/* * Note: * The mapping from Win32 locale ID numbers to POSIX locale strings should * be the faster one. * * Windows LCIDs are defined at https://msdn.microsoft.com/en-us/library/cc233965.aspx * [MS-LCID] Windows Language Code Identifier (LCID) Reference
*/
/* ///////////////////////////////////////////////// // // Easy macros to make the LCID <--> POSIX Mapping // /////////////////////////////////////////////////
*/
/** * The standard one language/one country mapping for LCID. * The first element must be the language, and the following * elements are the language with the country. * @param hostID LCID in host format such as 0x044d * @param languageID posix ID of just the language such as 'de' * @param posixID posix ID of the language_TERRITORY such as 'de_CH'
*/ #define ILCID_POSIX_ELEMENT_ARRAY(hostID, languageID, posixID) \
constexpr ILcidPosixElement locmap_ ## languageID [] = { \
{LANGUAGE_LCID(hostID), #languageID}, /* parent locale */ \
{hostID, #posixID}, \
};
/** * Define a subtable by ID * @param id the POSIX ID, either a language or language_TERRITORY
*/ #define ILCID_POSIX_SUBTABLE(id) \
constexpr ILcidPosixElement locmap_ ## id [] =
/** * Create the map for the posixID. This macro supposes that the language string * name is the same as the global variable name, and that the first element * in the ILcidPosixElement is just the language. * @param _posixID the full POSIX ID for this entry.
*/ #define ILCID_POSIX_MAP(_posixID) \
{UPRV_LENGTHOF(locmap_ ## _posixID), locmap_ ## _posixID}
/* //////////////////////////////////////////// // // Create the table of LCID to POSIX Mapping // None of it should be dynamically created. // // Keep static locale variables inside the function so that // it can be created properly during static init. // // Note: This table should be updated periodically. Check the [MS-LCID] Windows Language Code Identifier // (LCID) Reference defined at https://msdn.microsoft.com/en-us/library/cc233965.aspx // // Microsoft is moving away from LCID in favor of locale name as of Vista. This table needs to be // maintained for support of older Windows version. // Update: Windows 7 (091130) // // Note: Microsoft assign a different LCID if a locale has a sorting variant. POSIX IDs below may contain // @collation=XXX, but no other keywords are allowed (at least for now). When uprv_convertToLCID() is // called from uloc_getLCID(), keywords other than collation are already removed. If we really need // to support other keywords in this mapping data, we must update the implementation. ////////////////////////////////////////////
*/
// TODO: For Windows ideally this table would be a list of exceptions rather than a complete list as // LocaleNameToLCID and LCIDToLocaleName provide 90% of these.
ILCID_POSIX_SUBTABLE(az) {
{0x2c, "az"},
{0x082c, "az_Cyrl_AZ"}, /* Cyrillic based */
{0x742c, "az_Cyrl"}, /* Cyrillic based */
{0x042c, "az_Latn_AZ"}, /* Latin based */
{0x782c, "az_Latn"}, /* Latin based */
{0x042c, "az_AZ"} /* Latin based */
};
ILCID_POSIX_ELEMENT_ARRAY(0x046d, ba, ba_RU)
ILCID_POSIX_ELEMENT_ARRAY(0x0423, be, be_BY)
// ICU has chosen different names for these.
