Home | History | Annotate | Download | only in browser 
      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, "windows-949", 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-949", "Windows-949" },
     92   { "windows-1250", "Windows-1250" },
     93   { "windows-1251", "Windows-1251" },
     94   { "windows-1252", "Windows-1252" },
     95   { "windows-1253", "Windows-1253" },
     96   { "windows-1254", "Windows-1254" },
     97   { "windows-1255", "Windows-1255" },
     98   { "windows-1256", "Windows-1256" },
     99   { "windows-1257", "Windows-1257" },
    100   { "windows-1258", "Windows-1258" },
    101 };
    102 
    103 const int kCanonicalDisplayNameOverridesLength =
    104     arraysize(kCanonicalDisplayNameOverrides);
    105 
    106 typedef std::map<std::string, const char*> CanonicalNameDisplayNameMapType;
    107 
    108 class CanonicalEncodingMap {
    109  public:
    110   CanonicalEncodingMap() {}
    111   const IdToCanonicalEncodingNameMapType* GetIdToCanonicalEncodingNameMapData();
    112   const CanonicalEncodingNameToIdMapType* GetCanonicalEncodingNameToIdMapData();
    113   const CanonicalNameDisplayNameMapType* GetCanonicalNameDisplayNameMapData();
    114   std::vector<int>* locale_dependent_encoding_ids() {
    115     return &locale_dependent_encoding_ids_;
    116   }
    117 
    118   std::vector<CharacterEncoding::EncodingInfo>* current_display_encodings() {
    119     return &current_display_encodings_;
    120   }
    121 
    122  private:
    123   scoped_ptr<IdToCanonicalEncodingNameMapType> id_to_encoding_name_map_;
    124   scoped_ptr<CanonicalEncodingNameToIdMapType> encoding_name_to_id_map_;
    125   scoped_ptr<CanonicalNameDisplayNameMapType>
    126       encoding_name_to_display_name_map_;
    127   std::vector<int> locale_dependent_encoding_ids_;
    128   std::vector<CharacterEncoding::EncodingInfo> current_display_encodings_;
    129 
    130   DISALLOW_COPY_AND_ASSIGN(CanonicalEncodingMap);
    131 };
    132 
    133 const IdToCanonicalEncodingNameMapType*
    134     CanonicalEncodingMap::GetIdToCanonicalEncodingNameMapData() {
    135   // Testing and building map is not thread safe, this function is supposed to
    136   // only run in UI thread. Myabe I should add a lock in here for making it as
    137   // thread safe.
    138   if (!id_to_encoding_name_map_.get()) {
    139     id_to_encoding_name_map_.reset(new IdToCanonicalEncodingNameMapType);
    140     for (int i = 0; i < kCanonicalEncodingNamesLength; ++i) {
    141       int resource_id = kCanonicalEncodingNames[i].resource_id;
    142       (*id_to_encoding_name_map_)[resource_id] =
    143         std::make_pair(kCanonicalEncodingNames[i].name,
    144                        kCanonicalEncodingNames[i].category_string_id);
    145     }
    146   }
    147   return id_to_encoding_name_map_.get();
    148 }
    149 
    150 const CanonicalEncodingNameToIdMapType*
    151     CanonicalEncodingMap::GetCanonicalEncodingNameToIdMapData() {
    152   if (!encoding_name_to_id_map_.get()) {
    153     encoding_name_to_id_map_.reset(new CanonicalEncodingNameToIdMapType);
    154     for (int i = 0; i < kCanonicalEncodingNamesLength; ++i) {
    155       (*encoding_name_to_id_map_)[kCanonicalEncodingNames[i].name] =
    156           kCanonicalEncodingNames[i].resource_id;
    157     }
    158   }
    159   return encoding_name_to_id_map_.get();
    160 }
    161 
    162 const CanonicalNameDisplayNameMapType*
    163     CanonicalEncodingMap::GetCanonicalNameDisplayNameMapData() {
    164   if (!encoding_name_to_display_name_map_.get()) {
    165     encoding_name_to_display_name_map_.reset(
    166         new CanonicalNameDisplayNameMapType);
    167     // First store the names in the kCanonicalEncodingNames list.
