X-Git-Url: https://git.saurik.com/wxWidgets.git/blobdiff_plain/acf0c9acfcbfb745a564e084d26e0405ce1bada7..9d5507f7a2701395e1d5c121bd877bb9066ee6ea:/include/wx/string.h?ds=sidebyside diff --git a/include/wx/string.h b/include/wx/string.h index 204e4be293..2d79486f83 100644 --- a/include/wx/string.h +++ b/include/wx/string.h @@ -1,6 +1,6 @@ /////////////////////////////////////////////////////////////////////////////// // Name: wx/string.h -// Purpose: wxString and wxArrayString classes +// Purpose: wxString class // Author: Vadim Zeitlin // Modified by: // Created: 29/01/98 @@ -15,8 +15,8 @@ missing from string.h. */ -#ifndef _WX_WXSTRINGH__ -#define _WX_WXSTRINGH__ +#ifndef _WX_WXSTRING_H__ +#define _WX_WXSTRING_H__ // ---------------------------------------------------------------------------- // headers @@ -51,12 +51,21 @@ #include #endif -#include "wx/wxchar.h" // for wxChar +#include "wx/wxcrtbase.h" // for wxChar, wxStrlen() etc. #include "wx/strvararg.h" #include "wx/buffer.h" // for wxCharBuffer #include "wx/strconv.h" // for wxConvertXXX() macros and wxMBConv classes +#include "wx/stringimpl.h" +#include "wx/stringops.h" +#include "wx/unichar.h" -class WXDLLIMPEXP_BASE wxString; +class WXDLLIMPEXP_FWD_BASE wxString; + +// unless this symbol is predefined to disable the compatibility functions, do +// use them +#ifndef WXWIN_COMPATIBILITY_STRING_PTR_AS_ITER + #define WXWIN_COMPATIBILITY_STRING_PTR_AS_ITER 1 +#endif // --------------------------------------------------------------------------- // macros @@ -69,9 +78,13 @@ class WXDLLIMPEXP_BASE wxString; #define wxMBSTRINGCAST (char *)(const char *) #define wxWCSTRINGCAST (wchar_t *)(const wchar_t *) -// implementation only -#define wxASSERT_VALID_INDEX(i) \ - wxASSERT_MSG( (size_t)(i) <= length(), _T("invalid index in wxString") ) +// like _T(), but for literals in wxString's internal representation, i.e. +// char* in UTF-8 build and wxChar* otherwise: +#if wxUSE_UNICODE_UTF8 + #define wxSTRING_TEXT(str) str +#else + #define wxSTRING_TEXT(str) _T(str) +#endif // ---------------------------------------------------------------------------- // constants @@ -82,33 +95,30 @@ class WXDLLIMPEXP_BASE wxString; // deprecated in favour of wxString::npos, don't use in new code // // maximum possible length for a string means "take all string" everywhere -#define wxSTRING_MAXLEN wxStringBase::npos +#define wxSTRING_MAXLEN wxString::npos #endif // WXWIN_COMPATIBILITY_2_6 -// ---------------------------------------------------------------------------- -// global data -// ---------------------------------------------------------------------------- - -// global pointer to empty string -extern WXDLLIMPEXP_DATA_BASE(const wxChar*) wxEmptyString; - // --------------------------------------------------------------------------- // global functions complementing standard C string library replacements for // strlen() and portable strcasecmp() //--------------------------------------------------------------------------- -// Use wxXXX() functions from wxchar.h instead! These functions are for +#if WXWIN_COMPATIBILITY_2_8 +// Use wxXXX() functions from wxcrt.h instead! These functions are for // backwards compatibility only. // checks whether the passed in pointer is NULL and if the string is empty +wxDEPRECATED( inline bool IsEmpty(const char *p) ); inline bool IsEmpty(const char *p) { return (!p || !*p); } // safe version of strlen() (returns 0 if passed NULL pointer) +wxDEPRECATED( inline size_t Strlen(const char *psz) ); inline size_t Strlen(const char *psz) { return psz ? strlen(psz) : 0; } // portable strcasecmp/_stricmp +wxDEPRECATED( inline int Stricmp(const char *psz1, const char *psz2) ); inline int Stricmp(const char *psz1, const char *psz2) { #if defined(__VISUALC__) && defined(__WXWINCE__) @@ -166,131 +176,392 @@ inline int Stricmp(const char *psz1, const char *psz2) #endif // OS/compiler } +#endif // WXWIN_COMPATIBILITY_2_8 + // ---------------------------------------------------------------------------- -// deal with STL/non-STL/non-STL-but-wxUSE_STD_STRING +// wxCStrData // ---------------------------------------------------------------------------- -// in both cases we need to define wxStdString -#if wxUSE_STL || wxUSE_STD_STRING +// Lightweight object returned by wxString::c_str() and implicitly convertible +// to either const char* or const wchar_t*. +class WXDLLIMPEXP_BASE wxCStrData +{ +private: + // Ctors; for internal use by wxString and wxCStrData only + wxCStrData(const wxString *str, size_t offset = 0, bool owned = false) + : m_str(str), m_offset(offset), m_owned(owned) {} + +public: + // Ctor constructs the object from char literal; they are needed to make + // operator?: compile and they intentionally take char*, not const char* + inline wxCStrData(char *buf); + inline wxCStrData(wchar_t *buf); + inline wxCStrData(const wxCStrData& data); -#include "wx/beforestd.h" -#include -#include "wx/afterstd.h" + inline ~wxCStrData(); -#if wxUSE_UNICODE - #ifdef HAVE_STD_WSTRING - typedef std::wstring wxStdString; - #else - typedef std::basic_string wxStdString; - #endif -#else - typedef std::string wxStdString; + // methods defined inline below must be declared inline or mingw32 3.4.5 + // warns about " defined locally after being referenced with + // dllimport linkage" +#if wxUSE_UNICODE_WCHAR + inline #endif + const wchar_t* AsWChar() const; + operator const wchar_t*() const { return AsWChar(); } -#endif // need +#if !wxUSE_UNICODE || wxUSE_UTF8_LOCALE_ONLY + inline +#endif + const char* AsChar() const; + const unsigned char* AsUnsignedChar() const + { return (const unsigned char *) AsChar(); } + operator const char*() const { return AsChar(); } + operator const unsigned char*() const { return AsUnsignedChar(); } -#if wxUSE_STL + operator const void*() const { return AsChar(); } - // we don't need an extra ctor from std::string when copy ctor already does - // the work - #undef wxUSE_STD_STRING - #define wxUSE_STD_STRING 0 + inline const wxCharBuffer AsCharBuf() const; + inline const wxWCharBuffer AsWCharBuf() const; - #if (defined(__GNUG__) && (__GNUG__ < 3)) || \ - (defined(_MSC_VER) && (_MSC_VER <= 1200)) - #define wxSTRING_BASE_HASNT_CLEAR - #endif + inline wxString AsString() const; - typedef wxStdString wxStringBase; -#else // if !wxUSE_STL + // returns the value as C string in internal representation (equivalent + // to AsString().wx_str(), but more efficient) + const wxStringCharType *AsInternal() const; -#if !defined(HAVE_STD_STRING_COMPARE) && \ - (!defined(__WX_SETUP_H__) || wxUSE_STL == 0) - #define HAVE_STD_STRING_COMPARE -#endif + // allow expressions like "c_str()[0]": + inline wxUniChar operator[](size_t n) const; + wxUniChar operator[](int n) const { return operator[](size_t(n)); } + wxUniChar operator[](long n) const { return operator[](size_t(n)); } +#ifndef wxSIZE_T_IS_UINT + wxUniChar operator[](unsigned int n) const { return operator[](size_t(n)); } +#endif // size_t != unsigned int + // these operators are needed to emulate the pointer semantics of c_str(): + // expressions like "wxChar *p = str.c_str() + 1;" should continue to work + // (we need both versions to resolve ambiguities): + wxCStrData operator+(int n) const + { return wxCStrData(m_str, m_offset + n, m_owned); } + wxCStrData operator+(long n) const + { return wxCStrData(m_str, m_offset + n, m_owned); } + wxCStrData operator+(size_t n) const + { return wxCStrData(m_str, m_offset + n, m_owned); } + + // and these for "str.c_str() + n - 2": + wxCStrData operator-(int n) const + { + wxASSERT_MSG( n <= (int)m_offset, + _T("attempt to construct address before the beginning of the string") ); + return wxCStrData(m_str, m_offset - n, m_owned); + } + wxCStrData operator-(long n) const + { + wxASSERT_MSG( n <= (int)m_offset, + _T("attempt to construct address before the beginning of the string") ); + return wxCStrData(m_str, m_offset - n, m_owned); + } + wxCStrData operator-(size_t n) const + { + wxASSERT_MSG( n <= m_offset, + _T("attempt to construct address before the beginning of the string") ); + return wxCStrData(m_str, m_offset - n, m_owned); + } + + // this operator is needed to make expressions like "*c_str()" or + // "*(c_str() + 2)" work + inline wxUniChar operator*() const; + +private: + const wxString *m_str; + size_t m_offset; + bool m_owned; + + friend class WXDLLIMPEXP_FWD_BASE wxString; +}; + +// ---------------------------------------------------------------------------- +// wxStringPrintfMixin // --------------------------------------------------------------------------- -// string data prepended with some housekeeping info (used by wxString class), -// is never used directly (but had to be put here to allow inlining) -// --------------------------------------------------------------------------- -struct WXDLLIMPEXP_BASE wxStringData +// NB: VC6 has a bug that causes linker errors if you have template methods +// in a class using __declspec(dllimport). The solution is to split such +// class into two classes, one that contains the template methods and does +// *not* use WXDLLIMPEXP_BASE and another class that contains the rest +// (with DLL linkage). +// +// We only do this for VC6 here, because the code is less efficient +// (Printf() has to use dynamic_cast<>) and because OpenWatcom compiler +// cannot compile this code. + +#if defined(__VISUALC__) && __VISUALC__ < 1300 + #define wxNEEDS_WXSTRING_PRINTF_MIXIN +#endif + +#ifdef wxNEEDS_WXSTRING_PRINTF_MIXIN +// this class contains implementation of wxString's vararg methods, it's +// exported from wxBase DLL +class WXDLLIMPEXP_BASE wxStringPrintfMixinBase +{ +protected: + wxStringPrintfMixinBase() {} + +#if !wxUSE_UTF8_LOCALE_ONLY + int DoPrintfWchar(const wxChar *format, ...); + static wxString DoFormatWchar(const wxChar *format, ...); +#endif +#if wxUSE_UNICODE_UTF8 + int DoPrintfUtf8(const char *format, ...); + static wxString DoFormatUtf8(const char *format, ...); +#endif +}; + +// this class contains template wrappers for wxString's vararg methods, it's +// intentionally *not* exported from the DLL in order to fix the VC6 bug +// described above +class wxStringPrintfMixin : public wxStringPrintfMixinBase { - int nRefs; // reference count - size_t nDataLength, // actual string length - nAllocLength; // allocated memory size +private: + // to further complicate things, we can't return wxString from + // wxStringPrintfMixin::Format() because wxString is not yet declared at + // this point; the solution is to use this fake type trait template - this + // way the compiler won't know the return type until Format() is used + // (this doesn't compile with Watcom, but VC6 compiles it just fine): + template struct StringReturnType + { + typedef wxString type; + }; - // mimics declaration 'wxChar data[nAllocLength]' - wxChar* data() const { return (wxChar*)(this + 1); } +public: + // these are duplicated wxString methods, they're also declared below + // if !wxNEEDS_WXSTRING_PRINTF_MIXIN: - // empty string has a special ref count so it's never deleted - bool IsEmpty() const { return (nRefs == -1); } - bool IsShared() const { return (nRefs > 1); } + // static wxString Format(const wString& format, ...) ATTRIBUTE_PRINTF_1; + WX_DEFINE_VARARG_FUNC_SANS_N0(static typename StringReturnType::type, + Format, 1, (const wxFormatString&), + DoFormatWchar, DoFormatUtf8) + // We have to implement the version without template arguments manually + // because of the StringReturnType<> hack, although WX_DEFINE_VARARG_FUNC + // normally does it itself. It has to be a template so that we can use + // the hack, even though there's no real template parameter. We can't move + // it to wxStrig, because it would shadow these versions of Format() then. + template + inline static typename StringReturnType::type + Format(const T& fmt) + { + // NB: this doesn't compile if T is not (some form of) a string; + // this makes Format's prototype equivalent to + // Format(const wxFormatString& fmt) + return DoFormatWchar(wxFormatString(fmt)); + } - // lock/unlock - void Lock() { if ( !IsEmpty() ) nRefs++; } + // int Printf(const wxString& format, ...); + WX_DEFINE_VARARG_FUNC(int, Printf, 1, (const wxFormatString&), + DoPrintfWchar, DoPrintfUtf8) + // int sprintf(const wxString& format, ...) ATTRIBUTE_PRINTF_2; + WX_DEFINE_VARARG_FUNC(int, sprintf, 1, (const wxFormatString&), + DoPrintfWchar, DoPrintfUtf8) - // VC++ will refuse to inline Unlock but profiling shows that it is wrong -#if defined(__VISUALC__) && (__VISUALC__ >= 1200) - __forceinline -#endif - // VC++ free must take place in same DLL as allocation when using non dll - // run-time library (e.g. Multithreaded instead of Multithreaded DLL) -#if defined(__VISUALC__) && defined(_MT) && !defined(_DLL) - void Unlock() { if ( !IsEmpty() && --nRefs == 0) Free(); } - // we must not inline deallocation since allocation is not inlined - void Free(); -#else - void Unlock() { if ( !IsEmpty() && --nRefs == 0) free(this); } +protected: + wxStringPrintfMixin() : wxStringPrintfMixinBase() {} +}; +#endif // wxNEEDS_WXSTRING_PRINTF_MIXIN + + +// ---------------------------------------------------------------------------- +// wxString: string class trying to be compatible with std::string, MFC +// CString and wxWindows 1.x wxString all at once +// --------------------------------------------------------------------------- + +#ifdef wxNEEDS_WXSTRING_PRINTF_MIXIN + // "non dll-interface class 'wxStringPrintfMixin' used as base interface + // for dll-interface class 'wxString'" -- this is OK in our case + #pragma warning (disable:4275) #endif - // if we had taken control over string memory (GetWriteBuf), it's - // intentionally put in invalid state - void Validate(bool b) { nRefs = (b ? 