X-Git-Url: https://git.saurik.com/wxWidgets.git/blobdiff_plain/d4ffe273ca70deb8e2cff7be9a07b0640e24bee0..940ddb190cb3f67f3b1eb85c29133e829c3ab762:/src/common/string.cpp?ds=sidebyside diff --git a/src/common/string.cpp b/src/common/string.cpp index 4d80612280..783035d960 100644 --- a/src/common/string.cpp +++ b/src/common/string.cpp @@ -56,6 +56,11 @@ IMPLEMENT_DYNAMIC_CLASS(wxString, wxObject) #endif //WXSTRING_IS_WXOBJECT +#if wxUSE_UNICODE +#undef wxUSE_EXPERIMENTAL_PRINTF +#define wxUSE_EXPERIMENTAL_PRINTF 1 +#endif + // allocating extra space for each string consumes more memory but speeds up // the concatenation operations (nLen is the current string's length) // NB: EXTRA_ALLOC must be >= 0! @@ -89,6 +94,13 @@ extern const wxChar WXDLLEXPORT *g_szNul = &g_strEmpty.dummy; // conditional compilation // ---------------------------------------------------------------------------- +#if !defined(__WXSW__) && wxUSE_UNICODE + #ifdef wxUSE_EXPERIMENTAL_PRINTF + #undef wxUSE_EXPERIMENTAL_PRINTF + #endif + #define wxUSE_EXPERIMENTAL_PRINTF 1 +#endif + // we want to find out if the current platform supports vsnprintf()-like // function: for Unix this is done with configure, for Windows we test the // compiler explicitly. @@ -113,7 +125,7 @@ extern const wxChar WXDLLEXPORT *g_szNul = &g_strEmpty.dummy; #elif defined(__GNUG__) && !defined(__UNIX__) #warning "Using sprintf() because no snprintf()-like function defined" #elif defined(__MWERKS__) - #warning "Using sprintf() because no snprintf()-like function defined" + #warning "Using sprintf() because no snprintf()-like function defined" #endif //compiler #endif // no vsnprintf @@ -126,7 +138,7 @@ extern const wxChar WXDLLEXPORT *g_szNul = &g_strEmpty.dummy; // global functions // ---------------------------------------------------------------------------- -#ifdef wxSTD_STRING_COMPATIBILITY +#if defined(wxSTD_STRING_COMPATIBILITY) && wxUSE_STD_IOSTREAM // MS Visual C++ version 5.0 provides the new STL headers as well as the old // iostream ones. @@ -256,15 +268,14 @@ wxString::wxString(const void *pStart, const void *pEnd) wxString::wxString(const char *psz, wxMBConv& conv, size_t nLength) { // first get necessary size - - size_t nLen = conv.MB2WC((wchar_t *) NULL, psz, 0); + size_t nLen = psz ? conv.MB2WC((wchar_t *) NULL, psz, 0) : 0; // nLength is number of *Unicode* characters here! - if (nLen > nLength) + if ((nLen != (size_t)-1) && (nLen > nLength)) nLen = nLength; // empty? - if ( nLen != 0 ) { + if ( (nLen != 0) && (nLen != (size_t)-1) ) { AllocBuffer(nLen); conv.MB2WC(m_pchData, psz, nLen); } @@ -275,15 +286,15 @@ wxString::wxString(const char *psz, wxMBConv& conv, size_t nLength) #else +#if wxUSE_WCHAR_T // from wide string wxString::wxString(const wchar_t *pwz) { // first get necessary size - - size_t nLen = wxWC2MB((char *) NULL, pwz, 0); + size_t nLen = pwz ? wxWC2MB((char *) NULL, pwz, 0) : 0; // empty? - if ( nLen != 0 ) { + if ( (nLen != 0) && (nLen != (size_t)-1) ) { AllocBuffer(nLen); wxWC2MB(m_pchData, pwz, nLen); } @@ -291,6 +302,7 @@ wxString::wxString(const wchar_t *pwz) Init(); } } +#endif #endif @@ -497,12 +509,14 @@ wxString& wxString::operator=(const unsigned char* psz) return *this; } +#if