ILCID_POSIX_SUBTABLE(ckb) {
{0x92, "ckb"},
{0x7c92, "ckb_Arab"},
{0x0492, "ckb_Arab_IQ"}
};
/* Declared as cs_CZ to get around compiler errors on z/OS, which defines cs as a function */
ILCID_POSIX_ELEMENT_ARRAY(0x0405, cs, cs_CZ)
ILCID_POSIX_ELEMENT_ARRAY(0x0452, cy, cy_GB)
ILCID_POSIX_ELEMENT_ARRAY(0x0406, da, da_DK)
// Windows doesn't know POSIX or BCP47 Unicode phonebook sort names
ILCID_POSIX_SUBTABLE(de) {
{0x07, "de"},
{0x0c07, "de_AT"},
{0x0807, "de_CH"},
{0x0407, "de_DE"},
{0x1407, "de_LI"},
{0x1007, "de_LU"},
{0x10407,"de_DE@collation=phonebook"}, /*This is really de_DE_PHONEBOOK on Windows*/
{0x10407,"de@collation=phonebook"} /*This is really de_DE_PHONEBOOK on Windows*/
};
ILCID_POSIX_ELEMENT_ARRAY(0x0465, dv, dv_MV)
ILCID_POSIX_ELEMENT_ARRAY(0x0408, el, el_GR)
// Windows uses an empty string for 'invariant'
ILCID_POSIX_SUBTABLE(en) {
{0x09, "en"},
{0x0c09, "en_AU"},
{0x2809, "en_BZ"},
{0x1009, "en_CA"},
{0x0809, "en_GB"},
{0x3c09, "en_HK"},
{0x3809, "en_ID"},
{0x1809, "en_IE"},
{0x4009, "en_IN"},
{0x2009, "en_JM"},
{0x4409, "en_MY"},
{0x1409, "en_NZ"},
{0x3409, "en_PH"},
{0x4809, "en_SG"},
{0x2C09, "en_TT"},
{0x0409, "en_US"},
{0x007f, "en_US_POSIX"}, /* duplicate for round-tripping */
{0x2409, "en_029"},
{0x1c09, "en_ZA"},
{0x3009, "en_ZW"},
{0x2409, "en_VI"}, /* Virgin Islands AKA Caribbean Islands (en_CB). On Windows8+ This is 0x1000 or dynamically assigned */
{0x0409, "en_AS"}, /* Alias for en_US. Leave last. On Windows8+ This is 0x1000 or dynamically assigned */
{0x0409, "en_GU"}, /* Alias for en_US. Leave last. On Windows8+ This is 0x1000 or dynamically assigned */
{0x0409, "en_MH"}, /* Alias for en_US. Leave last. On Windows8+ This is 0x1000 or dynamically assigned */
{0x0409, "en_MP"}, /* Alias for en_US. Leave last. On Windows8+ This is 0x1000 or dynamically assigned */
{0x0409, "en_UM"} /* Alias for en_US. Leave last. On Windows8+ This is 0x1000 or dynamically assigned */
};
ILCID_POSIX_SUBTABLE(en_US_POSIX) {
{0x007f, "en_US_POSIX"} /* duplicate for roundtripping */
};
/* This LCID is really four different locales.*/
ILCID_POSIX_SUBTABLE(hr) {
{0x1a, "hr"},
{0x141a, "bs_Latn_BA"}, /* Bosnian, Bosnia and Herzegovina */
{0x681a, "bs_Latn"}, /* Bosnian, Bosnia and Herzegovina */
{0x141a, "bs_BA"}, /* Bosnian, Bosnia and Herzegovina */
{0x781a, "bs"}, /* Bosnian */
{0x201a, "bs_Cyrl_BA"}, /* Bosnian, Bosnia and Herzegovina */
{0x641a, "bs_Cyrl"}, /* Bosnian, Bosnia and Herzegovina */
{0x101a, "hr_BA"}, /* Croatian in Bosnia */
{0x041a, "hr_HR"}, /* Croatian*/
{0x2c1a, "sr_Latn_ME"},
{0x241a, "sr_Latn_RS"},
{0x181a, "sr_Latn_BA"}, /* Serbo-Croatian in Bosnia */
{0x081a, "sr_Latn_CS"}, /* Serbo-Croatian*/
{0x701a, "sr_Latn"}, /* It's 0x1a or 0x081a, pick one to make the test program happy. */
{0x1c1a, "sr_Cyrl_BA"}, /* Serbo-Croatian in Bosnia */
{0x0c1a, "sr_Cyrl_CS"}, /* Serbian*/
{0x301a, "sr_Cyrl_ME"},
{0x281a, "sr_Cyrl_RS"},
{0x6c1a, "sr_Cyrl"}, /* It's 0x1a or 0x0c1a, pick one to make the test program happy. */
{0x7c1a, "sr"} /* In CLDR sr is sr_Cyrl. */
};
/* The "no" locale split into nb and nn. By default in ICU, "no" is nb.*/ // TODO: Not all of these are needed on Windows, but I don't know how ICU treats preferred ones here.