    168     for (int i = 0; i < kCanonicalEncodingNamesLength; ++i) {
    169       (*encoding_name_to_display_name_map_)[kCanonicalEncodingNames[i].name] =
    170           kCanonicalEncodingNames[i].name;
    171     }
    172     // Then save in the overrides.
    173     for (int i = 0; i < kCanonicalDisplayNameOverridesLength; ++i) {
    174       (*encoding_name_to_display_name_map_)
    175           [kCanonicalDisplayNameOverrides[i].canonical_form] =
    176           kCanonicalDisplayNameOverrides[i].display_form;
    177     }
    178     DCHECK(static_cast<int>(encoding_name_to_display_name_map_->size()) ==
    179            kCanonicalEncodingNamesLength)
    180         << "Got an override that wasn't in the encoding list";
    181   }
    182   return encoding_name_to_display_name_map_.get();
    183 }
    184 
    185 // A static map object which contains all resourceid-nonsequenced canonical
    186 // encoding names.
    187 CanonicalEncodingMap* CanonicalEncodingMapSingleton() {
    188   DCHECK(BrowserThread::CurrentlyOn(BrowserThread::UI));
    189   static CanonicalEncodingMap* singleton = new CanonicalEncodingMap;
    190   return singleton;
    191 }
    192 
    193 const int kDefaultEncodingMenus[] = {
    194   IDC_ENCODING_UTF16LE,
    195   IDC_ENCODING_ISO88591,
    196   IDC_ENCODING_WINDOWS1252,
    197   IDC_ENCODING_GBK,
    198   IDC_ENCODING_GB18030,
    199   IDC_ENCODING_BIG5,
    200   IDC_ENCODING_BIG5HKSCS,
    201   IDC_ENCODING_KOREAN,
    202   IDC_ENCODING_SHIFTJIS,
    203   IDC_ENCODING_EUCJP,
    204   IDC_ENCODING_ISO2022JP,
    205   IDC_ENCODING_THAI,
    206   IDC_ENCODING_ISO885915,
    207   IDC_ENCODING_MACINTOSH,
    208   IDC_ENCODING_ISO88592,
    209   IDC_ENCODING_WINDOWS1250,
    210   IDC_ENCODING_ISO88595,
    211   IDC_ENCODING_WINDOWS1251,
    212   IDC_ENCODING_KOI8R,
    213   IDC_ENCODING_KOI8U,
    214   IDC_ENCODING_ISO88597,
    215   IDC_ENCODING_WINDOWS1253,
    216   IDC_ENCODING_WINDOWS1254,
    217   IDC_ENCODING_WINDOWS1256,
    218   IDC_ENCODING_ISO88596,
    219   IDC_ENCODING_WINDOWS1255,
    220   IDC_ENCODING_ISO88598I,
    221   IDC_ENCODING_ISO88598,
    222   IDC_ENCODING_WINDOWS1258,
    223   IDC_ENCODING_ISO88594,
    224   IDC_ENCODING_ISO885913,
    225   IDC_ENCODING_WINDOWS1257,
    226   IDC_ENCODING_ISO88593,
    227   IDC_ENCODING_ISO885910,
    228   IDC_ENCODING_ISO885914,
    229   IDC_ENCODING_ISO885916,
    230 };
    231 
    232 const int kDefaultEncodingMenusLength = arraysize(kDefaultEncodingMenus);
    233 
    234 // Parse the input |encoding_list| which is a encoding list separated with
    235 // comma, get available encoding ids and save them to |available_list|.
    236 // The parameter |maximum_size| indicates maximum size of encoding items we
    237 // want to get from the |encoding_list|.
    238 void ParseEncodingListSeparatedWithComma(
    239     const std::string& encoding_list, std::vector<int>* const available_list,
    240     size_t maximum_size) {
    241   base::StringTokenizer tokenizer(encoding_list, ",");
    242   while (tokenizer.GetNext()) {
    243     int id = CharacterEncoding::GetCommandIdByCanonicalEncodingName(
    244         tokenizer.token());
    245     // Ignore invalid encoding.