wxUSE_WCHAR_T wxString& wxString::operator=(const wchar_t *pwz) { wxString str(pwz); *this = str; return *this; } +#endif #endif @@ -983,14 +997,17 @@ int wxString::Printf(const wxChar *pszFormat, ...) int wxString::PrintfV(const wxChar* pszFormat, va_list argptr) { // static buffer to avoid dynamic memory allocation each time - static char s_szScratch[1024]; + char s_szScratch[1024]; // using static buffer causes internal compiler err +#if 0 #if wxUSE_THREADS // protect the static buffer static wxCriticalSection critsect; wxCriticalSectionLocker lock(critsect); #endif +#endif -#if 1 // the new implementation +#if wxUSE_EXPERIMENTAL_PRINTF +// the new implementation Reinit(); for (size_t n = 0; pszFormat[n]; n++) @@ -998,225 +1015,225 @@ int wxString::PrintfV(const wxChar* pszFormat, va_list argptr) static char s_szFlags[256] = "%"; size_t flagofs = 1; bool adj_left = FALSE, in_prec = FALSE, - prec_dot = FALSE, done = FALSE; + prec_dot = FALSE, done = FALSE; int ilen = 0; size_t min_width = 0, max_width = wxSTRING_MAXLEN; do { #define CHECK_PREC if (in_prec && !prec_dot) { s_szFlags[flagofs++] = '.'; prec_dot = TRUE; } - switch (pszFormat[++n]) { - case _T('\0'): - done = TRUE; - break; - case _T('%'): - *this += _T('%'); - done = TRUE; - break; - case _T('#'): - case _T('0'): - case _T(' '): - case _T('+'): - case _T('\''): - CHECK_PREC - s_szFlags[flagofs++] = pszFormat[n]; - break; - case _T('-'): - CHECK_PREC - adj_left = TRUE; - s_szFlags[flagofs++] = pszFormat[n]; - break; - case _T('.'): - CHECK_PREC - in_prec = TRUE; - prec_dot = FALSE; - max_width = 0; - // dot will be auto-added to s_szFlags if non-negative number follows - break; - case _T('h'): - ilen = -1; - CHECK_PREC - s_szFlags[flagofs++] = pszFormat[n]; - break; - case _T('l'): - ilen = 1; - CHECK_PREC - s_szFlags[flagofs++] = pszFormat[n]; - break; - case _T('q'): - case _T('L'): - ilen = 2; - CHECK_PREC - s_szFlags[flagofs++] = pszFormat[n]; - break; - case _T('Z'): - ilen = 3; - CHECK_PREC - s_szFlags[flagofs++] = pszFormat[n]; - break; - case _T('*'): - { - int len = va_arg(argptr, int); - if (in_prec) { - if (len<0) break; - CHECK_PREC - max_width = len; - } else { - if (len<0) { - adj_left = !adj_left; - s_szFlags[flagofs++] = '-'; - len = -len; - } - min_width = len; - } - flagofs += ::sprintf(s_szFlags+flagofs,"%d",len); - } - break; - case _T('1'): case _T('2'): case _T('3'): - case _T('4'): case _T('5'): case _T('6'): - case _T('7'): case _T('8'): case _T('9'): - { - int len = 0; - CHECK_PREC - while ((pszFormat[n]>=_T('0')) && (pszFormat[n]<=_T('9'))) { - s_szFlags[flagofs++] = pszFormat[n]; - len = len*10 + (pszFormat[n] - _T('0')); - n++; - } - if (in_prec) max_width = len; - else min_width = len; - n--; // the main loop pre-increments n again - } - break; - case _T('d'): - case _T('i'): - case _T('o'): - case _T('u'): - case _T('x'): - case _T('X'): - CHECK_PREC - s_szFlags[flagofs++] = pszFormat[n]; - s_szFlags[flagofs] = '\0'; - if (ilen == 0 ) { - int val = va_arg(argptr, int); - ::sprintf(s_szScratch, s_szFlags, val); - } - else if (ilen == -1) { - short int val = va_arg(argptr, short int); - ::sprintf(s_szScratch, s_szFlags, val); - } - else if (ilen == 1) { - long int val = va_arg(argptr, long int); - ::sprintf(s_szScratch, s_szFlags, val); - } - else if (ilen == 2) { + switch (pszFormat[++n]) { + case _T('\0'): + done = TRUE; + break; + case _T('%'): + *this += _T('%'); + done = TRUE; + break; + case _T('#'): + case _T('0'): + case _T(' '): + case _T('+'): + case _T('\''): + CHECK_PREC + s_szFlags[flagofs++] = pszFormat[n]; + break; + case _T('-'): + CHECK_PREC + adj_left = TRUE; + s_szFlags[flagofs++] = pszFormat[n]; + break; + case _T('.'): + CHECK_PREC + in_prec = TRUE; + prec_dot = FALSE; + max_width = 0; + // dot will be auto-added to s_szFlags if non-negative number follows + break; + case _T('h'): + ilen = -1; + CHECK_PREC + s_szFlags[flagofs++] = pszFormat[n]; + break; + case _T('l'): + ilen = 1; + CHECK_PREC + s_szFlags[flagofs++] = pszFormat[n]; + break; + case _T('q'): + case _T('L'): + ilen = 2; + CHECK_PREC + s_szFlags[flagofs++] = pszFormat[n]; + break; + case _T('Z'): + ilen = 3; + CHECK_PREC + s_szFlags[flagofs++] = pszFormat[n]; + break; + case _T('*'): + { + int len = va_arg(argptr, int); + if (in_prec) { + if (len<0) break; + CHECK_PREC + max_width = len; + } else { + if (len<0) { + adj_left = !adj_left; + s_szFlags[flagofs++] = '-'; + len = -len; + } + min_width = len; + } + flagofs += ::sprintf(s_szFlags+flagofs,"%d",len); + } + break; + case _T('1'): case _T('2'): case _T('3'): + case _T('4'): case _T('5'): case _T('6'): + case _T('7'): case _T('8'): case _T('9'): + { + int len = 0; + CHECK_PREC + while ((pszFormat[n]>=_T('0')) && (pszFormat[n]<=_T('9'))) { + s_szFlags[flagofs++] = pszFormat[n]; + len = len*10 + (pszFormat[n] - _T('0')); + n++; + } + if (in_prec) max_width = len; + else min_width = len; + n--; // the main loop pre-increments n again + } + break; + case _T('d'): + case _T('i'): + case _T('o'): + case _T('u'): + case _T('x'): + case _T('X'): + CHECK_PREC + s_szFlags[flagofs++] = pszFormat[n]; + s_szFlags[flagofs] = '\0'; + if (ilen == 0 ) { + int val = va_arg(argptr, int); + ::sprintf(s_szScratch, s_szFlags, val); + } + else if (ilen == -1) { + short int val = va_arg(argptr, short int); + ::sprintf(s_szScratch, s_szFlags, val); + } + else if (ilen == 1) { + long int val = va_arg(argptr, long int); + ::sprintf(s_szScratch, s_szFlags, val); + } + else if (ilen == 2) { #if SIZEOF_LONG_LONG - long long int val = va_arg(argptr, long long int); - ::sprintf(s_szScratch, s_szFlags, val); + long long int val = va_arg(argptr, long long int); + ::sprintf(s_szScratch, s_szFlags, val); #else - long int val = va_arg(argptr, long int); - ::sprintf(s_szScratch, s_szFlags, val); + long int val = va_arg(argptr, long int); + ::sprintf(s_szScratch, s_szFlags, val); #endif - } - else if (ilen == 3) { - size_t val = va_arg(argptr, size_t); - ::sprintf(s_szScratch, s_szFlags, val); - } - *this += wxString(s_szScratch); - done = TRUE; - break; - case _T('e'): - case _T('E'): - case _T('f'): - case _T('g'): - case _T('G'): - CHECK_PREC - s_szFlags[flagofs++] = pszFormat[n]; - s_szFlags[flagofs] = '\0'; - if (ilen == 2) { - long double val = va_arg(argptr, long double); - ::sprintf(s_szScratch, s_szFlags, val); - } else { - double val = va_arg(argptr, double); - ::sprintf(s_szScratch, s_szFlags, val); - } - *this += wxString(s_szScratch); - done = TRUE; - break; - case _T('p'): - { - void *val = va_arg(argptr, void *); - CHECK_PREC - s_szFlags[flagofs++] = pszFormat[n]; - s_szFlags[flagofs] = '\0'; - ::sprintf(s_szScratch, s_szFlags, val); - *this += wxString(s_szScratch); - done = TRUE; - } - break; - case _T('c'): - { - wxChar val = va_arg(argptr, int); - // we don't need to honor padding here, do we? - *this += val; - done = TRUE; - } - break; - case _T('s'): - if (ilen == -1) { - // wx extension: we'll let %hs mean non-Unicode strings - char *val = va_arg(argptr, char *); + } + else if (ilen == 3) { + size_t val = va_arg(argptr, size_t); + ::sprintf(s_szScratch, s_szFlags, val); + } + *this += wxString(s_szScratch); + done = TRUE; + break; + case _T('e'): + case _T('E'): + case _T('f'): + case _T('g'): + case _T('G'): + CHECK_PREC + s_szFlags[flagofs++] = pszFormat[n]; + s_szFlags[flagofs] = '\0'; + if (ilen == 2) { + long double val = va_arg(argptr, long double); + ::sprintf(s_szScratch, s_szFlags, val); + } else { + double val = va_arg(argptr, double); + ::sprintf(s_szScratch, s_szFlags, val); + } + *this += wxString(s_szScratch); + done = TRUE; + break; + case _T('p'): + { + void *val = va_arg(argptr, void *); + CHECK_PREC + s_szFlags[flagofs++] = pszFormat[n]; + s_szFlags[flagofs] = '\0'; + ::sprintf(s_szScratch, s_szFlags, val); + *this += wxString(s_szScratch); + done = TRUE; + } + break; + case _T('c'): + { + wxChar val = va_arg(argptr, int); + // we don't need to honor padding here, do we? + *this += val; + done = TRUE; + } + break; + case _T('s'): + if (ilen == -1) { + // wx extension: we'll let %hs mean non-Unicode strings + char *val = va_arg(argptr, char *); #if wxUSE_UNICODE - // ASCII->Unicode constructor handles max_width right - wxString s(val, wxConv_libc, max_width); + // ASCII->Unicode constructor handles max_width right + wxString s(val, wxConvLibc, max_width); #else - size_t len = wxSTRING_MAXLEN; - if (val) { - for (len = 0; val[len] && (len= 1) { - long int *val = va_arg(argptr, long int *); - *val = Len(); - } - done = TRUE; - break; - default: - if (wxIsalpha(pszFormat[n])) - // probably some flag not taken care of here yet - s_szFlags[flagofs++] = pszFormat[n]; - else { - // bad format - *this += _T('%'); // just to pass the glibc tst-printf.c - n--; - done = TRUE; - } - break; - } + if (s.Len() < min_width) + s.Pad(min_width - s.Len(), _T(' '), adj_left); + *this += s; + } else { + wxChar *val = va_arg(argptr, wxChar *); + size_t len = wxSTRING_MAXLEN; + if (val) { + for (len = 0; val[len] && (len= 1) { + long int *val = va_arg(argptr, long int *); + *val = Len(); + } + done = TRUE; + break; + default: + if (wxIsalpha(pszFormat[n])) + // probably some flag not taken care of here yet + s_szFlags[flagofs++] = pszFormat[n]; + else { + // bad format + *this += _T('%'); // just to pass the glibc tst-printf.c + n--; + done = TRUE; + } + break; + } #undef CHECK_PREC } while (!done); } else *this += pszFormat[n]; @@ -1408,7 +1425,7 @@ size_t wxString::rfind(const wxString& str, size_t nStart) const wxASSERT( str.GetStringData()->IsValid() ); wxASSERT( nStart <= Len() ); - // # could be quicker than that + // TODO could be made much quicker than that const wxChar *p = c_str() + (nStart == npos ? Len() : nStart); while ( p >= c_str() + str.Len() ) { if ( wxStrncmp(p - str.Len(), str, str.Len()) == 0 ) @@ -1423,19 +1440,131 @@ size_t wxString::rfind(const wxString& str, size_t nStart) const #if !defined(__VISUALC__) || defined(__WIN32__) size_t wxString::rfind(const wxChar* sz, size_t nStart, size_t n) const { - return rfind(wxString(sz, n == npos ? 