ILCID_POSIX_SUBTABLE(no) {
{0x14, "no"}, /* really nb_NO - actually Windows differentiates between neutral (no region) and specific (with region) */
{0x7c14, "nb"}, /* really nb */
{0x0414, "nb_NO"}, /* really nb_NO. Keep first in the 414 list. */
{0x0414, "no_NO"}, /* really nb_NO */
{0x0814, "nn_NO"}, /* really nn_NO. Keep first in the 814 list. */
{0x7814, "nn"}, /* It's 0x14 or 0x814, pick one to make the test program happy. */
{0x0814, "no_NO_NY"}/* really nn_NO */
};
ILCID_POSIX_SUBTABLE(quc) {
{0x93, "quc"},
{0x0493, "quc_CO"}, /* "quc_Latn_GT" is an exceptional case. Language ID of "quc" is 0x93, but LCID of "quc_Latn_GT" is 0x486, which should be under the group of "qut". "qut" is a retired ISO 639-3 language code for West Central Quiche, and merged to "quc". It looks Windows previously reserved "qut" for K'iche', but, decided to use "quc" when adding a locale for K'iche' (Guatemala).
This data structure used here assumes language ID bits in LCID is unique for alphabetic language code. But this is not true for "quc_Latn_GT". If we don't have the data below, LCID look up by alphabetic locale ID (POSIX) will fail. The same entry is found under "qut" below, which is required for reverse look up.
*/
{0x0486, "quc_Latn_GT"}
};
ILCID_POSIX_SUBTABLE(qut) {
{0x86, "qut"},
{0x0486, "qut_GT"}, /* See the note in "quc" above.
*/
{0x0486, "quc_Latn_GT"}
};
// TODO: This is almost certainly 'wrong'. 0 in Windows is a synonym for LOCALE_USER_DEFAULT. // More likely this is a similar concept to the Windows 0x7f Invariant locale "" // (Except that it's not invariant in ICU)
ILCID_POSIX_SUBTABLE(root) {
{0x00, "root"}
};
ILCID_POSIX_SUBTABLE(uz) {
{0x43, "uz"},
{0x0843, "uz_Cyrl_UZ"}, /* Cyrillic based */
{0x7843, "uz_Cyrl"}, /* Cyrillic based */
{0x0843, "uz_UZ"}, /* Cyrillic based */
{0x0443, "uz_Latn_UZ"}, /* Latin based */
{0x7c43, "uz_Latn"} /* Latin based */
};
ILCID_POSIX_SUBTABLE(ve) { /* TODO: Verify the country */
{0x33, "ve"},
{0x0433, "ve_ZA"},
{0x0433, "ven_ZA"}
};
// Windows & ICU tend to different names for some of these // TODO: Windows probably does not need all of these entries, but I don't know how the precedence works.
ILCID_POSIX_SUBTABLE(zh) {
{0x0004, "zh_Hans"},
{0x7804, "zh"},
{0x0804, "zh_CN"},
{0x0804, "zh_Hans_CN"},
{0x0c04, "zh_Hant_HK"},
{0x0c04, "zh_HK"},
{0x1404, "zh_Hant_MO"},
{0x1404, "zh_MO"},
{0x1004, "zh_Hans_SG"},
{0x1004, "zh_SG"},
{0x0404, "zh_Hant_TW"},
