X-Git-Url: https://git.saurik.com/wxWidgets.git/blobdiff_plain/ec37df5776f89de888cdef2ac1445896d283590a..ab3edacebaba6d67c081ed1853f1d331296e533a:/src/common/intl.cpp diff --git a/src/common/intl.cpp b/src/common/intl.cpp index d3cf3e156a..f41c6ba9c5 100644 --- a/src/common/intl.cpp +++ b/src/common/intl.cpp @@ -1,5 +1,5 @@ ///////////////////////////////////////////////////////////////////////////// -// Name: intl.cpp +// Name: src/common/intl.cpp // Purpose: Internationalization and localisation for wxWindows // Author: Vadim Zeitlin // Modified by: @@ -31,26 +31,35 @@ #if wxUSE_INTL // standard headers -#include -#include +#include +#include +#include +#ifdef HAVE_LANGINFO_H + #include +#endif // wxWindows -#include "wx/defs.h" -#include "wx/string.h" -#include "wx/tokenzr.h" -#include "wx/intl.h" +#ifndef WX_PRECOMP + #include "wx/string.h" + #include "wx/intl.h" + #include "wx/log.h" + #include "wx/debug.h" + #include "wx/utils.h" + #include "wx/dynarray.h" +#endif // WX_PRECOMP + #include "wx/file.h" -#include "wx/log.h" -#include "wx/debug.h" -#include "wx/utils.h" -#include "wx/dynarray.h" +#include "wx/tokenzr.h" +#include "wx/module.h" +#include "wx/fontmap.h" +#include "wx/encconv.h" + #ifdef __WIN32__ -#include "wx/msw/private.h" + #include "wx/msw/private.h" +#elif defined(__UNIX_LIKE__) + #include "wx/fontmap.h" // for CharsetToEncoding() #endif - -#include - // ---------------------------------------------------------------------------- // simple types // ---------------------------------------------------------------------------- @@ -85,9 +94,7 @@ typedef unsigned char size_t8; { // Asserting a sizeof directly causes some compilers to // issue a "using constant in a conditional expression" warning - size_t intsize = sizeof(int); - - wxASSERT_MSG( intsize == 4, + wxASSERT_MSG( wxAssertIsEqual(sizeof(int), 4), "size_t32 is incorrectly defined!" ); } } intsizechecker; @@ -334,6 +341,12 @@ static wxString GetFullSearchPath(const wxChar *lang) // then take the current directory // FIXME it should be the directory of the executable +#ifdef __WXMAC__ + wxChar cwd[512] ; + wxGetWorkingDirectory( cwd , sizeof( cwd ) ) ; + searchPath << GetAllMsgCatalogSubdirs(cwd, lang); + // generic search paths could be somewhere in the system folder preferences +#else searchPath << GetAllMsgCatalogSubdirs(wxT("."), lang); // and finally add some standard ones @@ -341,7 +354,7 @@ static wxString GetFullSearchPath(const wxChar *lang) << GetAllMsgCatalogSubdirs(wxT("/usr/share/locale"), lang) << GetAllMsgCatalogSubdirs(wxT("/usr/lib/locale"), lang) << GetAllMsgCatalogSubdirs(wxT("/usr/local/share/locale"), lang); - +#endif return searchPath; } @@ -499,29 +512,15 @@ const char *wxMsgCatalog::GetString(const char *szOrig) const return NULL; } - -#if wxUSE_GUI -#include "wx/fontmap.h" -#include "wx/encconv.h" -#endif - void wxMsgCatalog::ConvertEncoding() { -#if wxUSE_GUI - wxFontEncoding enc; - // first, find encoding header: const char *hdr = StringAtOfs(m_pOrigTable, 0); if ( hdr == NULL || hdr[0] != 0 ) { - // not supported by this catalog, does not have non-fuzzy header + // not supported by this catalog, does not have correct header return; } - /* - we support catalogs with header (msgid "") that is _not_ marked as "#, - fuzzy" (otherwise the string would not be included into compiled - catalog) - */ wxString header(StringAtOfs(m_pTransTable, 0)); wxString charset; int pos = header.Find(wxT("Content-Type: text/plain; charset=")); @@ -531,23 +530,28 @@ void wxMsgCatalog::ConvertEncoding() while (header[n] != wxT('\n')) charset << header[n++]; - enc = wxTheFontMapper->CharsetToEncoding(charset, FALSE); +#if wxUSE_FONTMAP + wxFontEncoding enc = wxTheFontMapper->CharsetToEncoding(charset, FALSE); if ( enc == wxFONTENCODING_SYSTEM ) return; // unknown encoding - wxFontEncodingArray a = wxEncodingConverter::GetPlatformEquivalents(enc); - if (a[0] == enc) - return; // no conversion needed, locale uses native encoding - - if (a.GetCount() == 0) - return; // we don't know common equiv. under this platform + wxFontEncoding targetEnc = wxLocale::GetSystemEncoding(); + if (targetEnc == wxFONTENCODING_SYSTEM) + { + wxFontEncodingArray a = wxEncodingConverter::GetPlatformEquivalents(enc); + if (a[0] == enc) + return; // no conversion needed, locale uses native encoding + if (a.GetCount() == 0) + return; // we don't know common equiv. under this platform + targetEnc = a[0]; + } wxEncodingConverter converter; + converter.Init(enc, targetEnc); - converter.Init(enc, a[0]); for (size_t i = 0; i < m_numStrings; i++) converter.Convert((char*)StringAtOfs(m_pTransTable, i)); -#endif // wxUSE_GUI +#endif // wxUSE_FONTMAP } @@ -559,11 +563,28 @@ void wxMsgCatalog::ConvertEncoding() WX_DECLARE_EXPORTED_OBJARRAY(wxLanguageInfo, wxLanguageInfoArray); WX_DEFINE_OBJARRAY(wxLanguageInfoArray); +wxLanguageInfoArray *wxLocale::ms_languagesDB = NULL; + +/*static*/ void wxLocale::CreateLanguagesDB() +{ + if (ms_languagesDB == NULL) + { + ms_languagesDB = new wxLanguageInfoArray; + InitLanguagesDB(); + } +} + +/*static*/ void wxLocale::DestroyLanguagesDB() +{ + delete ms_languagesDB; + ms_languagesDB = NULL; +} + + wxLocale::wxLocale() { m_pszOldLocale = NULL; m_pMsgCat = NULL; - m_languagesDB = NULL; m_language = wxLANGUAGE_UNKNOWN; } @@ -594,7 +615,12 @@ bool wxLocale::Init(const wxChar *szName, if ( m_strShort.IsEmpty() ) { // FIXME I don't know how these 2 letter abbreviations are formed, // this wild guess is surely wrong - m_strShort = tolower(szLocale[0]) + tolower(szLocale[1]); + if ( szLocale[0] ) + { + m_strShort += (wxChar)wxTolower(szLocale[0]); + if ( szLocale[1] ) + m_strShort += (wxChar)wxTolower(szLocale[1]); + } } // save the old locale to be able to restore it later @@ -614,11 +640,7 @@ bool wxLocale::Init(int language, int flags) wxLanguageInfo *info = NULL; int lang = language; - if (m_languagesDB == NULL) - { - m_languagesDB = new wxLanguageInfoArray; - InitLanguagesDB(); - } + CreateLanguagesDB(); if (lang == wxLANGUAGE_DEFAULT) { @@ -634,11 +656,11 @@ bool wxLocale::Init(int language, int flags) if (lang != wxLANGUAGE_DEFAULT) { - for (size_t i = 0; i < m_languagesDB->GetCount(); i++) + for (size_t i = 0; i < ms_languagesDB->GetCount(); i++) { - if (m_languagesDB->Item(i).Language == lang) + if (ms_languagesDB->Item(i).Language == lang) { - info = &m_languagesDB->Item(i); + info = &ms_languagesDB->Item(i); break; } } @@ -654,7 +676,7 @@ bool wxLocale::Init(int language, int flags) wxString name = info->Description; wxString canonical = info->CanonicalName; wxString locale; - wxChar *retloc; + const wxChar *retloc; // Set the locale: #ifdef __UNIX__ @@ -738,7 +760,8 @@ bool wxLocale::Init(int language, int flags) wxLogError(wxT("Cannot set locale to language %s."), name.c_str()); return FALSE; } - +#elif defined(__WXMAC__) + retloc = wxSetlocale(LC_ALL , wxEmptyString); #else return FALSE; #endif @@ -759,14 +782,13 @@ void wxLocale::AddCatalogLookupPathPrefix(const wxString& prefix) //else: already have it } -int wxLocale::GetSystemLanguage() const +/*static*/ int wxLocale::GetSystemLanguage() { - wxCHECK_MSG( m_languagesDB != NULL, wxLANGUAGE_UNKNOWN, - _T("Languages DB not initialized, call wxLocale::Init!") ); + CreateLanguagesDB(); // init i to avoid compiler warning size_t i = 0, - count = m_languagesDB->GetCount(); + count = ms_languagesDB->GetCount(); #if defined(__UNIX__) // first get the string identifying the language from the environment @@ -775,30 +797,43 @@ int wxLocale::GetSystemLanguage() const !wxGetEnv(wxT("LC_MESSAGES"), &langFull) && !wxGetEnv(wxT("LANG"), &langFull)) { - // no language specified - return wxLANGUAGE_UNKNOWN; + // no language specified, threat it as English + return wxLANGUAGE_ENGLISH; } - if ( langFull == _T("C") ) + if ( langFull == _T("C") || langFull == _T("POSIX") ) { // default C locale - return wxLANGUAGE_DEFAULT; + return wxLANGUAGE_ENGLISH; } // the language string has the following form // - // lang[_LANG[.encoding]] + // lang[_LANG][.encoding][@modifier] + // + // (see environ(5) in the Open Unix specification) // - // where lang is the primary language, LANG is a sublang + // where lang is the primary language, LANG is a sublang/territory, + // encoding is the charset to use and modifier "allows the user to select + // a specific instance of localization data within a single category" // // for example, the following strings are valid: // fr // fr_FR // de_DE.iso88591 + // de_DE@euro + // de_DE.iso88591@euro // for now we don't use the encoding, although we probably should (doing // translations of the msg catalogs on the fly as required) (TODO) - langFull = langFull.BeforeFirst(_T('.')); + // + // we don't use the modifiers neither but we probably should translate + // "euro" into iso885915 + size_t posEndLang = langFull.find_first_of(_T("@.")); + if ( posEndLang != wxString::npos ) + { + langFull.Truncate(posEndLang); + } // in addition to the format above, we also can have full language names // in LANG env var - for example, SuSE is known to use LANG="german" - so @@ -835,7 +870,7 @@ int wxLocale::GetSystemLanguage() const // 1. Try to find the language either as is: for ( i = 0; i < count; i++ ) { - if ( m_languagesDB->Item(i).CanonicalName == langFull ) + if ( ms_languagesDB->Item(i).CanonicalName == langFull ) { break; } @@ -846,7 +881,7 @@ int wxLocale::GetSystemLanguage() const { for ( i = 0; i < count; i++ ) { - if ( m_languagesDB->Item(i).CanonicalName == lang ) + if ( ms_languagesDB->Item(i).CanonicalName == lang ) { break; } @@ -858,7 +893,7 @@ int wxLocale::GetSystemLanguage() const { for ( i = 0; i < count; i++ ) { - if ( ExtractLang(m_languagesDB->Item(i).CanonicalName) + if ( ExtractLang(ms_languagesDB->Item(i).CanonicalName) == langFull ) { break; @@ -871,12 +906,298 @@ int wxLocale::GetSystemLanguage() const // try to find the name in verbose description for ( i = 0; i < count; i++ ) { - if (m_languagesDB->Item(i).Description.CmpNoCase(langFull) == 0) + if (ms_languagesDB->Item(i).Description.CmpNoCase(langFull) == 0) { break; } } } +#elif defined(__WXMAC__) + char* lc = NULL ; + long lang = GetScriptVariable( smSystemScript, smScriptLang) ; + switch( GetScriptManagerVariable( smRegionCode ) ) { + case verUS : + lc = "en_US" ; + break ; + case verFrance : + lc = "fr_FR" ; + break ; + case verBritain : + lc = "en_GB" ; + break ; + case verGermany : + lc = "de_DE" ; + break ; + case verItaly : + lc = "it_IT" ; + break ; + case verNetherlands : + lc = "nl_NL" ; + break ; + case verFlemish : + lc = "nl_BE" ; + break ; + case verSweden : + lc = "sv_SE" ; + break ; + case verSpain : + lc = "es_ES" ; + break ; + case verDenmark : + lc = "da_DK" ; + break ; + case verPortugal : + lc = "pt_PT" ; + break ; + case verFrCanada: + lc = "fr_CA" ; + break ; + case verNorway: + lc = "no_NO" ; + break ; + case verIsrael: + lc = "iw_IL" ; + break ; + case verJapan: + lc = "ja_JP" ; + break ; + case verAustralia: + lc = "en_AU" ; + break ; + case verArabic: + lc = "ar" ; + break ; + case verFinland: + lc = "fi_FI" ; + break ; + case verFrSwiss: + lc = "fr_CH" ; + break ; + case verGrSwiss: + lc = "de_CH" ; + break ; + case verGreece: + lc = "el_GR" ; + break ; + case verIceland: + lc = "is_IS" ; + break ; + case verMalta: + lc = "mt_MT" ; + break ; + case verCyprus: + // _CY is not part of wx, so we have to translate according to the system language + if ( lang == langGreek ) { + lc = "el_GR" ; + } + else if ( lang == langTurkish ) { + lc = "tr_TR" ; + } + break ; + case verTurkey: + lc = "tr_TR" ; + break ; + case verYugoCroatian: + lc = "hr_HR" ; + break ; + case verIndiaHindi: + lc = "hi_IN" ; + break ; + case verPakistanUrdu: + lc = "ur_PK" ; + break ; + case verTurkishModified: + lc = "tr_TR" ; + break ; + case verItalianSwiss: + lc = "it_CH" ; + break ; + case verInternational: + lc = "en" ; + break ; + case verRomania: + lc = "ro_RO" ; + break ; + case verGreecePoly: + lc = "el_GR" ; + break ; + case verLithuania: + lc = "lt_LT" ; + break ; + case verPoland: + lc = "pl_PL" ; + break ; + case verMagyar : + case verHungary: + lc = "hu_HU" ; + break ; + case verEstonia: + lc = "et_EE" ; + break ; + case verLatvia: + lc = "lv_LV" ; + break ; + case verSami: + // not known + break ; + case verFaroeIsl: + lc = "fo_FO" ; + break ; + case verIran: + lc = "fa_IR" ; + break ; + case verRussia: + lc = "ru_RU" ; + break ; + case verIreland: + lc = "ga_IE" ; + break ; + case verKorea: + lc = "ko_KR" ; + break ; + case verChina: + lc = "zh_CN" ; + break ; + case verTaiwan: + lc = "zh_TW" ; + break ; + case verThailand: + lc = "th_TH" ; + break ; + case verCzech: + lc = "cs_CZ" ; + break ; + case verSlovak: + lc = "sk_SK" ; + break ; + case verBengali: + lc = "bn" ; + break ; + case verByeloRussian: + lc = "be_BY" ; + break ; + case verUkraine: + lc = "uk_UA" ; + break ; + case verGreeceAlt: + lc = "el_GR" ; + break ; + case verSerbian: + lc = "sr_YU" ; + break ; + case verSlovenian: + lc = "sl_SI" ; + break ; + case verMacedonian: + lc = "mk_MK" ; + break ; + case verCroatia: + lc = "hr_HR" ; + break ; + case verBrazil: + lc = "pt_BR " ; + break ; + case verBulgaria: + lc = "bg_BG" ; + break ; + case verCatalonia: + lc = "ca_ES" ; + break ; + case verScottishGaelic: + lc = "gd" ; + break ; + case verManxGaelic: + lc = "gv" ; + break ; + case verBreton: + lc = "br" ; + break ; + case verNunavut: + lc = "iu_CA" ; + break ; + case verWelsh: + lc = "cy" ; + break ; + case verIrishGaelicScript: + lc = "ga_IE" ; + break ; + case verEngCanada: + lc = "en_CA" ; + break ; + case verBhutan: + lc = "dz_BT" ; + break ; + case verArmenian: + lc = "hy_AM" ; + break ; + case verGeorgian: + lc = "ka_GE" ; + break ; + case verSpLatinAmerica: + lc = "es_AR" ; + break ; + case verTonga: + lc = "to_TO" ; + break ; + case verFrenchUniversal: + lc = "fr_FR" ; + break ; + case verAustria: + lc = "de_AT" ; + break ; + case verGujarati: + lc = "gu_IN" ; + break ; + case verPunjabi: + lc = "pa" ; + break ; + case verIndiaUrdu: + lc = "ur_IN" ; + break ; + case verVietnam: + lc = "vi_VN" ; + break ; + case verFrBelgium: + lc = "fr_BE" ; + break ; + case verUzbek: + lc = "uz_UZ" ; + break ; + case verSingapore: + lc = "zh_SG" ; + break ; + case verNynorsk: + lc = "nn_NO" ; + break ; + case verAfrikaans: + lc = "af_ZA" ; + break ; + case verEsperanto: + lc = "eo" ; + break ; + case verMarathi: + lc = "mr_IN" ; + break ; + case verTibetan: + lc = "bo" ; + break ; + case verNepal: + lc = "ne_NP" ; + break ; + case verGreenland: + lc = "kl_GL" ; + break ; + default : + break ; + } + for ( i = 0; i < count; i++ ) + { + if ( ms_languagesDB->Item(i).CanonicalName == lc ) + { + break; + } + } + #elif defined(__WIN32__) LCID lcid = GetUserDefaultLCID(); if ( lcid != 0 ) @@ -886,8 +1207,8 @@ int wxLocale::GetSystemLanguage() const for ( i = 0; i < count; i++ ) { - if (m_languagesDB->Item(i).WinLang == lang && - m_languagesDB->Item(i).WinSublang == sublang) + if (ms_languagesDB->Item(i).WinLang == lang && + ms_languagesDB->Item(i).WinSublang == sublang) { break; } @@ -899,17 +1220,122 @@ int wxLocale::GetSystemLanguage() const if ( i < count ) { // we did find a matching entry, use it - return m_languagesDB->Item(i).Language; + return ms_languagesDB->Item(i).Language; } // no info about this language in the database return wxLANGUAGE_UNKNOWN; } -void wxLocale::AddLanguage(const wxLanguageInfo& info) +// ---------------------------------------------------------------------------- +// encoding stuff +// ---------------------------------------------------------------------------- + +// this is a bit strange as under Windows we get the encoding name using its +// numeric value and under Unix we do it the other way round, but this just +// reflects the way different systems provide he encoding info + +/* static */ +wxString wxLocale::GetSystemEncodingName() { - wxASSERT_MSG(m_languagesDB != NULL, "Languages DB not initialized, call wxLocale::Init!"); - m_languagesDB->Add(info); + wxString encname; + +#if defined(__WIN32__) && !defined(__WXMICROWIN__) + // FIXME: what is the error return value for GetACP()? + UINT codepage = ::GetACP(); + encname.Printf(_T("windows-%u"), codepage); +#elif defined(__UNIX_LIKE__) + +#if defined(HAVE_LANGINFO_H) && defined(CODESET) + // GNU libc provides current character set this way (this conforms + // to Unix98) + char *oldLocale = strdup(setlocale(LC_CTYPE, NULL)); + setlocale(LC_CTYPE, ""); + char *alang = nl_langinfo(CODESET); + setlocale(LC_CTYPE, oldLocale); + free(oldLocale); + if (alang) + { + encname = wxConvLibc.cMB2WX(alang); + } + else +#endif // HAVE_LANGINFO_H + { + // if we can't get at the character set directly, try to see if it's in + // the environment variables (in most cases this won't work, but I was + // out of ideas) + wxChar *lang = wxGetenv(wxT("LC_ALL")); + wxChar *dot = lang ? wxStrchr(lang, wxT('.')) : (wxChar *)NULL; + if (!dot) + { + lang = wxGetenv(wxT("LC_CTYPE")); + if ( lang ) + dot = wxStrchr(lang, wxT('.')); + } + if (!dot) + { + lang = wxGetenv(wxT("LANG")); + if ( lang ) + dot = wxStrchr(lang, wxT('.')); + } + + if ( dot ) + { + encname = dot+1; + } + } +#endif // Win32/Unix + + return encname; +} + +/* static */ +wxFontEncoding wxLocale::GetSystemEncoding() +{ +#if