical_encoding_name_map_singleton.
    276           GetIdToCanonicalEncodingNameMapData();
    277   DCHECK(map);
    278 
    279   IdToCanonicalEncodingNameMapType::const_iterator found_name = map->find(id);
    280   if (found_name != map->end())
    281     return found_name->second.second;
    282   return 0;
    283 }
    284 
    285 std::string GetEncodingCategoryStringByCommandId(int id) {
    286   int category_id = GetEncodingCategoryStringIdByCommandId(id);
    287   if (category_id)
    288     return l10n_util::GetStringUTF8(category_id);
    289   return std::string();
    290 }
    291 
    292 }  // namespace
    293 
    294 CharacterEncoding::EncodingInfo::EncodingInfo(int id)
    295     : encoding_id(id) {
    296   encoding_category_name =
    297       UTF8ToUTF16(GetEncodingCategoryStringByCommandId(id));
    298   encoding_display_name = GetCanonicalEncodingDisplayNameByCommandId(id);
    299 }
    300 
    301 // Static.
    302 int CharacterEncoding::GetCommandIdByCanonicalEncodingName(
    303     const std::string& encoding_name) {
    304   const CanonicalEncodingNameToIdMapType* map =
    305       canonical_encoding_name_map_singleton.
    306           GetCanonicalEncodingNameToIdMapData();
    307   DCHECK(map);
    308 
    309   CanonicalEncodingNameToIdMapType::const_iterator found_id =
    310       map->find(encoding_name);
    311   if (found_id != map->end())
    312     return found_id->second;
    313   return 0;
    314 }
    315 
    316 // Static.
    317 std::string CharacterEncoding::GetCanonicalEncodingNameByCommandId(int id) {
    318   const IdToCanonicalEncodingNameMapType* map =
    319       canonical_encoding_name_map_singleton.
    320           GetIdToCanonicalEncodingNameMapData();
    321   DCHECK(map);
    322 
    323   IdToCanonicalEncodingNameMapType::const_iterator found_name = map->find(id);
    324   if (found_name != map->end())
    325     return found_name->second.first;
    326   return std::string();
    327 }
    328 
    329 // Static.
    330 string16 CharacterEncoding::GetCanonicalEncodingDisplayNameByCommandId(
    331     int id) {
    332   const IdToCanonicalEncodingNameMapType* map =
    333       canonical_encoding_name_map_singleton.
    334           GetIdToCanonicalEncodingNameMapData();
    335   DCHECK(map);
    336 
    337   IdToCanonicalEncodingNameMapType::const_iterator found_name = map->find(id);
    338   if (found_name != map->end())
    339     return GetEncodingDisplayName(found_name->second.first,
    340                                   found_name->second.second);
    341   return string16();
    342 }
    343 
    344 // Static.
    345 // Return count number of all supported canonical encoding.
    346 int CharacterEncoding::GetSupportCanonicalEncodingCount() {
    347   return canonical_encoding_names_length;
    348 }
    349 
    350 // Static.
    351 std::string CharacterEncoding::GetCanonicalEncodingNameByIndex(int index) {
    352   if (index < canonical_encoding_names_length)
    353     return canonical_encoding_names[index].name;
    354   return std::string();
    355 }
    356 
    357 // Static.
    358 string16 CharacterEncoding::GetCanonicalEncodingDisplayNameByIndex(
    359     int index) {
    360   if (index < canonical_encoding_names_length)
    361     return GetEncodingDisplayName(canonical_encoding_names[index].name,
    362         canonical_encoding_names[index].category_string_id);
    363   return string16();
    364 }
    365 
    366 // Static.
    367 int CharacterEncoding::GetEncodingCommandIdByIndex(int index) {
    368   if (index < canonical_encoding_names_length)
    369     return canonical_encoding_names[index].resource_id;
    370   return 0;
    371 }
    372 
    373 // Static.
    374 std::string CharacterEncoding::GetCanonicalEncodingNameByAliasName(
    375     const std::string& alias_name) {
    376   // If the input alias_name is already canonical encoding name, just return it.
    377   const CanonicalEncodingNameToIdMapType* map =
    378       canonical_encoding_name_map_singleton.