1 : 0); } - bool IsValid() const { return (nRefs != 0); } +#if wxUSE_UNICODE_UTF8 +// see the comment near wxString::iterator for why we need this +class WXDLLIMPEXP_BASE wxStringIteratorNode +{ +public: + wxStringIteratorNode() + : m_str(NULL), m_citer(NULL), m_iter(NULL), m_prev(NULL), m_next(NULL) {} + wxStringIteratorNode(const wxString *str, + wxStringImpl::const_iterator *citer) + { DoSet(str, citer, NULL); } + wxStringIteratorNode(const wxString *str, wxStringImpl::iterator *iter) + { DoSet(str, NULL, iter); } + ~wxStringIteratorNode() + { clear(); } + + inline void set(const wxString *str, wxStringImpl::const_iterator *citer) + { clear(); DoSet(str, citer, NULL); } + inline void set(const wxString *str, wxStringImpl::iterator *iter) + { clear(); DoSet(str, NULL, iter); } + + const wxString *m_str; + wxStringImpl::const_iterator *m_citer; + wxStringImpl::iterator *m_iter; + wxStringIteratorNode *m_prev, *m_next; + +private: + inline void clear(); + inline void DoSet(const wxString *str, + wxStringImpl::const_iterator *citer, + wxStringImpl::iterator *iter); + + // the node belongs to a particular iterator instance, it's not copied + // when a copy of the iterator is made + DECLARE_NO_COPY_CLASS(wxStringIteratorNode) }; +#endif // wxUSE_UNICODE_UTF8 -class WXDLLIMPEXP_BASE wxStringBase +class WXDLLIMPEXP_BASE wxString +#ifdef wxNEEDS_WXSTRING_PRINTF_MIXIN + : public wxStringPrintfMixin +#endif { -public : + // NB: special care was taken in arranging the member functions in such order + // that all inline functions can be effectively inlined, verify that all + // performance critical functions are still inlined if you change order! +public: // an 'invalid' value for string index, moved to this place due to a CW bug static const size_t npos; -protected: - // points to data preceded by wxStringData structure with ref count info - wxChar *m_pchData; - - // accessor to string data - wxStringData* GetStringData() const { return (wxStringData*)m_pchData - 1; } - - // string (re)initialization functions - // initializes the string to the empty value (must be called only from - // ctors, use Reinit() otherwise) - void Init() { m_pchData = (wxChar *)wxEmptyString; } - // initializes the string with (a part of) C-string - void InitWith(const wxChar *psz, size_t nPos = 0, size_t nLen = npos); - // as Init, but also frees old data - void Reinit() { GetStringData()->Unlock(); Init(); } - - // memory allocation - // allocates memory for string of length nLen - bool AllocBuffer(size_t nLen); - // copies data to another string - bool AllocCopy(wxString&, int, int) const; - // effectively copies data to string - bool AssignCopy(size_t, const wxChar *); - - // append a (sub)string - bool ConcatSelf(size_t nLen, const wxChar *src, size_t nMaxLen); - bool ConcatSelf(size_t nLen, const wxChar *src) - { return ConcatSelf(nLen, src, nLen); } - - // functions called before writing to the string: they copy it if there - // are other references to our data (should be the only owner when writing) - bool CopyBeforeWrite(); - bool AllocBeforeWrite(size_t); - - // compatibility with wxString - bool Alloc(size_t nLen); + +private: + // if we hadn't made these operators private, it would be possible to + // compile "wxString s; s = 17;" without any warnings as 17 is implicitly + // converted to char in C and we do have operator=(char) + // + // NB: we don't need other versions (short/long and unsigned) as attempt + // to assign another numeric type to wxString will now result in + // ambiguity between operator=(char) and operator=(int) + wxString& operator=(int); + + // these methods are not implemented - there is _no_ conversion from int to + // string, you're doing something wrong if the compiler wants to call it! + // + // try `s << i' or `s.Printf("%d", i)' instead + wxString(int); + + + // buffer for holding temporary substring when using any of the methods + // that take (char*,size_t) or (wchar_t*,size_t) arguments: + template + struct SubstrBufFromType + { + T data; + size_t len; + + SubstrBufFromType(const T& data_, size_t len_) + : data(data_), len(len_) {} + }; + +#if wxUSE_UNICODE_UTF8 + // even char* -> char* needs conversion, from locale charset to UTF-8 + typedef SubstrBufFromType SubstrBufFromWC; + typedef SubstrBufFromType SubstrBufFromMB; +#elif wxUSE_UNICODE_WCHAR + typedef SubstrBufFromType SubstrBufFromWC; + typedef SubstrBufFromType SubstrBufFromMB; +#else + typedef SubstrBufFromType SubstrBufFromMB; + typedef SubstrBufFromType SubstrBufFromWC; +#endif + + + // Functions implementing primitive operations on string data; wxString + // methods and iterators are implemented in terms of it. The differences + // between UTF-8 and wchar_t* representations of the string are mostly + // contained here. + +#if wxUSE_UNICODE_UTF8 + static SubstrBufFromMB ConvertStr(const char *psz, size_t nLength, + const wxMBConv& conv); + static SubstrBufFromWC ConvertStr(const wchar_t *pwz, size_t nLength, + const wxMBConv& conv); +#elif wxUSE_UNICODE_WCHAR + static SubstrBufFromMB ConvertStr(const char *psz, size_t nLength, + const wxMBConv& conv); +#else + static SubstrBufFromWC ConvertStr(const wchar_t *pwz, size_t nLength, + const wxMBConv& conv); +#endif + +#if !wxUSE_UNICODE_UTF8 // wxUSE_UNICODE_WCHAR or !wxUSE_UNICODE + // returns C string encoded as the implementation expects: + #if wxUSE_UNICODE + static const wchar_t* ImplStr(const wchar_t* str) + { return str ? str : wxT(""); } + static const SubstrBufFromWC ImplStr(const wchar_t* str, size_t n) + { return SubstrBufFromWC(str, (str && n == npos) ? wxWcslen(str) : n); } + static wxWCharBuffer ImplStr(const