{0x7c04, "zh_Hant"},
{0x0404, "zh_TW"},
{0x30404,"zh_Hant_TW"}, /* Bopomofo order */
{0x30404,"zh_TW"}, /* Bopomofo order */
{0x20004,"zh@collation=stroke"},
{0x20404,"zh_Hant@collation=stroke"},
{0x20404,"zh_Hant_TW@collation=stroke"},
{0x20404,"zh_TW@collation=stroke"},
{0x20804,"zh_Hans@collation=stroke"},
{0x20804,"zh_Hans_CN@collation=stroke"},
{0x20804,"zh_CN@collation=stroke"} // TODO: Alternate collations for other LCIDs are missing, eg: 0x50804
};
ILCID_POSIX_ELEMENT_ARRAY(0x0435, zu, zu_ZA)
/* This must be static and grouped by LCID. */
constexpr ILcidPosixMap gPosixIDmap[] = {
ILCID_POSIX_MAP(af), /* af Afrikaans 0x36 */
ILCID_POSIX_MAP(am), /* am Amharic 0x5e */
ILCID_POSIX_MAP(ar), /* ar Arabic 0x01 */
ILCID_POSIX_MAP(arn), /* arn Araucanian/Mapudungun 0x7a */
ILCID_POSIX_MAP(as), /* as Assamese 0x4d */
ILCID_POSIX_MAP(az), /* az Azerbaijani 0x2c */
ILCID_POSIX_MAP(ba), /* ba Bashkir 0x6d */
ILCID_POSIX_MAP(be), /* be Belarusian 0x23 */ /* ILCID_POSIX_MAP(ber), ber Berber/Tamazight 0x5f */
ILCID_POSIX_MAP(bg), /* bg Bulgarian 0x02 */
ILCID_POSIX_MAP(bin), /* bin Edo 0x66 */
ILCID_POSIX_MAP(bn), /* bn Bengali; Bangla 0x45 */
ILCID_POSIX_MAP(bo), /* bo Tibetan 0x51 */
ILCID_POSIX_MAP(br), /* br Breton 0x7e */
ILCID_POSIX_MAP(ca), /* ca Catalan 0x03 */
ILCID_POSIX_MAP(chr), /* chr Cherokee 0x5c */
ILCID_POSIX_MAP(ckb), /* ckb Sorani (Central Kurdish) 0x92 */
ILCID_POSIX_MAP(co), /* co Corsican 0x83 */
ILCID_POSIX_MAP(cs), /* cs Czech 0x05 */
ILCID_POSIX_MAP(cy), /* cy Welsh 0x52 */
ILCID_POSIX_MAP(da), /* da Danish 0x06 */
ILCID_POSIX_MAP(de), /* de German 0x07 */
ILCID_POSIX_MAP(dv), /* dv Divehi 0x65 */
ILCID_POSIX_MAP(el), /* el Greek 0x08 */
ILCID_POSIX_MAP(en), /* en English 0x09 */
ILCID_POSIX_MAP(en_US_POSIX), /* invariant 0x7f */
ILCID_POSIX_MAP(es), /* es Spanish 0x0a */
ILCID_POSIX_MAP(et), /* et Estonian 0x25 */
ILCID_POSIX_MAP(eu), /* eu Basque 0x2d */
ILCID_POSIX_MAP(fa), /* fa Persian/Farsi 0x29 */
ILCID_POSIX_MAP(fa_AF), /* fa Persian/Dari 0x8c */
ILCID_POSIX_MAP(ff), /* ff Fula 0x67 */
ILCID_POSIX_MAP(fi), /* fi Finnish 0x0b */
ILCID_POSIX_MAP(fil), /* fil Filipino 0x64 */
ILCID_POSIX_MAP(fo), /* fo Faroese 0x38 */
ILCID_POSIX_MAP(fr), /* fr French 0x0c */
ILCID_POSIX_MAP(fuv), /* fuv Fulfulde - Nigeria 0x67 */
ILCID_POSIX_MAP(fy), /* fy Frisian 0x62 */
ILCID_POSIX_MAP(ga), /* * Gaelic (Ireland,Scotland) 0x3c */
ILCID_POSIX_MAP(gd), /* gd Gaelic (United Kingdom) 0x91 */