0 : n), nStart); + return rfind(wxString(sz, n == npos ? 0 : n), nStart); } size_t wxString::rfind(wxChar ch, size_t nStart) const { - wxASSERT( nStart <= Len() ); + if ( nStart == npos ) + { + nStart = Len(); + } + else + { + wxASSERT( nStart <= Len() ); + } - const wxChar *p = wxStrrchr(c_str() + nStart, ch); + const wxChar *p = wxStrrchr(c_str(), ch); - return p == NULL ? npos : p - c_str(); + if ( p == NULL ) + return npos; + + size_t result = p - c_str(); + return ( result > nStart ) ? npos : result; } #endif // VC++ 1.5 +size_t wxString::find_first_of(const wxChar* sz, size_t nStart) const +{ + const wxChar *start = c_str() + nStart; + const wxChar *firstOf = wxStrpbrk(start, sz); + if ( firstOf ) + return firstOf - start; + else + return npos; +} + +size_t wxString::find_last_of(const wxChar* sz, size_t nStart) const +{ + if ( nStart == npos ) + { + nStart = Len(); + } + else + { + wxASSERT( nStart <= Len() ); + } + + for ( const wxChar *p = c_str() + length() - 1; p >= c_str(); p-- ) + { + if ( wxStrchr(sz, *p) ) + return p - c_str(); + } + + return npos; +} + +size_t wxString::find_first_not_of(const wxChar* sz, size_t nStart) const +{ + if ( nStart == npos ) + { + nStart = Len(); + } + else + { + wxASSERT( nStart <= Len() ); + } + + size_t nAccept = wxStrspn(c_str() + nStart, sz); + if ( nAccept >= length() - nStart ) + return npos; + else + return nAccept; +} + +size_t wxString::find_first_not_of(wxChar ch, size_t nStart) const +{ + wxASSERT( nStart <= Len() ); + + for ( const wxChar *p = c_str() + nStart; *p; p++ ) + { + if ( *p != ch ) + return p - c_str(); + } + + return npos; +} + +size_t wxString::find_last_not_of(const wxChar* sz, size_t nStart) const +{ + if ( nStart == npos ) + { + nStart = Len(); + } + else + { + wxASSERT( nStart <= Len() ); + } + + for ( const wxChar *p = c_str() + nStart - 1; p >= c_str(); p-- ) + { + if ( !wxStrchr(sz, *p) ) + return p - c_str(); + } + + return npos; +} + +size_t wxString::find_last_not_of(wxChar ch, size_t nStart) const +{ + if ( nStart == npos ) + { + nStart = Len(); + } + else + { + wxASSERT( nStart <= Len() ); + } + + for ( const wxChar *p = c_str() + nStart - 1; p >= c_str(); p-- ) + { + if ( *p != ch ) + return p - c_str(); + } + + return npos; +} + wxString wxString::substr(size_t nStart, size_t nLen) const { // npos means 'take all' @@ -1803,13 +1932,18 @@ void wxArrayString::DoSort() // MBConv // ============================================================================ -WXDLLEXPORT_DATA(wxMBConv *) wxConv_current = &wxConv_libc; +WXDLLEXPORT_DATA(wxMBConv *) wxConvCurrent = &wxConvLibc; +#if !wxUSE_WCHAR_T +WXDLLEXPORT_DATA(wxMBConv) wxConvLibc, wxConvFile; +#endif + +#if wxUSE_WCHAR_T // ---------------------------------------------------------------------------- // standard libc conversion // ---------------------------------------------------------------------------- -WXDLLEXPORT_DATA(wxMBConv) wxConv_libc; +WXDLLEXPORT_DATA(wxMBConv) wxConvLibc; size_t wxMBConv::MB2WC(wchar_t *buf, const char *psz, size_t n) const { @@ -1825,15 +1959,15 @@ size_t wxMBConv::WC2MB(char *buf, const wchar_t *psz, size_t n) const // standard file conversion // ---------------------------------------------------------------------------- -WXDLLEXPORT_DATA(wxMBConv_file) wxConv_file; +WXDLLEXPORT_DATA(wxMBConvFile) wxConvFile; // just use the libc conversion for now -size_t wxMBConv_file::MB2WC(wchar_t *buf, const char *psz, size_t n) const +size_t wxMBConvFile::MB2WC(wchar_t *buf, const char *psz, size_t n) const { return wxMB2WC(buf, psz, n); } -size_t wxMBConv_file::WC2MB(char *buf, const wchar_t *psz, size_t n) const +size_t wxMBConvFile::WC2MB(char *buf, const wchar_t *psz, size_t n) const { return wxWC2MB(buf, psz, n); } @@ -1842,12 +1976,12 @@ size_t wxMBConv_file::WC2MB(char *buf, const wchar_t *psz, size_t n) const // standard gdk conversion // ---------------------------------------------------------------------------- -#ifdef __WXGTK__ -WXDLLEXPORT_DATA(wxMBConv_gdk) wxConv_gdk; +#ifdef __WXGTK12__ +WXDLLEXPORT_DATA(wxMBConvGdk) wxConvGdk; #include -size_t wxMBConv_gdk::MB2WC(wchar_t *buf, const char *psz, size_t n) const +size_t wxMBConvGdk::MB2WC(wchar_t *buf, const char *psz, size_t n) const { if (buf) { return gdk_mbstowcs((GdkWChar *)buf, psz, n); @@ -1859,7 +1993,7 @@ size_t wxMBConv_gdk::MB2WC(wchar_t *buf, const char *psz, size_t n) const } } -size_t wxMBConv_gdk::WC2MB(char *buf, const wchar_t *psz, size_t n) const +size_t wxMBConvGdk::WC2MB(char *buf, const wchar_t *psz, size_t n) const { char *mbstr = gdk_wcstombs((GdkWChar *)psz); size_t len = mbstr ? strlen(mbstr) : 0; @@ -1870,21 +2004,35 @@ size_t wxMBConv_gdk::WC2MB(char *buf, const wchar_t *psz, size_t n) const } return len; } -#endif +#endif // GTK > 1.0 // ---------------------------------------------------------------------------- // UTF-7 // ---------------------------------------------------------------------------- -WXDLLEXPORT_DATA(wxMBConv_UTF7) wxConv_UTF7; +WXDLLEXPORT_DATA(wxMBConvUTF7) wxConvUTF7; + +#if 0 +static char utf7_setD[]="ABCDEFGHIJKLMNOPQRSTUVWXYZ" + "abcdefghijklmnopqrstuvwxyz" + "0123456789'(),-./:?"; +static char utf7_setO[]="!\"#$%&*;<=>@[]^_`{|}"; +static char utf7_setB[]="ABCDEFGHIJKLMNOPQRSTUVWXYZ" + "abcdefghijklmnopqrstuvwxyz" + "0123456789+/"; +#endif // TODO: write actual implementations of UTF-7 here -size_t wxMBConv_UTF7::MB2WC(wchar_t *buf, const char *psz, size_t n) const +size_t wxMBConvUTF7::MB2WC(wchar_t * WXUNUSED(buf), + const char * WXUNUSED(psz), + size_t WXUNUSED(n)) const { return 0; } -size_t wxMBConv_UTF7::WC2MB(char *buf, const wchar_t *psz, size_t n) const +size_t wxMBConvUTF7::WC2MB(char * WXUNUSED(buf), + const wchar_t * WXUNUSED(psz), + size_t WXUNUSED(n)) const { return 0; } @@ -1893,17 +2041,74 @@ size_t wxMBConv_UTF7::WC2MB(char *buf, const wchar_t *psz, size_t n) const // UTF-8 // ---------------------------------------------------------------------------- -WXDLLEXPORT_DATA(wxMBConv_UTF8) wxConv_UTF8; +WXDLLEXPORT_DATA(wxMBConvUTF8) wxConvUTF8; + +static unsigned long utf8_max[]={0x7f,0x7ff,0xffff,0x1fffff,0x3ffffff,0x7fffffff,0xffffffff}; -// TODO: write actual implementations of UTF-8 here -size_t wxMBConv_UTF8::MB2WC(wchar_t *buf, const char *psz, size_t n) const +size_t wxMBConvUTF8::MB2WC(wchar_t *buf, const char *psz, size_t n) const { - return wxMB2WC(buf, psz, n); + size_t len = 0; + + while (*psz && ((!buf) || (len>cnt); + while (cnt--) { + cc = *psz++; + if ((cc&0xC0)!