char* str, + const wxMBConv& conv = wxConvLibc) + { return ConvertStr(str, npos, conv).data; } + static SubstrBufFromMB ImplStr(const char* str, size_t n, + const wxMBConv& conv = wxConvLibc) + { return ConvertStr(str, n, conv); } + #else + static const char* ImplStr(const char* str, + const wxMBConv& WXUNUSED(conv) = wxConvLibc) + { return str ? str : ""; } + static const SubstrBufFromMB ImplStr(const char* str, size_t n, + const wxMBConv& WXUNUSED(conv) = wxConvLibc) + { return SubstrBufFromMB(str, (str && n == npos) ? wxStrlen(str) : n); } + static wxCharBuffer ImplStr(const wchar_t* str) + { return ConvertStr(str, npos, wxConvLibc).data; } + static SubstrBufFromWC ImplStr(const wchar_t* str, size_t n) + { return ConvertStr(str, n, wxConvLibc); } + #endif + + // translates position index in wxString to/from index in underlying + // wxStringImpl: + static size_t PosToImpl(size_t pos) { return pos; } + static void PosLenToImpl(size_t pos, size_t len, + size_t *implPos, size_t *implLen) + { *implPos = pos; *implLen = len; } + static size_t LenToImpl(size_t len) { return len; } + static size_t PosFromImpl(size_t pos) { return pos; } + +#else // wxUSE_UNICODE_UTF8 + + static wxCharBuffer ImplStr(const char* str, + const wxMBConv& conv = wxConvLibc) + { return ConvertStr(str, npos, conv).data; } + static SubstrBufFromMB ImplStr(const char* str, size_t n, + const wxMBConv& conv = wxConvLibc) + { return ConvertStr(str, n, conv); } + + static wxCharBuffer ImplStr(const wchar_t* str) + { return ConvertStr(str, npos, wxMBConvUTF8()).data; } + static SubstrBufFromWC ImplStr(const wchar_t* str, size_t n) + { return ConvertStr(str, n, wxMBConvUTF8()); } + + size_t PosToImpl(size_t pos) const + { + if ( pos == 0 || pos == npos ) + return pos; + else + return (begin() + pos).impl() - m_impl.begin(); + } + + void PosLenToImpl(size_t pos, size_t len, size_t *implPos, size_t *implLen) const; + + size_t LenToImpl(size_t len) const + { + size_t pos, len2; + PosLenToImpl(0, len, &pos, &len2); + return len2; + } + + size_t PosFromImpl(size_t pos) const + { + if ( pos == 0 || pos == npos ) + return pos; + else + return const_iterator(this, m_impl.begin() + pos) - begin(); + } +#endif // !wxUSE_UNICODE_UTF8/wxUSE_UNICODE_UTF8 + public: // standard types typedef wxUniChar value_type; @@ -302,47 +573,71 @@ public: typedef size_t size_type; typedef wxUniChar const_reference; - #define WX_STR_ITERATOR_IMPL(iterator_name, pointer_type, \ - reference_type, reference_ctor) \ +#if wxUSE_STL + #if wxUSE_UNICODE_UTF8 + // random access is not O(1), as required by Random Access Iterator + #define WX_STR_ITERATOR_TAG std::bidirectional_iterator_tag + #else + #define WX_STR_ITERATOR_TAG std::random_access_iterator_tag + #endif +#else + #define WX_STR_ITERATOR_TAG void /* dummy type */ +#endif + + #define WX_STR_ITERATOR_IMPL(iterator_name, pointer_type, reference_type) \ + private: \ + typedef wxStringImpl::iterator_name underlying_iterator; \ public: \ + typedef WX_STR_ITERATOR_TAG iterator_category; \ typedef wxUniChar value_type; \ + typedef int difference_type; \ typedef reference_type reference; \ typedef pointer_type pointer; \ \ - iterator_name(const iterator_name& i) : m_cur(i.m_cur) {} \ - \ - reference operator*() const { return reference_ctor; } \ + reference operator[](size_t n) const { return *(*this + n); } \ \ iterator_name& operator++() \ - { ++m_cur; return *this; } \ - iterator_name operator++(int) \ - { iterator_name tmp = *this; ++m_cur; return tmp; } \ + { wxStringOperations::IncIter(m_cur); return *this; } \ iterator_name& operator--() \ - { --m_cur; return *this; } \ + { wxStringOperations::DecIter(m_cur); return *this; } \ + iterator_name operator++(int) \ + { \ + iterator_name tmp = *this; \ + wxStringOperations::IncIter(m_cur); \ + return tmp; \ + } \ iterator_name operator--(int) \ - { iterator_name tmp = *this; --m_cur; return tmp; } \ + { \ + iterator_name tmp = *this; \ + wxStringOperations::DecIter(m_cur); \ + return tmp; \ + } \ \ - iterator_name operator+(int n) const \ - { return iterator_name(m_cur + n); } \ - iterator_name operator+(size_t n) const \ - { return iterator_name(m_cur + n); } \ - iterator_name operator-(int n) const \ - { return iterator_name(m_cur - n); } \ - iterator_name operator-(size_t n) const \ - { return iterator_name(m_cur - n); } \ - iterator_name operator+=(int n) \ - { m_cur += n; return *this; } \ - iterator_name operator+=(size_t n) \ - { m_cur += n; return *this; } \ - iterator_name operator-=(int n) \ - { m_cur -= n; return *this; } \ - iterator_name operator-=(size_t n) \ - { m_cur -= n; return *this; } \ + iterator_name& operator+=(int n) \ + { \ + m_cur = wxStringOperations::AddToIter(m_cur, n); \ + return *this; \ + } \ + iterator_name& operator+=(size_t n) \ + { \ + m_cur = wxStringOperations::AddToIter(m_cur, (int)n); \ + return *this; \ + } \ + iterator_name& operator-=(int n) \ + { \ + m_cur = wxStringOperations::AddToIter(m_cur, -n); \ + return *this; \ + } \ + iterator_name& operator-=(size_t n) \ + { \ + m_cur = wxStringOperations::AddToIter(m_cur, -(int)n); \ + return *this; \ + } \ \ - unsigned operator-(const iterator_name& i) const \ - { return m_cur - i.m_cur; } \ + difference_type operator-(const iterator_name& i) const \ + { return wxStringOperations::DiffIters(m_cur, i.m_cur); } \ \ - bool operator==(const iterator_name&i) const \ + bool operator==(const iterator_name& i) const \ { return m_cur == i.m_cur; } \ bool operator!=(const iterator_name& i) const \ { return m_cur != i.m_cur; } \ @@ -356,50 +651,190 @@ public: bool operator>=(const iterator_name& i) const \ { return m_cur >= i.m_cur; } \ \ - protected: \ + private: \ /* for internal wxString use only: */ \ - iterator_name(pointer ptr) : m_cur(ptr) {} \ - operator pointer() const { return m_cur; } \ + underlying_iterator impl() const { return m_cur; } \ \ - friend class WXDLLIMPEXP_BASE wxString; \ - friend class WXDLLIMPEXP_BASE wxStringBase; \ - friend class WXDLLIMPEXP_BASE wxCStrData; \ + friend class wxString; \ + friend class wxCStrData; \ \ - protected: \ - pointer m_cur; + private: \ + underlying_iterator m_cur + + class WXDLLIMPEXP_FWD_BASE const_iterator; + +#if wxUSE_UNICODE_UTF8 + // NB: In UTF-8 build, (non-const) iterator needs to keep reference + // to the underlying wxStringImpl, because UTF-8 is variable-length + // encoding and changing the value pointer to by an iterator (using + // its operator*) requires calling