    379           GetCanonicalEncodingNameToIdMapData();
    380   DCHECK(map);
    381 
    382   CanonicalEncodingNameToIdMapType::const_iterator found_id =
    383       map->find(alias_name);
    384   if (found_id != map->end())
    385     return alias_name;
    386 
    387   UErrorCode error_code = U_ZERO_ERROR;
    388 
    389   const char* canonical_name = ucnv_getCanonicalName(
    390       alias_name.c_str(), "MIME", &error_code);
    391   // If failed,  then try IANA next.
    392   if (U_FAILURE(error_code) || !canonical_name) {
    393     error_code = U_ZERO_ERROR;
    394     canonical_name = ucnv_getCanonicalName(
    395       alias_name.c_str(), "IANA", &error_code);
    396   }
    397 
    398   if (canonical_name) {
    399     // TODO(jnd) use a map to handle all customized {alias, canonical}
    400     // encoding mappings if we have more than one pair.
    401     // We don't want to add an unnecessary charset to the encoding menu, so we
    402     // alias 'US-ASCII' to 'ISO-8859-1' in our UI without touching WebKit.
    403     // http://crbug.com/15801.
    404     if (alias_name == "US-ASCII")
    405       return GetCanonicalEncodingNameByCommandId(IDC_ENCODING_ISO88591);
    406     return canonical_name;
    407   } else {
    408     return std::string();
    409   }
    410 }
    411 
    412 // Static
    413 // According to the behavior of user recently selected encoding short list in
    414 // FireFox, we always put UTF-8 as toppest position, after then put user
    415 // recently selected encodings, then put local dependent encoding items.
    416 // At last, we put all rest encoding items.
    417 const std::vector<CharacterEncoding::EncodingInfo>*
    418     CharacterEncoding::GetCurrentDisplayEncodings(
    419     const std::string& locale,
    420     const std::string& locale_encodings,
    421     const std::string& recently_select_encodings) {
    422   std::vector<int>* const locale_dependent_encoding_list =
    423       canonical_encoding_name_map_singleton.locale_dependent_encoding_ids();
    424   std::vector<CharacterEncoding::EncodingInfo>* const encoding_list =
    425       canonical_encoding_name_map_singleton.current_display_encodings();
    426 
    427   // Initialize locale dependent static encoding list.
    428   if (locale_dependent_encoding_list->empty() && !locale_encodings.empty())
    429     ParseEncodingListSeparatedWithComma(locale_encodings,
    430                                         locale_dependent_encoding_list,
    431                                         kUserSelectedEncodingsMaxLength);
    432 
    433   static std::string cached_user_selected_encodings;
    434   // Build current display encoding list.
    435   if (encoding_list->empty() ||
    436       cached_user_selected_encodings != recently_select_encodings) {
    437     // Update user recently selected encodings.
    438     cached_user_selected_encodings = recently_select_encodings;
    439     // Clear old encoding list since user recently selected encodings changed.
    440     encoding_list->clear();
    441     // Always add UTF-8 to first encoding position.
    442     encoding_list->push_back(EncodingInfo(IDC_ENCODING_UTF8));
    443     std::set<int> inserted_encoding;
    444     inserted_encoding.insert(IDC_ENCODING_UTF8);
    445 
    446     // Parse user recently selected encodings and get list
    447     std::vector<int> recently_select_encoding_list;
    448     ParseEncodingListSeparatedWithComma(recently_select_encodings,
    449                                         &recently_select_encoding_list,
    450                                         kUserSelectedEncodingsMaxLength);
    451 
    452     // Put 'cached encodings' (dynamic encoding list) after 'local dependent
    453     // encoding list'.
    454     recently_select_encoding_list.insert(recently_select_encoding_list.begin(),
    455         locale_dependent_encoding_list->begin(),
    456         locale_dependent_encoding_list->end());
    457     for (std::vector<int>::iterator it = recently_select_encoding_list.begin();
    458          it != recently_select_encoding_list.end(); ++it) {
    459         // Test whether we have met this encoding id.
    460         bool ok = inserted_encoding.insert(*it).second;
    461         // Duplicated encoding, ignore it. Ideally, this situation should not
    462         // happened, but just in case some one manually edit preference file.