=0x80) { + // invalid UTF-8 sequence + return (size_t)-1; + } + res=(res<<6)|(cc&0x3f); + } + if (res<=utf8_max[ocnt]) { + // illegal UTF-8 encoding + return (size_t)-1; + } + if (buf) *buf++=res; + len++; + } + } + } + if (buf && (lenutf8_max[cnt]; cnt++); + if (!cnt) { + // plain ASCII char + if (buf) *buf++=cc; + len++; + } else { + len+=cnt+1; + if (buf) { + *buf++=(-128>>cnt)|((cc>>(cnt*6))&(0x3f>>cnt)); + while (cnt--) + *buf++=0x80|((cc>>(cnt*6))&0x3f); + } + } + } + if (buf && (lennames.Add(name); - } - } - else if (cmd == _T("")) - cset->names.Add(token.GetNextToken()); - else if (cmd == _T("")) - comchar = token.GetNextToken(); - else if (cmd == _T("")) - escchar = token.GetNextToken(); - else if (cmd == _T("data = (wchar_t *)calloc(256, sizeof(wxChar)); - in_charset = TRUE; - } - else if (cmd == _T("END")) { - if (token.GetNextToken() == _T("CHARMAP")) - in_charset = FALSE; - } - else if (in_charset) { - // format: /x00 NULL (NUL) - wxString hex = token.GetNextToken(); - wxString uni = token.GetNextToken(); - // just assume that we've got the right format - int pos = ::wxHexToDec(hex.Mid(2,2)); - unsigned long uni1 = ::wxHexToDec(uni.Mid(2,2)); - unsigned long uni2 = ::wxHexToDec(uni.Mid(4,2)); - cset->data[pos] = (uni1 << 16) | uni2; - } + // wxFprintf(stderr,_T("Loaded: %s\n"),fname.c_str()); + + wxString line; + for (line = cmap.GetFirstLine(); + !cmap.Eof(); + line = cmap.GetNextLine()) { + // wxFprintf(stderr,_T("line contents: %s\n"),line.c_str()); + wxStringTokenizer token(line); + wxString cmd = token.GetNextToken(); + if (cmd == comchar) { + if (token.GetNextToken() == _T("alias")) + cset->names.Add(token.GetNextToken()); + } + else if (cmd == _T("")) + cset->names.Add(token.GetNextToken()); + else if (cmd == _T("")) + comchar = token.GetNextToken(); + else if (cmd == _T("")) + escchar = token.GetNextToken(); + else if (cmd == _T("")) { + delete cset; + cset = (wxCharacterSet *) NULL; + break; // we don't support multibyte charsets ourselves (yet) + } + else if (cmd == _T("CHARMAP")) { + cset->data = (wchar_t *)calloc(256, sizeof(wchar_t)); + in_charset = TRUE; + } + else if (cmd == _T("END")) { + if (token.GetNextToken() == _T("CHARMAP")) + in_charset = FALSE; + } + else if (in_charset) { + // format: /x00 NULL (NUL) + // /x41 LATIN CAPITAL LETTER A + wxString hex = token.GetNextToken(); + // skip whitespace (why doesn't wxStringTokenizer do this?) + while (wxIsEmpty(hex) && token.HasMoreTokens()) hex = token.GetNextToken(); + wxString uni = token.GetNextToken(); + // skip whitespace again + while (wxIsEmpty(uni) && token.HasMoreTokens()) uni = token.GetNextToken(); + if ((hex.Len() > 2) && (wxString(hex.GetChar(0)) == escchar) && (hex.GetChar(1) == _T('x')) && + (uni.Left(2) == _T("=0) { + unsigned long uni1 = ::wxHexToDec(uni.Mid(2,2)); + unsigned long uni2 = ::wxHexToDec(uni.Mid(4,2)); + cset->data[pos] = (uni1 << 16) | uni2; + // wxFprintf(stderr,_T("char %02x mapped to %04x (%c)\n"),pos,cset->data[pos],cset->data[pos]); + } + } + } + } + if (cset) { + cset->names.Shrink(); + wxCharsets.Add(cset); } - cset->names.Shrink(); - wxCharsets.Add(cset); - forget_it: - continue; } } #endif wxCharsets.Shrink(); - already_loaded = TRUE; } -static wxCharacterSet *wxFindCharacterSet(const wxString& charset) +static wxCharacterSet *wxFindCharacterSet(const wxChar *charset) { + if (!charset) return (wxCharacterSet *)NULL; + wxLoadCharacterSets(); for (size_t n=0; n4) { - if (wxString(charset,4) == _T("8859")) { - codeset << _T("8859-"); - if (*charset == _T('-')) charset++; + // first, convert the character set name to standard form + wxString codeset; + if (wxString(charset,3).CmpNoCase(_T("ISO")) == 0) { + // make sure it's represented in the standard form: ISO_8859-1 + codeset = _T("ISO_"); + charset += 3; + if ((*charset == _T('-')) || (*charset == _T('_'))) charset++; + if (wxStrlen(charset)>4) { + if (wxString(charset,4) == _T("8859")) { + codeset << _T("8859-"); + if (*charset == _T('-')) charset++; + } } } - } - codeset << charset; - codeset.MakeUpper(); - cset = wxFindCharacterSet(codeset); + codeset << charset; + codeset.MakeUpper(); + m_name = wxStrdup(codeset.c_str()); + m_deferred = TRUE; #endif + } } -wxCSConv::~wxCSConv(void) +void wxCSConv::LoadNow() { +// wxPrintf(_T("Conversion request\n")); + if (m_deferred) { + if (!m_name) { +#ifdef __UNIX__ + wxChar *lang = wxGetenv(_T("LANG")); + wxChar *dot = lang ? wxStrchr(lang, _T('.')) : (wxChar *)NULL; + if (dot) SetName(dot+1); +#endif + } + m_cset = wxFindCharacterSet(m_name); + m_deferred = FALSE; + } } size_t wxCSConv::MB2WC(wchar_t *buf, const char *psz, size_t n) const { + ((wxCSConv *)this)->LoadNow(); // discard constness if (buf) { - if (cset) { - for (size_t c=0; c<=n; c++) - buf[c] = cset->data[psz[c]]; + if (m_cset) { + for (size_t c=0; cdata[(unsigned char)(psz[c])]; } else { // latin-1 (direct) - for (size_t c=0; c<=n; c++) - buf[c] = psz[c]; + for (size_t c=0; cLoadNow(); // discard constness if (buf) { - if (cset) { - for (size_t c=0; c<=n; c++) { - size_t n; - for (n=0; (n<256) && (cset->data[n] != psz[c]); n++); - buf[c] = (n>0xff) ? '?' : n; + if (m_cset) { + for (size_t c=0; cdata[n] != psz[c]); n++); + buf[c] = (n>0xff) ? '?' : n; } } else { // latin-1 (direct) - for (size_t c=0; c<=n; c++) - buf[c] = (psz[c]>0xff) ? '?' : psz[c]; + for (size_t c=0; c0xff) ? '?' : psz[c]; } + return n; } - return n; + return wcslen(psz); } +#endif//wxUSE_WCHAR_T + +#if wxUSE_WCHAR_T +const wxWCharBuffer wxMBConv::cMB2WC(const char *psz) const + { + if (psz) { + size_t nLen = MB2WC((wchar_t *) NULL, psz, 0); + wxWCharBuffer buf(nLen); + MB2WC(WCSTRINGCAST buf, psz, nLen); + return buf; + } else return wxWCharBuffer((wchar_t *) NULL); + } + +const wxCharBuffer wxMBConv::cWC2MB(const wchar_t *psz) const + { + if (psz) { + size_t nLen = WC2MB((char *) NULL, psz, 0); + wxCharBuffer buf(nLen); + WC2MB(MBSTRINGCAST buf, psz, nLen); + return buf; + } else return wxCharBuffer((char *) NULL); + } + +#endif//wxUSE_WCHAR_T +