wxStringImpl::replace() if the old + // and new values differ in their encoding's length. + // + // Furthermore, the replace() call may invalid all iterators for the + // string, so we have to keep track of outstanding iterators and update + // them if replace() happens. + // + // This is implemented by maintaining linked list of iterators for every + // string and traversing it in wxUniCharRef::operator=(). Head of the + // list is stored in wxString. (FIXME-UTF8) + + class WXDLLIMPEXP_BASE iterator + { + WX_STR_ITERATOR_IMPL(iterator, wxChar*, wxUniCharRef); + + public: + iterator() {} + iterator(const iterator& i) + : m_cur(i.m_cur), m_node(i.str(), &m_cur) {} + iterator& operator=(const iterator& i) + { m_cur = i.m_cur; m_node.set(i.str(), &m_cur); return *this; } + + reference operator*() + { return wxUniCharRef::CreateForString(m_node, m_cur); } + + iterator operator+(int n) const + { return iterator(str(), wxStringOperations::AddToIter(m_cur, n)); } + iterator operator+(size_t n) const + { return iterator(str(), wxStringOperations::AddToIter(m_cur, (int)n)); } + iterator operator-(int n) const + { return iterator(str(), wxStringOperations::AddToIter(m_cur, -n)); } + iterator operator-(size_t n) const + { return iterator(str(), wxStringOperations::AddToIter(m_cur, -(int)n)); } + + private: + iterator(wxString *str, underlying_iterator ptr) + : m_cur(ptr), m_node(str, &m_cur) {} + + wxString* str() const { return wx_const_cast(wxString*, m_node.m_str); } + + wxStringIteratorNode m_node; + + friend class const_iterator; + }; + + class WXDLLIMPEXP_BASE const_iterator + { + // NB: reference_type is intentionally value, not reference, the character + // may be encoded differently in wxString data: + WX_STR_ITERATOR_IMPL(const_iterator, const wxChar*, wxUniChar); + + public: + const_iterator() {} + const_iterator(const const_iterator& i) + : m_cur(i.m_cur), m_node(i.str(), &m_cur) {} + const_iterator(const iterator& i) + : m_cur(i.m_cur), m_node(i.str(), &m_cur) {} + + const_iterator& operator=(const const_iterator& i) + { m_cur = i.m_cur; m_node.set(i.str(), &m_cur); return *this; } + const_iterator& operator=(const iterator& i) + { m_cur = i.m_cur; m_node.set(i.str(), &m_cur); return *this; } + + reference operator*() const + { return wxStringOperations::DecodeChar(m_cur); } + + const_iterator operator+(int n) const + { return const_iterator(str(), wxStringOperations::AddToIter(m_cur, n)); } + const_iterator operator+(size_t n) const + { return const_iterator(str(), wxStringOperations::AddToIter(m_cur, (int)n)); } + const_iterator operator-(int n) const + { return const_iterator(str(), wxStringOperations::AddToIter(m_cur, -n)); } + const_iterator operator-(size_t n) const + { return const_iterator(str(), wxStringOperations::AddToIter(m_cur, -(int)n)); } - class const_iterator; + private: + // for internal wxString use only: + const_iterator(const wxString *str, underlying_iterator ptr) + : m_cur(ptr), m_node(str, &m_cur) {} + + const wxString* str() const { return m_node.m_str; } + + wxStringIteratorNode m_node; + }; - class iterator + size_t IterToImplPos(wxString::iterator i) const + { return wxStringImpl::const_iterator(i.impl()) - m_impl.begin(); } + +#else // !wxUSE_UNICODE_UTF8 + + class WXDLLIMPEXP_BASE iterator { - WX_STR_ITERATOR_IMPL(iterator, wxChar*, wxUniCharRef, - wxUniCharRef::CreateForString(m_cur)) + WX_STR_ITERATOR_IMPL(iterator, wxChar*, wxUniCharRef); + + public: + iterator() {} + iterator(const iterator& i) : m_cur(i.m_cur) {} + + reference operator*() + { return wxUniCharRef::CreateForString(m_cur); } + + iterator operator+(int n) const + { return iterator(wxStringOperations::AddToIter(m_cur, n)); } + iterator operator+(size_t n) const + { return iterator(wxStringOperations::AddToIter(m_cur, (int)n)); } + iterator operator-(int n) const + { return iterator(wxStringOperations::AddToIter(m_cur, -n)); } + iterator operator-(size_t n) const + { return iterator(wxStringOperations::AddToIter(m_cur, -(int)n)); } + + private: + // for internal wxString use only: + iterator(underlying_iterator ptr) : m_cur(ptr) {} + iterator(wxString *WXUNUSED(str), underlying_iterator ptr) : m_cur(ptr) {} friend class const_iterator; }; - class const_iterator + class WXDLLIMPEXP_BASE const_iterator { // NB: reference_type is intentionally value, not reference, the character // may be encoded differently in wxString data: - WX_STR_ITERATOR_IMPL(const_iterator, const wxChar*, wxUniChar, - wxUniChar(*m_cur)) + WX_STR_ITERATOR_IMPL(const_iterator, const wxChar*, wxUniChar); public: + const_iterator() {} + const_iterator(const const_iterator& i) : m_cur(i.m_cur) {} const_iterator(const iterator& i) : m_cur(i.m_cur) {} + + reference operator*() const + { return wxStringOperations::DecodeChar(m_cur); } + + const_iterator operator+(int n) const + { return const_iterator(wxStringOperations::AddToIter(m_cur, n)); } + const_iterator operator+(size_t n) const + { return const_iterator(wxStringOperations::AddToIter(m_cur, (int)n)); } + const_iterator operator-(int n) const + { return const_iterator(wxStringOperations::AddToIter(m_cur, -n)); } + const_iterator operator-(size_t n) const + { return const_iterator(wxStringOperations::AddToIter(m_cur, -(int)n)); } + + private: + // for internal wxString use only: + const_iterator(underlying_iterator ptr) : m_cur(ptr) {} + const_iterator(const wxString *WXUNUSED(str), underlying_iterator ptr) + : m_cur(ptr) {} }; +#endif // wxUSE_UNICODE_UTF8/!wxUSE_UNICODE_UTF8 - #undef WX_STR_ITERATOR + #undef WX_STR_ITERATOR_TAG + #undef WX_STR_ITERATOR_IMPL + + friend class iterator; + friend class const_iterator; template class reverse_iterator_impl { public: typedef T iterator_type; + + typedef typename T::iterator_category iterator_category; typedef typename T::value_type value_type; + typedef typename T::difference_type difference_type; typedef typename T::reference reference; typedef typename T::pointer *pointer; + reverse_iterator_impl() {} reverse_iterator_impl(iterator_type i) : m_cur(i) {} reverse_iterator_impl(const reverse_iterator_impl& ri) : m_cur(ri.m_cur) {} @@ -407,6 +842,7 @@ public: iterator_type base() const { return m_cur; } reference operator*() const { return *(m_cur-1); } + reference operator[](size_t n) const { return *(*this + n); } reverse_iterator_impl& operator++() { --m_cur; return *this; } @@ -417,11 +853,41 @@ public: reverse_iterator_impl operator--(int) { reverse_iterator_impl tmp = *this; ++m_cur; return tmp; } + // NB: explicit in the functions below is to keep BCC 5.5 happy + reverse_iterator_impl operator+(int n) const + { return reverse_iterator_impl(m_cur - n); } + reverse_iterator_impl operator+(size_t n) const + { return reverse_iterator_impl(m_cur - n); } + reverse_iterator_impl operator-(int n) const + { return reverse_iterator_impl(m_cur + n); } + reverse_iterator_impl operator-(size_t n) const + { return reverse_iterator_impl(m_cur + n); } + reverse_iterator_impl operator+=(int n) + { m_cur -= n; return *this; } + reverse_iterator_impl operator+=(size_t n) + { m_cur -= n; return *this; } + reverse_iterator_impl operator-=(int n) + { m_cur += n; return *this; } + reverse_iterator_impl operator-=(size_t n) + { m_cur += n; return *this; } + + unsigned operator-(const reverse_iterator_impl& i) const + { return i.m_cur - m_cur; } + bool operator==(const reverse_iterator_impl& ri) const { return m_cur == ri.m_cur; } bool operator!