defined(__WIN32__) && !defined(__WXMICROWIN__) + UINT codepage = ::GetACP(); + + // wxWindows only knows about CP1250-1257, 932, 936, 949, 950 + if ( codepage >= 1250 && codepage <= 1257 ) + { + return (wxFontEncoding)(wxFONTENCODING_CP1250 + codepage - 1250); + } + + if ( codepage == 932 ) + { + return wxFONTENCODING_CP932; + } + + if ( codepage == 936 ) + { + return wxFONTENCODING_CP936; + } + + if ( codepage == 949 ) + { + return wxFONTENCODING_CP949; + } + + if ( codepage == 950 ) + { + return wxFONTENCODING_CP950; + } +#elif defined(__UNIX_LIKE__) && wxUSE_FONTMAP + wxString encname = GetSystemEncodingName(); + if ( !encname.empty() ) + { + return wxTheFontMapper-> + CharsetToEncoding(encname, FALSE /* not interactive */); + } +#endif // Win32/Unix + + return wxFONTENCODING_SYSTEM; +} + +/*static*/ void wxLocale::AddLanguage(const wxLanguageInfo& info) +{ + CreateLanguagesDB(); + ms_languagesDB->Add(info); } wxString wxLocale::GetSysName() const @@ -928,8 +1354,6 @@ wxLocale::~wxLocale() delete pTmpCat; } - delete m_languagesDB; - // restore old locale wxSetLocale(m_pOldLocale); wxSetlocale(LC_ALL, m_pszOldLocale); @@ -986,12 +1410,13 @@ const wxMB2WXbuf wxLocale::GetString(const wxChar *szOrigString, return (wxMB2WXbuf)(szOrigString); } - else - { - return wxConvertMB2WX(pszTrans); // or preferably wxCSConv(charset).cMB2WX(pszTrans) or something, - // a macro similar to wxConvertMB2WX could be written for that - } + // or preferably wxCSConv(charset).cMB2WX(pszTrans) or something, a macro + // similar to wxConvertMB2WX could be written for that + + return wxConvertMB2WX(pszTrans); + + // undo the hack from the beginning of this function #undef szOrgString } @@ -1057,15 +1482,37 @@ wxLocale *wxSetLocale(wxLocale *pLocale) return pOld; } + + +// ---------------------------------------------------------------------------- +// wxLocale module (for lazy destruction of languagesDB) +// ---------------------------------------------------------------------------- + +class wxLocaleModule: public wxModule +{ + DECLARE_DYNAMIC_CLASS(wxLocaleModule) + public: + wxLocaleModule() {} + bool OnInit() { return TRUE; } + void OnExit() { wxLocale::DestroyLanguagesDB(); } +}; + +IMPLEMENT_DYNAMIC_CLASS(wxLocaleModule, wxModule) + + + // ---------------------------------------------------------------------------- // default languages table & initialization // ---------------------------------------------------------------------------- + +// --- --- --- generated code begins here --- --- --- + // This table is generated by misc/languages/genlang.py // When making changes, please put them into misc/languages/langtabl.txt -#ifndef __WIN32__ +#if !defined(__WIN32__) || defined(__WXMICROWIN__) #define SETWINLANG(info,lang,sublang) @@ -1567,7 +2014,7 @@ void wxLocale::InitLanguagesDB() wxLanguageInfo info; wxStringTokenizer tkn; - LNG(wxLANGUAGE_ABKHAZIAN, "ab" , 0 , 0 , "Abkhazian") + LNG(wxLANGUAGE_ABKHAZIAN, "ab" , 0 , 0 , "Abkhazian") LNG(wxLANGUAGE_AFAR, "aa" , 0 , 0 , "Afar") LNG(wxLANGUAGE_AFRIKAANS, "af_ZA", LANG_AFRIKAANS , SUBLANG_DEFAULT , "Afrikaans") LNG(wxLANGUAGE_ALBANIAN, "sq_AL", LANG_ALBANIAN , SUBLANG_DEFAULT , "Albanian") @@ -1795,10 +2242,11 @@ void wxLocale::InitLanguagesDB() LNG(wxLANGUAGE_YORUBA, "yo" , 0 , 0 , "Yoruba") LNG(wxLANGUAGE_ZHUANG, "za" , 0 , 0 , "Zhuang") LNG(wxLANGUAGE_ZULU, "zu" , 0 , 0 , "Zulu") - + }; #undef LNG +// --- --- --- generated code ends here --- --- --- #endif // wxUSE_INTL