mirror of
https://github.com/kiwix/libkiwix.git
synced 2026-01-06 05:18:07 -05:00
04bf1be9d6
ICU package contains a special code "mul" with a self-name of "multiple languages". libkiwix now suppresses that. As a result the self-name of "mul" (like for any other unknown language code) is the code itself (i.e. "mul"). The most prominent user-visible effect of this change is that the language filter in the library page no longer contains a "Multiple languages" entry if there is a legacy ZIM file with the language set to "mul" - that entry now shows up as "Mul".
87 lines
2.4 KiB
C++
87 lines
2.4 KiB
C++
#include "tools.h"
|
||
#include "stringTools.h"
|
||
#include <mutex>
|
||
|
||
namespace kiwix
|
||
{
|
||
|
||
namespace
|
||
{
|
||
|
||
// These mappings are not provided by the ICU library, any such mappings can be manually added here
|
||
std::map<std::string, std::string> iso639_3 = {
|
||
{"ami", "Amis"},
|
||
{"atj", "atikamekw"},
|
||
{"azb", "آذربایجان دیلی"},
|
||
{"bcl", "central bikol"},
|
||
{"bgs", "tagabawa"},
|
||
{"blk", "ပအိုဝ်ႏ"},
|
||
{"bxr", "буряад хэлэн"},
|
||
{"cbk", "chavacano"},
|
||
{"cdo", "閩東語"},
|
||
{"dag", "Dagbani"},
|
||
{"diq", "dimli"},
|
||
{"dty", "डोटेली"},
|
||
{"eml", "emiliân-rumagnōl"},
|
||
{"fbs", "српскохрватски"},
|
||
{"fon", "fɔ̀ngbè"},
|
||
{"gcr", "Kriyòl gwiyannen"},
|
||
{"guw", "Gungbe"},
|
||
{"hbs", "srpskohrvatski"},
|
||
{"hyw", "հայերէն/հայերեն"},
|
||
{"ido", "ido"},
|
||
{"kbp", "kabɩyɛ"},
|
||
{"kld", "Gamilaraay"},
|
||
{"lbe", "лакку маз"},
|
||
{"lbj", "ལ་དྭགས་སྐད་"},
|
||
{"lld", "ladin"},
|
||
{"map", "Austronesian"},
|
||
{"mhr", "марий йылме"},
|
||
{"mnw", "ဘာသာမန်"},
|
||
{"myn", "mayan"},
|
||
{"nah", "nahuatl"},
|
||
{"nai", "north American Indian"},
|
||
{"nds", "plattdütsch"},
|
||
{"nrm", "bhasa narom"},
|
||
{"olo", "livvi"},
|
||
{"pih", "Pitcairn-Norfolk"},
|
||
{"pnb", "Western Panjabi"},
|
||
{"pwn", "Pinayuanan"},
|
||
{"rmr", "Caló"},
|
||
{"rmy", "romani shib"},
|
||
{"roa", "romance languages"},
|
||
{"skr", "سرائیکی"},
|
||
{"szy", "Sakizaya"},
|
||
{"tay", "Tayal"},
|
||
{"tgl", "Wikang Tagalog"},
|
||
{"twi", "Akwapem Twi"},
|
||
// ICU for Ubuntu versions <= focal (20.04) returns "" for the language code ""
|
||
// unlike the later versions - which returns "und". We map this value to "Undetermined" for a common ground.
|
||
{"", "Undetermined"},
|
||
};
|
||
|
||
std::once_flag fillLanguagesFlag;
|
||
|
||
void fillLanguagesMap()
|
||
{
|
||
for (auto icuLangPtr = icu::Locale::getISOLanguages(); *icuLangPtr != NULL; ++icuLangPtr) {
|
||
const kiwix::ICULanguageInfo lang(*icuLangPtr);
|
||
iso639_3.insert({lang.iso3Code(), lang.selfName()});
|
||
}
|
||
iso639_3.erase("mul");
|
||
}
|
||
|
||
} // unnamed namespace
|
||
|
||
std::string getLanguageSelfName(const std::string& lang)
|
||
{
|
||
std::call_once(fillLanguagesFlag, fillLanguagesMap);
|
||
const auto itr = iso639_3.find(lang);
|
||
if (itr != iso639_3.end()) {
|
||
return itr->second;
|
||
}
|
||
return lang;
|
||
};
|
||
|
||
} // namespace kiwix
|