=(const reverse_iterator_impl& ri) const { return !(*this == ri); } + bool operator<(const reverse_iterator_impl& i) const + { return m_cur > i.m_cur; } + bool operator>(const reverse_iterator_impl& i) const + { return m_cur < i.m_cur; } + bool operator<=(const reverse_iterator_impl& i) const + { return m_cur >= i.m_cur; } + bool operator>=(const reverse_iterator_impl& i) const + { return m_cur <= i.m_cur; } + private: iterator_type m_cur; }; @@ -429,559 +895,227 @@ public: typedef reverse_iterator_impl reverse_iterator; typedef reverse_iterator_impl const_reverse_iterator; +private: + // used to transform an expression built using c_str() (and hence of type + // wxCStrData) to an iterator into the string + static const_iterator CreateConstIterator(const wxCStrData& data) + { + return const_iterator(data.m_str, + (data.m_str->begin() + data.m_offset).impl()); + } + + // in UTF-8 STL build, creation from std::string requires conversion under + // non-UTF8 locales, so we can't have and use wxString(wxStringImpl) ctor; + // instead we define dummy type that lets us have wxString ctor for creation + // from wxStringImpl that couldn't be used by user code (in all other builds, + // "standard" ctors can be used): +#if wxUSE_UNICODE_UTF8 && wxUSE_STL_BASED_WXSTRING + struct CtorFromStringImplTag {}; + wxString(CtorFromStringImplTag* WXUNUSED(dummy), const wxStringImpl& src) + : m_impl(src) {} + + static wxString FromImpl(const wxStringImpl& src) + { return wxString((CtorFromStringImplTag*)NULL, src); } +#else + #if !wxUSE_STL_BASED_WXSTRING + wxString(const wxStringImpl& src) : m_impl(src) { } + // else: already defined as wxString(wxStdString) below + #endif + static wxString FromImpl(const wxStringImpl& src) { return wxString(src); } +#endif + +public: // constructors and destructor // ctor for an empty string - wxStringBase() { Init(); } + wxString() {} + // copy ctor - wxStringBase(const wxStringBase& stringSrc) - { - wxASSERT_MSG( stringSrc.GetStringData()->IsValid(), - _T("did you forget to call UngetWriteBuf()?") ); + wxString(const wxString& stringSrc) : m_impl(stringSrc.m_impl) { } - if ( stringSrc.empty() ) { - // nothing to do for an empty string - Init(); - } - else { - m_pchData = stringSrc.m_pchData; // share same data - GetStringData()->Lock(); // => one more copy - } - } // string containing nRepeat copies of ch - wxStringBase(size_type nRepeat, wxUniChar ch); - // ctor takes first nLength characters from C string - // (default value of npos means take all the string) - wxStringBase(const wxChar *psz) - { InitWith(psz, 0, npos); } - wxStringBase(const wxChar *psz, size_t nLength) - { InitWith(psz, 0, nLength); } - wxStringBase(const wxChar *psz, - const wxMBConv& WXUNUSED(conv), - size_t nLength = npos) - { InitWith(psz, 0, nLength); } - // take nLen chars starting at nPos - wxStringBase(const wxStringBase& str, size_t nPos, size_t nLen) + wxString(wxUniChar ch, size_t nRepeat = 1 ) + { assign(nRepeat, ch); } + wxString(size_t nRepeat, wxUniChar ch) + { assign(nRepeat, ch); } + wxString(wxUniCharRef ch, size_t nRepeat = 1) + { assign(nRepeat, ch); } + wxString(size_t nRepeat, wxUniCharRef ch) + { assign(nRepeat, ch); } + wxString(char ch, size_t nRepeat = 1) + { assign(nRepeat, ch); } + wxString(size_t nRepeat, char ch) + { assign(nRepeat, ch); } + wxString(wchar_t ch, size_t nRepeat = 1) + { assign(nRepeat, ch); } + wxString(size_t nRepeat, wchar_t ch) + { assign(nRepeat, ch); } + + // ctors from char* strings: + wxString(const char *psz) + : m_impl(ImplStr(psz)) {} + wxString(const char *psz, const wxMBConv& conv) + : m_impl(ImplStr(psz, conv)) {} + wxString(const char *psz, size_t nLength) + { assign(psz, nLength); } + wxString(const char *psz, const wxMBConv& conv, size_t nLength) { - wxASSERT_MSG( str.GetStringData()->IsValid(), - _T("did you forget to call UngetWriteBuf()?") ); - Init(); - size_t strLen = str.length() - nPos; nLen = strLen < nLen ? strLen : nLen; - InitWith(str.c_str(), nPos, nLen); + SubstrBufFromMB str(ImplStr(psz, nLength, conv)); + m_impl.assign(str.data, str.len); } - // take all characters from pStart to pEnd - wxStringBase(const void *pStart, const void *pEnd); - // dtor is not virtual, this class must not be inherited from! - ~wxStringBase() + // and unsigned char*: + wxString(const unsigned char *psz) + : m_impl(ImplStr((const char*)psz)) {} + wxString(const unsigned char *psz, const wxMBConv& conv) + : m_impl(ImplStr((const char*)psz, conv)) {} + wxString(const unsigned char *psz, size_t nLength) + { assign((const char*)psz, nLength); } + wxString(const unsigned char *psz, const wxMBConv& conv, size_t nLength) { -#if defined(__VISUALC__) && (__VISUALC__ >= 1200) - //RN - according to the above VC++ does indeed inline this, - //even though it spits out two warnings - #pragma warning (disable:4714) -#endif - - GetStringData()->Unlock(); + SubstrBufFromMB str(ImplStr((const char*)psz, nLength, conv)); + m_impl.assign(str.data, str.len); } -#if defined(__VISUALC__) && (__VISUALC__ >= 1200) - //re-enable inlining warning - #pragma warning (default:4714) -#endif - // overloaded assignment - // from another wxString - wxStringBase& operator=(const wxStringBase& stringSrc); - // from a character - wxStringBase& operator=(wxUniChar ch); - // from a C string - wxStringBase& operator=(const wxChar *psz); - - // return the length of the string - size_type length() const { return GetStringData()->nDataLength; } - // return the length of the string - size_type size() const { return length(); } - // return the maximum size of the string - size_type max_size() const { return npos; } - // resize the string, filling the space with c if c != 0 - void resize(size_t nSize, wxUniChar ch = wxT('\0')); - // delete the contents of the string - void clear() { erase(0, npos); } - // returns true if the string is empty - bool empty() const { return length() == 0; } - // inform string about planned change in size - void reserve(size_t sz) { Alloc(sz); } - size_type capacity() const { return GetStringData()->nAllocLength; } - - // lib.string.access - // return the character at position n - value_type at(size_type n) const - { wxASSERT_VALID_INDEX( n ); return m_pchData[n]; } - // returns the writable character at position n - reference at(size_type n) - { - wxASSERT_VALID_INDEX( n ); - CopyBeforeWrite(); - return wxUniCharRef::CreateForString(&m_pchData[n]); - } - - // lib.string.modifiers - // append elements str[pos], ..., str[pos+n] - wxStringBase& append(const wxStringBase& str, size_t pos, size_t n) - { - wxASSERT(pos <= str.length()); - ConcatSelf(n, str.c_str() + pos, str.length() - pos); - return *this; - } - // append a string - wxStringBase& append(const wxStringBase& str) - { ConcatSelf(str.length(), str.c_str()); return *this; } - // append first n (or all if n == npos) characters of sz - wxStringBase& append(const wxChar *sz) - { ConcatSelf(wxStrlen(sz), sz); return *this; } - wxStringBase& append(const wxChar *sz, size_t n) - { ConcatSelf(n, sz); return *this; } - // append n copies of ch - wxStringBase& append(size_t n, wxUniChar ch); - // append from first to last - wxStringBase& append(const_iterator first, const_iterator last) - { ConcatSelf(last - first, first); return *this; } - - // same as `this_string = str' - wxStringBase& assign(const wxStringBase& str) - { return *this = str; } - // same as ` = str[pos..pos + n] - wxStringBase& assign(const wxStringBase& str, size_t pos, size_t n) - { clear(); return append(str, pos, n); } - // same as `= first n (or all if n == npos) characters of sz' - wxStringBase& assign(const wxChar *sz) - { clear(); return append(sz, wxStrlen(sz)); } - wxStringBase& assign(const wxChar *sz, size_t n) - { clear(); return append(sz, n); } - // same as `= n copies of ch' - wxStringBase& assign(size_t n, wxUniChar ch) - { clear(); return append(n, ch); } - // assign from first to last - wxStringBase& assign(const_iterator first, const_iterator last) - { clear(); return append(first, last); } - - // first valid index position - const_iterator begin() const { return m_pchData; } - iterator begin(); - // position one after the last valid one - const_iterator end() const { return m_pchData + length(); } - iterator end(); - - // first element of the reversed string - const_reverse_iterator rbegin() const { return const_reverse_iterator(end()); } - reverse_iterator rbegin() { return reverse_iterator(end()); } - // one beyond the end of the reversed string - const_reverse_iterator rend() const { return const_reverse_iterator(begin()); } - reverse_iterator rend() { return reverse_iterator(begin()); } - - // insert another string - wxStringBase& insert(size_t nPos, const wxStringBase& str) - { - wxASSERT( str.GetStringData()->IsValid() ); - return insert(nPos, str.c_str(), str.length()); - } - // insert n chars of str starting at nStart (in str) - wxStringBase& insert(size_t nPos, const wxStringBase& str, size_t nStart, size_t n) - { - wxASSERT( str.GetStringData()->IsValid() ); - wxASSERT( nStart < str.length() ); - size_t strLen = str.length() - nStart; - n = strLen < n ? strLen : n; - return insert(nPos, str.c_str() + nStart, n); - } - // insert first n (or all if n == npos) characters of sz - wxStringBase& insert(size_t nPos, const wxChar *sz, size_t n = npos); - // insert n copies of ch - wxStringBase& insert(size_t nPos, size_t n, wxUniChar ch) - { return insert(nPos, wxStringBase(n, ch)); } - iterator insert(iterator it, wxUniChar ch) - { size_t idx = it - begin(); insert(idx, 1, ch); return begin() + idx; } - void insert(iterator it, const_iterator first, const_iterator last) - { insert(it - begin(), first, last - first); } - void insert(iterator it, size_type n, wxUniChar ch) - { insert(it - begin(), n, ch); } - - // delete characters from nStart to nStart + nLen - wxStringBase& erase(size_type pos = 0, size_type n = npos); - iterator erase(iterator first, iterator last) - { - size_t idx = first - begin(); - erase(idx, last - first); - return begin() + idx; - } - iterator erase(iterator first); - - // explicit conversion to C string (use this with printf()!) - const wxChar* c_str() const { return m_pchData; } - const wxChar* data() const { return m_pchData; } - - // replaces the substring of length nLen starting at nStart - wxStringBase& replace(size_t nStart, size_t nLen, const wxChar* sz); - // replaces the substring of length nLen starting at nStart - wxStringBase& replace(size_t nStart, size_t nLen, const wxStringBase& str) - { return replace(nStart, nLen, str.c_str()); } - // replaces the substring with nCount copies of ch - wxStringBase& replace(size_t nStart, size_t nLen, size_t nCount, wxUniChar ch); - // replaces a substring with another substring - wxStringBase& replace(size_t nStart, size_t nLen, - const wxStringBase& str, size_t nStart2, size_t nLen2); - // replaces the substring with first nCount chars of sz - wxStringBase& replace(size_t nStart, size_t nLen, - const wxChar* sz, size_t nCount); - wxStringBase& replace(iterator first, iterator last, const_pointer s) - { return replace(first - begin(), last - first, s); } - wxStringBase& replace(iterator first, iterator last, const_pointer s, - size_type n) - { return replace(first - begin(), last - first, s, n); } - wxStringBase& replace(iterator first, iterator last, const wxStringBase& s) - { return replace(first - begin(), last - first, s); } - wxStringBase& replace(iterator first, iterator last, size_type n, wxUniChar c) - { return replace(first - begin(), last - first, n, c); } - wxStringBase& replace(iterator first, iterator last, - const_iterator first1, const_iterator last1) - { return replace(first - begin(), last - first, first1, last1 - first1); } - - // swap two strings - void swap(wxStringBase& str); - - // All find() functions take the nStart argument which specifies the - // position to start the search on, the default value is 0. All functions - // return npos if there were no match. - - // find a