ILCID_POSIX_MAP(gl), /* gl Galician 0x56 */
ILCID_POSIX_MAP(gn), /* gn Guarani 0x74 */
ILCID_POSIX_MAP(gsw), /* gsw Alemanic/Alsatian/Swiss German 0x84 */
ILCID_POSIX_MAP(gu), /* gu Gujarati 0x47 */
ILCID_POSIX_MAP(ha), /* ha Hausa 0x68 */
ILCID_POSIX_MAP(haw), /* haw Hawaiian 0x75 */
ILCID_POSIX_MAP(he), /* he Hebrew (formerly iw) 0x0d */
ILCID_POSIX_MAP(hi), /* hi Hindi 0x39 */
ILCID_POSIX_MAP(hr), /* * Croatian and others 0x1a */
ILCID_POSIX_MAP(hsb), /* hsb Upper Sorbian 0x2e */
ILCID_POSIX_MAP(hu), /* hu Hungarian 0x0e */
ILCID_POSIX_MAP(hy), /* hy Armenian 0x2b */
ILCID_POSIX_MAP(ibb), /* ibb Ibibio - Nigeria 0x69 */
ILCID_POSIX_MAP(id), /* id Indonesian (formerly in) 0x21 */
ILCID_POSIX_MAP(ig), /* ig Igbo 0x70 */
ILCID_POSIX_MAP(ii), /* ii Sichuan Yi 0x78 */
ILCID_POSIX_MAP(is), /* is Icelandic 0x0f */
ILCID_POSIX_MAP(it), /* it Italian 0x10 */
ILCID_POSIX_MAP(iu), /* iu Inuktitut 0x5d */
ILCID_POSIX_MAP(iw), /* iw Hebrew 0x0d */
ILCID_POSIX_MAP(ja), /* ja Japanese 0x11 */
ILCID_POSIX_MAP(ka), /* ka Georgian 0x37 */
ILCID_POSIX_MAP(kk), /* kk Kazakh 0x3f */
ILCID_POSIX_MAP(kl), /* kl Kalaallisut 0x6f */
ILCID_POSIX_MAP(km), /* km Khmer 0x53 */
ILCID_POSIX_MAP(kn), /* kn Kannada 0x4b */
ILCID_POSIX_MAP(ko), /* ko Korean 0x12 */
ILCID_POSIX_MAP(kok), /* kok Konkani 0x57 */
ILCID_POSIX_MAP(kr), /* kr Kanuri 0x71 */
ILCID_POSIX_MAP(ks), /* ks Kashmiri 0x60 */
ILCID_POSIX_MAP(ky), /* ky Kyrgyz 0x40 */
ILCID_POSIX_MAP(lb), /* lb Luxembourgish 0x6e */
ILCID_POSIX_MAP(la), /* la Latin 0x76 */
ILCID_POSIX_MAP(lo), /* lo Lao 0x54 */
ILCID_POSIX_MAP(lt), /* lt Lithuanian 0x27 */
ILCID_POSIX_MAP(lv), /* lv Latvian, Lettish 0x26 */
ILCID_POSIX_MAP(mi), /* mi Maori 0x81 */
ILCID_POSIX_MAP(mk), /* mk Macedonian 0x2f */
ILCID_POSIX_MAP(ml), /* ml Malayalam 0x4c */
ILCID_POSIX_MAP(mn), /* mn Mongolian 0x50 */
ILCID_POSIX_MAP(mni), /* mni Manipuri 0x58 */
ILCID_POSIX_MAP(moh), /* moh Mohawk 0x7c */
ILCID_POSIX_MAP(mr), /* mr Marathi 0x4e */
ILCID_POSIX_MAP(ms), /* ms Malay 0x3e */
ILCID_POSIX_MAP(mt), /* mt Maltese 0x3a */
ILCID_POSIX_MAP(my), /* my Burmese 0x55 */ /* ILCID_POSIX_MAP(nb), // no Norwegian 0x14 */
ILCID_POSIX_MAP(ne), /* ne Nepali 0x61 */
ILCID_POSIX_MAP(nl), /* nl Dutch 0x13 */ /* ILCID_POSIX_MAP(nn), // no Norwegian 0x14 */
ILCID_POSIX_MAP(no), /* * Norwegian 0x14 */
ILCID_POSIX_MAP(nso), /* nso Sotho, Northern (Sepedi dialect) 0x6c */
ILCID_POSIX_MAP(oc), /* oc Occitan 0x82 */
ILCID_POSIX_MAP(om), /* om Oromo 0x72 */