    463         if (!ok)
    464           continue;
    465         encoding_list->push_back(EncodingInfo(*it));
    466     }
    467     // Append a separator;
    468     encoding_list->push_back(EncodingInfo(0));
    469 
    470     // We need to keep "Unicode (UTF-16LE)" always at the top (among the rest
    471     // of encodings) instead of being sorted along with other encodings. So if
    472     // "Unicode (UTF-16LE)" is already in previous encodings, sort the rest
    473     // of encodings. Otherwise Put "Unicode (UTF-16LE)" on the first of the
    474     // rest of encodings, skip "Unicode (UTF-16LE)" and sort all left encodings.
    475     int start_sorted_index = encoding_list->size();
    476     if (inserted_encoding.find(IDC_ENCODING_UTF16LE) ==
    477         inserted_encoding.end()) {
    478       encoding_list->push_back(EncodingInfo(IDC_ENCODING_UTF16LE));
    479       inserted_encoding.insert(IDC_ENCODING_UTF16LE);
    480       start_sorted_index++;
    481     }
    482 
    483     // Add the rest of encodings that are neither in the static encoding list
    484     // nor in the list of recently selected encodings.
    485     // Build the encoding list sorted in the current locale sorting order.
    486     for (int i = 0; i < default_encoding_menus_length; ++i) {
    487       int id = default_encoding_menus[i];
    488       // We have inserted this encoding, skip it.
    489       if (inserted_encoding.find(id) != inserted_encoding.end())
    490         continue;
    491       encoding_list->push_back(EncodingInfo(id));
    492     }
    493     // Sort the encoding list.
    494     l10n_util::SortVectorWithStringKey(locale,
    495                                        encoding_list,
    496                                        start_sorted_index,
    497                                        encoding_list->size(),
    498                                        true);
    499   }
    500   DCHECK(!encoding_list->empty());
    501   return encoding_list;
    502 }
    503 
    504 // Static
    505 bool CharacterEncoding::UpdateRecentlySelectedEncoding(
    506     const std::string& original_selected_encodings,
    507     int new_selected_encoding_id,
    508     std::string* selected_encodings) {
    509   // Get encoding name.
    510   std::string encoding_name =
    511       GetCanonicalEncodingNameByCommandId(new_selected_encoding_id);
    512   DCHECK(!encoding_name.empty());
    513   // Check whether the new encoding is in local dependent encodings or original
    514   // recently selected encodings. If yes, do not add it.
    515   std::vector<int>* locale_dependent_encoding_list =
    516       canonical_encoding_name_map_singleton.locale_dependent_encoding_ids();
    517   DCHECK(locale_dependent_encoding_list);
    518   std::vector<int> selected_encoding_list;
    519   ParseEncodingListSeparatedWithComma(original_selected_encodings,
    520                                       &selected_encoding_list,
    521                                       kUserSelectedEncodingsMaxLength);
    522   // Put 'cached encodings' (dynamic encoding list) after 'local dependent
    523   // encoding list' for check.
    524   std::vector<int> top_encoding_list(*locale_dependent_encoding_list);
    525   // UTF8 is always in our optimized encoding list.
    526   top_encoding_list.insert(top_encoding_list.begin(), IDC_ENCODING_UTF8);
    527   top_encoding_list.insert(top_encoding_list.end(),
    528                            selected_encoding_list.begin(),
    529                            selected_encoding_list.end());
    530   for (std::vector<int>::const_iterator it = top_encoding_list.begin();
    531        it != top_encoding_list.end(); ++it)
    532     if (*it == new_selected_encoding_id)
    533       return false;
    534   // Need to add the encoding id to recently selected encoding list.
    535   // Remove the last encoding in original list.
    536   if (selected_encoding_list.size() == kUserSelectedEncodingsMaxLength)
    537     selected_encoding_list.pop_back();
    538   // Insert new encoding to head of selected encoding list.
    539   *selected_encodings = encoding_name;
    540   // Generate the string for rest selected encoding list.
    541   for (std::vector<int>::const_iterator it = selected_encoding_list.begin();
    542        it != selected_encoding_list.end(); ++it) {
    543     selected_encodings->append(1, L',');
    544     selected_encodings->append(GetCanonicalEncodingNameByCommandId(*it));
    545   }
    546   return true;
    547 }
    548