    246     if (!id)
    247       continue;
    248     available_list->push_back(id);
    249     if (available_list->size() == maximum_size)
    250       return;
    251   }
    252 }
    253 
    254 string16 GetEncodingDisplayName(const std::string& encoding_name,
    255                                 int category_string_id) {
    256   string16 category_name = l10n_util::GetStringUTF16(category_string_id);
    257   if (category_string_id != IDS_ENCODING_KOREAN &&
    258       category_string_id != IDS_ENCODING_THAI &&
    259       category_string_id != IDS_ENCODING_TURKISH) {
    260     const CanonicalNameDisplayNameMapType* map =
    261         CanonicalEncodingMapSingleton()->GetCanonicalNameDisplayNameMapData();
    262     DCHECK(map);
    263 
    264     CanonicalNameDisplayNameMapType::const_iterator found_name =
    265         map->find(encoding_name);
    266     DCHECK(found_name != map->end());
    267     return l10n_util::GetStringFUTF16(IDS_ENCODING_DISPLAY_TEMPLATE,
    268                                       category_name,
    269                                       ASCIIToUTF16(found_name->second));
    270   }
    271   return category_name;
    272 }
    273 
    274 int GetEncodingCategoryStringIdByCommandId(int id) {
    275   const IdToCanonicalEncodingNameMapType* map =
    276       CanonicalEncodingMapSingleton()->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       CanonicalEncodingMapSingleton()->GetCanonicalEncodingNameToIdMapData();
    306   DCHECK(map);
    307 
    308   CanonicalEncodingNameToIdMapType::const_iterator found_id =
    309       map->find(encoding_name);
    310   if (found_id != map->end())
    311     return found_id->second;
    312   return 0;
    313 }
    314 
    315 // Static.
    316 std::string CharacterEncoding::GetCanonicalEncodingNameByCommandId(int id) {
    317   const IdToCanonicalEncodingNameMapType* map =
    318       CanonicalEncodingMapSingleton()->GetIdToCanonicalEncodingNameMapData();
    319   DCHECK(map);
    320 
    321   IdToCanonicalEncodingNameMapType::const_iterator found_name = map->find(id);
    322   if (found_name != map->end())
    323     return found_name->second.first;
    324   return std::string();
    325 }
    326 
    327 // Static.
    328 string16 CharacterEncoding::GetCanonicalEncodingDisplayNameByCommandId(
    329     int id) {
    330   const IdToCanonicalEncodingNameMapType* map =
    331       CanonicalEncodingMapSingleton()->GetIdToCanonicalEncodingNameMapData();
    332   DCHECK(map);
    333 
    334   IdToCanonicalEncodingNameMapType::const_iterator found_name = map->find(id);
    335   if (found_name != map->end())
    336     return GetEncodingDisplayName(found_name->second.first,
    337                                   found_name->second.second);
    338   return string16();
    339 }
    340 
    341 // Static.
    342 // Return count number of all supported canonical encoding.
    343 int CharacterEncoding::GetSupportCanonicalEncodingCount() {
    344   return kCanonicalEncodingNamesLength;
    345 }
    346 
    347 // Static.
    348 std::string CharacterEncoding::GetCanonicalEncodingNameByIndex(int index) {
    349   if (index < kCanonicalEncodingNamesLength)
    350     return kCanonicalEncodingNames[index].name;
    351   return std::string();
    352 }
    353 
    354 // Static.
    355 string16 CharacterEncoding::GetCanonicalEncodingDisplayNameByIndex(
    356     int index) {
    357   if (index < kCanonicalEncodingNamesLength)
    358     return GetEncodingDisplayName(kCanonicalEncodingNames[index].name,
    359         kCanonicalEncodingNames[index].category_string_id);
    360   return string16();
    361 }
    362 
    363 // Static.
    364 int CharacterEncoding::GetEncodingCommandIdByIndex(int index) {
    365   if (index < kCanonicalEncodingNamesLength)
    366     return kCanonicalEncodingNames[index].resource_id;
    367   return 0;
    368 }
    369 
    370 // Static.
    371 std::string CharacterEncoding::GetCanonicalEncodingNameByAliasName(
    372     const std::string& alias_name) {
    373   // If the input alias_name is already canonical encoding name, just return it.