ILCID_POSIX_MAP(or_IN), /* or Oriya 0x48 */
ILCID_POSIX_MAP(pa), /* pa Punjabi 0x46 */
ILCID_POSIX_MAP(pap), /* pap Papiamentu 0x79 */
ILCID_POSIX_MAP(pl), /* pl Polish 0x15 */
ILCID_POSIX_MAP(ps), /* ps Pashto 0x63 */
ILCID_POSIX_MAP(pt), /* pt Portuguese 0x16 */
ILCID_POSIX_MAP(qu), /* qu Quechua 0x6B */
ILCID_POSIX_MAP(quc), /* quc K'iche 0x93 */
ILCID_POSIX_MAP(qut), /* qut K'iche 0x86 */
ILCID_POSIX_MAP(rm), /* rm Raeto-Romance/Romansh 0x17 */
ILCID_POSIX_MAP(ro), /* ro Romanian 0x18 */
ILCID_POSIX_MAP(root), /* root 0x00 */
ILCID_POSIX_MAP(ru), /* ru Russian 0x19 */
ILCID_POSIX_MAP(rw), /* rw Kinyarwanda 0x87 */
ILCID_POSIX_MAP(sa), /* sa Sanskrit 0x4f */
ILCID_POSIX_MAP(sah), /* sah Yakut 0x85 */
ILCID_POSIX_MAP(sd), /* sd Sindhi 0x59 */
ILCID_POSIX_MAP(se), /* se Sami 0x3b */ /* ILCID_POSIX_MAP(sh), // sh Serbo-Croatian 0x1a */
ILCID_POSIX_MAP(si), /* si Sinhalese 0x5b */
ILCID_POSIX_MAP(sk), /* sk Slovak 0x1b */
ILCID_POSIX_MAP(sl), /* sl Slovenian 0x24 */
ILCID_POSIX_MAP(so), /* so Somali 0x77 */
ILCID_POSIX_MAP(sq), /* sq Albanian 0x1c */ /* ILCID_POSIX_MAP(sr), // sr Serbian 0x1a */
ILCID_POSIX_MAP(st), /* st Sutu 0x30 */
ILCID_POSIX_MAP(sv), /* sv Swedish 0x1d */
ILCID_POSIX_MAP(sw), /* sw Swahili 0x41 */
ILCID_POSIX_MAP(syr), /* syr Syriac 0x5A */
ILCID_POSIX_MAP(ta), /* ta Tamil 0x49 */
ILCID_POSIX_MAP(te), /* te Telugu 0x4a */
ILCID_POSIX_MAP(tg), /* tg Tajik 0x28 */
ILCID_POSIX_MAP(th), /* th Thai 0x1e */
ILCID_POSIX_MAP(ti), /* ti Tigrigna 0x73 */
ILCID_POSIX_MAP(tk), /* tk Turkmen 0x42 */
ILCID_POSIX_MAP(tn), /* tn Tswana 0x32 */
ILCID_POSIX_MAP(tr), /* tr Turkish 0x1f */
ILCID_POSIX_MAP(ts), /* ts Tsonga 0x31 */
ILCID_POSIX_MAP(tt), /* tt Tatar 0x44 */
ILCID_POSIX_MAP(tzm), /* tzm Tamazight 0x5f */
ILCID_POSIX_MAP(ug), /* ug Uighur 0x80 */
ILCID_POSIX_MAP(uk), /* uk Ukrainian 0x22 */
ILCID_POSIX_MAP(ur), /* ur Urdu 0x20 */
ILCID_POSIX_MAP(uz), /* uz Uzbek 0x43 */
ILCID_POSIX_MAP(ve), /* ve Venda 0x33 */
ILCID_POSIX_MAP(vi), /* vi Vietnamese 0x2a */
ILCID_POSIX_MAP(wo), /* wo Wolof 0x88 */
ILCID_POSIX_MAP(xh), /* xh Xhosa 0x34 */
ILCID_POSIX_MAP(yi), /* yi Yiddish 0x3d */
ILCID_POSIX_MAP(yo), /* yo Yoruba 0x6a */
ILCID_POSIX_MAP(zh), /* zh Chinese 0x04 */
ILCID_POSIX_MAP(zu), /* zu Zulu 0x35 */
};
/** * Do not call this function. It is called by hostID. * The function is not private because this struct must stay as a C struct, * and this is an internal class.
