/////////////////////////////////////////////////////////////////////////////
-// Name: intl.cpp
+// Name: src/common/intl.cpp
// Purpose: Internationalization and localisation for wxWindows
// Author: Vadim Zeitlin
// Modified by:
#if wxUSE_INTL
// standard headers
-#include <locale.h>
-#include <ctype.h>
+#include <locale.h>
+#include <ctype.h>
+#include <stdlib.h>
+#ifdef HAVE_LANGINFO_H
+ #include <langinfo.h>
+#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 <stdlib.h>
+#if defined(__WXMAC__)
+ #include "wx/mac/private.h" // includes mac headers
+#endif
// ----------------------------------------------------------------------------
// simple types
{
// 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;
// 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
<< GetAllMsgCatalogSubdirs(wxT("/usr/share/locale"), lang)
<< GetAllMsgCatalogSubdirs(wxT("/usr/lib/locale"), lang)
<< GetAllMsgCatalogSubdirs(wxT("/usr/local/share/locale"), lang);
-
+#endif
return searchPath;
}
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="));
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
- wxFontEncoding targetEnc = wxFONTENCODING_SYSTEM;
-#ifdef __UNIX__
- wxString langFull;
- if (wxGetEnv(wxT("LC_ALL"), &langFull) ||
- wxGetEnv(wxT("LC_CTYPE"), &langFull) ||
- wxGetEnv(wxT("LANG"), &langFull))
- {
- wxString lcharset = langFull.AfterFirst(wxT('.')).BeforeFirst(wxT('@'));
- if (!lcharset.IsEmpty())
- targetEnc = wxTheFontMapper->CharsetToEncoding(lcharset, FALSE);
- }
-#endif
-
+ wxFontEncoding targetEnc = wxLocale::GetSystemEncoding();
if (targetEnc == wxFONTENCODING_SYSTEM)
{
wxFontEncodingArray a = wxEncodingConverter::GetPlatformEquivalents(enc);
for (size_t i = 0; i < m_numStrings; i++)
converter.Convert((char*)StringAtOfs(m_pTransTable, i));
-#endif // wxUSE_GUI
+#endif // wxUSE_FONTMAP
}
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
wxString name = info->Description;
wxString canonical = info->CanonicalName;
wxString locale;
- wxChar *retloc;
+ const wxChar *retloc;
// Set the locale:
#ifdef __UNIX__
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
return wxLANGUAGE_ENGLISH;
}
- if ( langFull == _T("C") )
+ if ( langFull == _T("C") || langFull == _T("POSIX") )
{
// default C locale
return wxLANGUAGE_ENGLISH;
// the language string has the following form
//
- // lang[_LANG[.encoding]]
+ // lang[_LANG][.encoding][@modifier]
//
- // where lang is the primary language, LANG is a sublang
+ // (see environ(5) in the Open Unix specification)
+ //
+ // 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
}
}
}
+#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 )
return wxLANGUAGE_UNKNOWN;
}
+// ----------------------------------------------------------------------------
+// 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()
+{
+ 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)
+ {
+#ifdef __SOLARIS__
+ // nl_langinfo() under Solaris returns 646 by default which stands for
+ // ISO-646, i.e. 7 bit ASCII and we should recognize it to avoid
+ // warnings about unrecognized encoding on each program startup
+ if ( strcmp(alang, "646") == 0 )
+ {
+ encname = _T("US-ASCII");
+ }
+#endif // __SOLARIS__
+ else
+ {
+ 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() )
+ {
+ wxFontEncoding enc = wxTheFontMapper->
+ CharsetToEncoding(encname, FALSE /* not interactive */);
+
+ // this should probably be considered as a bug in CharsetToEncoding():
+ // it shouldn't return wxFONTENCODING_DEFAULT at all - but it does it
+ // for US-ASCII charset
+ //
+ // we, OTOH, definitely shouldn't return it as it doesn't make sense at
+ // all (which encoding is it?)
+ if ( enc != wxFONTENCODING_DEFAULT )
+ {
+ return enc;
+ }
+ //else: return wxFONTENCODING_SYSTEM below
+ }
+#endif // Win32/Unix
+
+ return wxFONTENCODING_SYSTEM;
+}
+
/*static*/ void wxLocale::AddLanguage(const wxLanguageInfo& info)
{
CreateLanguagesDB();
const wxChar *szDomain) const
{
if ( wxIsEmpty(szOrigString) )
- return szDomain;
+ return _T("");
const char *pszTrans = NULL;
#if wxUSE_UNICODE
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
}
// 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)
// --- --- --- generated code ends here --- --- ---
-
-
#endif // wxUSE_INTL