    374   const CanonicalEncodingNameToIdMapType* map =
    375       CanonicalEncodingMapSingleton()->GetCanonicalEncodingNameToIdMapData();
    376   DCHECK(map);
    377 
    378   CanonicalEncodingNameToIdMapType::const_iterator found_id =
    379       map->find(alias_name);
    380   if (found_id != map->end())
    381     return alias_name;
    382 
    383   UErrorCode error_code = U_ZERO_ERROR;
    384 
    385   const char* canonical_name = ucnv_getCanonicalName(
    386       alias_name.c_str(), "MIME", &error_code);
    387   // If failed,  then try IANA next.
    388   if (U_FAILURE(error_code) || !canonical_name) {
    389     error_code = U_ZERO_ERROR;
    390     canonical_name = ucnv_getCanonicalName(
    391       alias_name.c_str(), "IANA", &error_code);
    392   }
    393 
    394   if (canonical_name) {
    395     // TODO(jnd) use a map to handle all customized {alias, canonical}
    396     // encoding mappings if we have more than one pair.
    397     // We don't want to add an unnecessary charset to the encoding menu, so we
    398     // alias 'US-ASCII' to 'ISO-8859-1' in our UI without touching WebKit.
    399     // http://crbug.com/15801.
    400     if (alias_name == "US-ASCII")
    401       return GetCanonicalEncodingNameByCommandId(IDC_ENCODING_ISO88591);
    402     return canonical_name;
    403   } else {
    404     return std::string();
    405   }
    406 }
    407 
    408 // Static
    409 // According to the behavior of user recently selected encoding short list in
    410 // Firefox, we always put UTF-8 as top position, after then put user
    411 // recent selected encodings, then put local dependent encoding items.
    412 // At last, we put all remaining encoding items.
    413 const std::vector<CharacterEncoding::EncodingInfo>*
    414     CharacterEncoding::GetCurrentDisplayEncodings(
    415     const std::string& locale,
    416     const std::string& locale_encodings,
    417     const std::string& recently_select_encodings) {
    418   std::vector<int>* const locale_dependent_encoding_list =
    419       CanonicalEncodingMapSingleton()->locale_dependent_encoding_ids();
    420   std::vector<CharacterEncoding::EncodingInfo>* const encoding_list =
    421       CanonicalEncodingMapSingleton()->current_display_encodings();
    422 
    423   // Initialize locale dependent static encoding list.
    424   if (locale_dependent_encoding_list->empty() && !locale_encodings.empty())
    425     ParseEncodingListSeparatedWithComma(locale_encodings,
    426                                         locale_dependent_encoding_list,
    427                                         kUserSelectedEncodingsMaxLength);
    428 
    429   CR_DEFINE_STATIC_LOCAL(std::string, cached_user_selected_encodings, ());
    430   // Build current display encoding list.
    431   if (encoding_list->empty() ||
    432       cached_user_selected_encodings != recently_select_encodings) {
    433     // Update user recently selected encodings.
    434     cached_user_selected_encodings = recently_select_encodings;
    435     // Clear old encoding list since user recently selected encodings changed.
    436     encoding_list->clear();
    437     // Always add UTF-8 to first encoding position.
    438     encoding_list->push_back(EncodingInfo(IDC_ENCODING_UTF8));
    439     std::set<int> inserted_encoding;
    440     inserted_encoding.insert(IDC_ENCODING_UTF8);
    441 
    442     // Parse user recently selected encodings and get list
    443     std::vector<int> recently_select_encoding_list;
    444     ParseEncodingListSeparatedWithComma(recently_select_encodings,
    445                                         &recently_select_encoding_list,
    446                                         kUserSelectedEncodingsMaxLength);
    447 
    448     // Put 'cached encodings' (dynamic encoding list) after 'local dependent
    449     // encoding list'.
    450     recently_select_encoding_list.insert(recently_select_encoding_list.begin(),
    451         locale_dependent_encoding_list->begin(),
    452         locale_dependent_encoding_list->end());
    453     for (std::vector<int>::iterator it = recently_select_encoding_list.begin();
    454          it != recently_select_encoding_list.end(); ++it) {
    455         // Test whether we have met this encoding id.
    456         bool ok = inserted_encoding.insert(*it).second;
    457         // Duplicated encoding, ignore it. Ideally, this situation should not
    458         // happened, but just in case some one manually edit preference file.