*/
int32_t
idCmp(constchar* id1, constchar* id2)
{
int32_t diffIdx = 0; while (*id1 == *id2 && *id1 != 0) {
diffIdx++;
id1++;
id2++;
} return diffIdx;
}
/** * Searches for a Windows LCID * * @param posixID the Posix style locale id. * @param status gets set to U_ILLEGAL_ARGUMENT_ERROR when the Posix ID has * no equivalent Windows LCID. * @return the LCID
*/
uint32_t
getHostID(const ILcidPosixMap *this_0, constchar* posixID, UErrorCode& status)
{ if (U_FAILURE(status)) { return locmap_root->hostID; }
int32_t bestIdx = 0;
int32_t bestIdxDiff = 0;
int32_t posixIDlen = static_cast<int32_t>(uprv_strlen(posixID));
uint32_t idx;
for (idx = 0; idx < this_0->numRegions; idx++ ) {
int32_t sameChars = idCmp(posixID, this_0->regionMaps[idx].posixID); if (sameChars > bestIdxDiff && this_0->regionMaps[idx].posixID[sameChars] == 0) { if (posixIDlen == sameChars) { /* Exact match */ return this_0->regionMaps[idx].hostID;
}
bestIdxDiff = sameChars;
bestIdx = idx;
}
} /* We asked for something unusual, like en_ZZ, and we try to return the number for the same language. */ /* We also have to make sure that sid and si and similar string subsets don't match. */ if ((posixID[bestIdxDiff] == '_' || posixID[bestIdxDiff] == '@')
&& this_0->regionMaps[bestIdx].posixID[bestIdxDiff] == 0)
{
status = U_USING_FALLBACK_WARNING; return this_0->regionMaps[bestIdx].hostID;
}
/*no match found */
status = U_ILLEGAL_ARGUMENT_ERROR; return locmap_root->hostID;
}
constchar*
getPosixID(const ILcidPosixMap *this_0, uint32_t hostID)
{
uint32_t i; for (i = 0; i < this_0->numRegions; i++)
{ if (this_0->regionMaps[i].hostID == hostID)
{ return this_0->regionMaps[i].posixID;
}
}
/* If you get here, then no matching region was found,
so return the language id with the wild card region. */ return this_0->regionMaps[0].posixID;
}
#if U_PLATFORM_HAS_WIN32_API && UCONFIG_USE_WINDOWS_LCID_MAPPING_API
static_assert(ULOC_FULLNAME_CAPACITY > LOCALE_NAME_MAX_LENGTH, "Windows locale names have smaller length than ICU locale names.");
char locName[LOCALE_NAME_MAX_LENGTH] = {};
// Note: Windows primary lang ID 0x92 in LCID is used for Central Kurdish and // GetLocaleInfo() maps such LCID to "ku". However, CLDR uses "ku" for // Northern Kurdish and "ckb" for Central Kurdish. For this reason, we cannot // use the Windows API to resolve locale ID for this specific case. if ((hostid & 0x3FF) != 0x92) {
int32_t tmpLen = 0;
char16_t windowsLocaleName[LOCALE_NAME_MAX_LENGTH] = {};
// Note: LOCALE_ALLOW_NEUTRAL_NAMES was enabled in Windows7+, prior versions did not handle neutral (no-region) locale names.
tmpLen = LCIDToLocaleName(hostid, (PWSTR)windowsLocaleName, UPRV_LENGTHOF(windowsLocaleName), LOCALE_ALLOW_NEUTRAL_NAMES); if (tmpLen > 1) {
int32_t i = 0; // Only need to look up in table if have _, eg for de-de_phoneb type alternate sort.
bLookup = false; for (i = 0; i < UPRV_LENGTHOF(locName); i++)
{
locName[i] = (char)(windowsLocaleName[i]);
// Windows locale name may contain sorting variant, such as "es-ES_tradnl". // In such cases, we need special mapping data found in the hardcoded table // in this source file. if (windowsLocaleName[i] == L'_')
{ // Keep the base locale, without variant // TODO: Should these be mapped from _phoneb to @collation=phonebook, etc.?
locName[i] = '\0';
tmpLen = i;
bLookup = true; break;
} elseif (windowsLocaleName[i] == L'-')
{ // Windows names use -, ICU uses _
locName[i] = '_';
} elseif (windowsLocaleName[i] == L'\0')
{ // No point in doing more work than necessary break;
}
} // TODO: Need to understand this better, why isn't it an alias?