    459         if (!ok)
    460           continue;
    461         encoding_list->push_back(EncodingInfo(*it));
    462     }
    463     // Append a separator;
    464     encoding_list->push_back(EncodingInfo(0));
    465 
    466     // We need to keep "Unicode (UTF-16LE)" always at the top (among the rest
    467     // of encodings) instead of being sorted along with other encodings. So if
    468     // "Unicode (UTF-16LE)" is already in previous encodings, sort the rest
    469     // of encodings. Otherwise Put "Unicode (UTF-16LE)" on the first of the
    470     // rest of encodings, skip "Unicode (UTF-16LE)" and sort all left encodings.
    471     int start_sorted_index = encoding_list->size();
    472     if (inserted_encoding.find(IDC_ENCODING_UTF16LE) ==
    473         inserted_encoding.end()) {
    474       encoding_list->push_back(EncodingInfo(IDC_ENCODING_UTF16LE));
    475       inserted_encoding.insert(IDC_ENCODING_UTF16LE);
    476       start_sorted_index++;
    477     }
    478 
    479     // Add the rest of encodings that are neither in the static encoding list
    480     // nor in the list of recently selected encodings.
    481     // Build the encoding list sorted in the current locale sorting order.
    482     for (int i = 0; i < kDefaultEncodingMenusLength; ++i) {
    483       int id = kDefaultEncodingMenus[i];
    484       // We have inserted this encoding, skip it.
    485       if (inserted_encoding.find(id) != inserted_encoding.end())
    486         continue;
    487       encoding_list->push_back(EncodingInfo(id));
    488     }
    489     // Sort the encoding list.
    490     l10n_util::SortVectorWithStringKey(locale,
    491                                        encoding_list,
    492                                        start_sorted_index,
    493                                        encoding_list->size(),
    494                                        true);
    495   }
    496   DCHECK(!encoding_list->empty());
    497   return encoding_list;
    498 }
    499 
    500 // Static
    501 bool CharacterEncoding::UpdateRecentlySelectedEncoding(
    502     const std::string& original_selected_encodings,
    503     int new_selected_encoding_id,
    504     std::string* selected_encodings) {
    505   // Get encoding name.
    506   std::string encoding_name =
    507       GetCanonicalEncodingNameByCommandId(new_selected_encoding_id);
    508   DCHECK(!encoding_name.empty());
    509   // Check whether the new encoding is in local dependent encodings or original
    510   // recently selected encodings. If yes, do not add it.
    511   std::vector<int>* locale_dependent_encoding_list =
    512       CanonicalEncodingMapSingleton()->locale_dependent_encoding_ids();
    513   DCHECK(locale_dependent_encoding_list);
    514   std::vector<int> selected_encoding_list;
    515   ParseEncodingListSeparatedWithComma(original_selected_encodings,
    516                                       &selected_encoding_list,
    517                                       kUserSelectedEncodingsMaxLength);
    518   // Put 'cached encodings' (dynamic encoding list) after 'local dependent
    519   // encoding list' for check.
    520   std::vector<int> top_encoding_list(*locale_dependent_encoding_list);
    521   // UTF8 is always in our optimized encoding list.
    522   top_encoding_list.insert(top_encoding_list.begin(), IDC_ENCODING_UTF8);
    523   top_encoding_list.insert(top_encoding_list.end(),
    524                            selected_encoding_list.begin(),
    525                            selected_encoding_list.end());
    526   for (std::vector<int>::const_iterator it = top_encoding_list.begin();
    527        it != top_encoding_list.end(); ++it)
    528     if (*it == new_selected_encoding_id)
    529       return false;
    530   // Need to add the encoding id to recently selected encoding list.
    531   // Remove the last encoding in original list.
    532   if (selected_encoding_list.size() == kUserSelectedEncodingsMaxLength)
    533     selected_encoding_list.pop_back();
    534   // Insert new encoding to head of selected encoding list.
    535   *selected_encodings = encoding_name;
    536   // Generate the string for rest selected encoding list.
    537   for (std::vector<int>::const_iterator it = selected_encoding_list.begin();
    538        it != selected_encoding_list.end(); ++it) {
    539     selected_encodings->append(1, L',');
    540     selected_encodings->append(GetCanonicalEncodingNameByCommandId(*it));
    541   }
    542   return true;
    543 }
    544