FIX_LANGUAGE_ID_TAG(locName, tmpLen);
pPosixID = locName;
}
} #endif
if (bLookup) { constchar *pCandidate = nullptr;
langID = LANGUAGE_LCID(hostid);
/* On Windows, when locale name has a variant, we still look up the hardcoded table. If a match in the hardcoded table is longer than the Windows locale name without
variant, we use the one as the result */ if (pCandidate && (pPosixID == nullptr || uprv_strlen(pCandidate) > uprv_strlen(pPosixID))) {
pPosixID = pCandidate;
}
}
/* no match found */
*status = U_ILLEGAL_ARGUMENT_ERROR; return 0;
}
/* ////////////////////////////////////// // // POSIX --> LCID // This should only be called from uloc_getLCID. // The locale ID must be in canonical form. // /////////////////////////////////////
*/
U_CAPI uint32_t
uprv_convertToLCIDPlatform(constchar* localeID, UErrorCode* status)
{ if (U_FAILURE(*status)) { return 0;
}
// The purpose of this function is to leverage the Windows platform name->lcid // conversion functionality when available. #if U_PLATFORM_HAS_WIN32_API && UCONFIG_USE_WINDOWS_LCID_MAPPING_API
int32_t len;
icu::CharString baseName; constchar * mylocaleID = localeID;
// Check any for keywords. if (uprv_strchr(localeID, '@'))
{
icu::CharString collVal = ulocimp_getKeywordValue(localeID, "collation", *status); if (U_SUCCESS(*status) && !collVal.isEmpty())
{ // If it contains the keyword collation, return 0 so that the LCID lookup table will be used. return 0;
} else
{ // If the locale ID contains keywords other than collation, just use the base name.
baseName = ulocimp_getBaseName(localeID, *status); if (U_SUCCESS(*status) && !baseName.isEmpty())
{
mylocaleID = baseName.data();
}
}
}
// this will change it from de_DE@collation=phonebook to de-DE-u-co-phonebk form
icu::CharString asciiBCP47Tag = ulocimp_toLanguageTag(mylocaleID, false, *status);
if (U_SUCCESS(*status))
{ // Need it to be UTF-16, not 8-bit wchar_t bcp47Tag[LOCALE_NAME_MAX_LENGTH] = {};
int32_t i; for (i = 0; i < UPRV_LENGTHOF(bcp47Tag); i++)
{ if (asciiBCP47Tag[i] == '\0')
{ break;
} else
{ // Copy the character
bcp47Tag[i] = static_cast<wchar_t>(asciiBCP47Tag[i]);
}
}
if (i < (UPRV_LENGTHOF(bcp47Tag) - 1))
{ // Ensure it's null terminated
bcp47Tag[i] = L'\0';
LCID lcid = LocaleNameToLCID(bcp47Tag, LOCALE_ALLOW_NEUTRAL_NAMES); if (lcid > 0)
{ // Found LCID from windows, return that one, unless its completely ambiguous // LOCALE_USER_DEFAULT and transients are OK because they will round trip // for this process. if (lcid != LOCALE_CUSTOM_UNSPECIFIED)
{ return lcid;
}
}
}
} #else
(void) localeID; // Suppress unused variable warning. #endif
// This function does the table lookup when native platform name->lcid conversion isn't available, // or for locales that don't follow patterns the platform expects.
uint32_t low = 0;
uint32_t high = gLocaleCount;
uint32_t mid;
uint32_t oldmid = 0;
int32_t compVal;
/* * Sometimes we can't do a binary search on posixID because some LCIDs * go to different locales. We hit one of those special cases.
*/ for (idx = 0; idx < gLocaleCount; idx++ ) {
myStatus = U_ZERO_ERROR;
value = getHostID(&gPosixIDmap[idx], posixID, myStatus); if (myStatus == U_ZERO_ERROR) { return value;
} elseif (myStatus == U_USING_FALLBACK_WARNING) {
fallbackValue = value;
}
}
Die Informationen auf dieser Webseite wurden
nach bestem Wissen sorgfältig zusammengestellt. Es wird jedoch weder Vollständigkeit, noch Richtigkeit,
noch Qualität der bereit gestellten Informationen zugesichert.
Bemerkung:
Die farbliche Syntaxdarstellung und die Messung sind noch experimentell.