substring - size_t find(const wxStringBase& str, size_t nStart = 0) const; - - // find first n characters of sz - size_t find(const wxChar* sz, size_t nStart = 0, size_t n = npos) const; - - // find the first occurence of character ch after nStart - size_t find(wxUniChar ch, size_t nStart = 0) const; - - // rfind() family is exactly like find() but works right to left - - // as find, but from the end - size_t rfind(const wxStringBase& str, size_t nStart = npos) const; - - // as find, but from the end - size_t rfind(const wxChar* sz, size_t nStart = npos, - size_t n = npos) const; - // as find, but from the end - size_t rfind(wxUniChar ch, size_t nStart = npos) const; - - // find first/last occurence of any character in the set - - // as strpbrk() but starts at nStart, returns npos if not found - size_t find_first_of(const wxStringBase& str, size_t nStart = 0) const - { return find_first_of(str.c_str(), nStart); } - // same as above - size_t find_first_of(const wxChar* sz, size_t nStart = 0) const; - size_t find_first_of(const wxChar* sz, size_t nStart, size_t n) const; - // same as find(char, size_t) - size_t find_first_of(wxUniChar c, size_t nStart = 0) const - { return find(c, nStart); } - // find the last (starting from nStart) char from str in this string - size_t find_last_of (const wxStringBase& str, size_t nStart = npos) const - { return find_last_of(str.c_str(), nStart); } - // same as above - size_t find_last_of (const wxChar* sz, size_t nStart = npos) const; - size_t find_last_of(const wxChar* sz, size_t nStart, size_t n) const; - // same as above - size_t find_last_of(wxUniChar c, size_t nStart = npos) const - { return rfind(c, nStart); } - - // find first/last occurence of any character not in the set - - // as strspn() (starting from nStart), returns npos on failure - size_t find_first_not_of(const wxStringBase& str, size_t nStart = 0) const - { return find_first_not_of(str.c_str(), nStart); } - // same as above - size_t find_first_not_of(const wxChar* sz, size_t nStart = 0) const; - size_t find_first_not_of(const wxChar* sz, size_t nStart, size_t n) const; - // same as above - size_t find_first_not_of(wxUniChar ch, size_t nStart = 0) const; - // as strcspn() - size_t find_last_not_of(const wxStringBase& str, size_t nStart = npos) const - { return find_last_not_of(str.c_str(), nStart); } - // same as above - size_t find_last_not_of(const wxChar* sz, size_t nStart = npos) const; - size_t find_last_not_of(const wxChar* sz, size_t nStart, size_t n) const; - // same as above - size_t find_last_not_of(wxUniChar ch, size_t nStart = npos) const; - - // All compare functions return -1, 0 or 1 if the [sub]string is less, - // equal or greater than the compare() argument. - - // comparison with another string - int compare(const wxStringBase& str) const; - // comparison with a substring - int compare(size_t nStart, size_t nLen, const wxStringBase& str) const; - // comparison of 2 substrings - int compare(size_t nStart, size_t nLen, - const wxStringBase& str, size_t nStart2, size_t nLen2) const; - // comparison with a c string - int compare(const wxChar* sz) const; - // substring comparison with first nCount characters of sz - int compare(size_t nStart, size_t nLen, - const wxChar* sz, size_t nCount = npos) const; - - size_type copy(wxChar* s, size_type n, size_type pos = 0); - - // substring extraction - wxStringBase substr(size_t nStart = 0, size_t nLen = npos) const; - - // string += string - wxStringBase& operator+=(const wxStringBase& s) { return append(s); } - // string += C string - wxStringBase& operator+=(const wxChar *psz) { return append(psz); } - // string += char - wxStringBase& operator+=(wxUniChar ch) { return append(1, ch); } - wxStringBase& operator+=(wxUniCharRef ch) { return append(1, ch); } - wxStringBase& operator+=(char ch) { return append(1, ch); } - wxStringBase& operator+=(wchar_t ch) { return append(1, ch); } -}; - -#endif // !wxUSE_STL - -// ---------------------------------------------------------------------------- -// wxCStrData -// ---------------------------------------------------------------------------- - -// Lightweight object returned by wxString::c_str() and implicitly convertible -// to either const char* or const wchar_t*. -class WXDLLIMPEXP_BASE wxCStrData -{ -private: - // Ctors; for internal use by wxString and wxCStrData only - wxCStrData(const wxString *str, size_t offset = 0, bool owned = false) - : m_str(str), m_offset(offset), m_owned(owned) {} - -public: - // Ctor constructs the object from char literal; they are needed to make - // operator?: compile and they intentionally take char*, not const char* - wxCStrData(char *buf); - wxCStrData(wchar_t *buf); + // ctors from wchar_t* strings: + wxString(const wchar_t *pwz) + : m_impl(ImplStr(pwz)) {} + wxString(const wchar_t *pwz, const wxMBConv& WXUNUSED(conv)) + : m_impl(ImplStr(pwz)) {} + wxString(const wchar_t *pwz, size_t nLength) + { assign(pwz, nLength); } + wxString(const wchar_t *pwz, const wxMBConv& WXUNUSED(conv), size_t nLength) + { assign(pwz, nLength); } + + wxString(const wxCharBuffer& buf) + { assign(buf.data()); } // FIXME-UTF8: fix for embedded NUL and buffer length + wxString(const wxWCharBuffer& buf) + { assign(buf.data()); } // FIXME-UTF8: fix for embedded NUL and buffer length + + wxString(const wxCStrData& cstr) + : m_impl(cstr.AsString().m_impl) { } + + // as we provide both ctors with this signature for both char and unsigned + // char string, we need to provide one for wxCStrData to resolve ambiguity + wxString(const wxCStrData& cstr, size_t nLength) + : m_impl(cstr.AsString().Mid(0, nLength).m_impl) {} + + // and because wxString is convertible to wxCStrData and const wxChar * + // we also need to provide this one + wxString(const wxString& str, size_t nLength) + { assign(str, nLength); } - ~wxCStrData(); + // even if we're not built with wxUSE_STL == 1 it is very convenient to allow + // implicit conversions from std::string to wxString and vice verse as this + // allows to use the same strings in non-GUI and GUI code, however we don't + // want to unconditionally add this ctor as it would make wx lib dependent on + // libstdc++ on some Linux versions which is bad, so instead we ask the + // client code to define this wxUSE_STD_STRING symbol if they need it +#if wxUSE_STD_STRING + #if wxUSE_UNICODE_WCHAR + wxString(const wxStdWideString& str) : m_impl(str) {} + #else // UTF-8 or ANSI + wxString(const wxStdWideString& str) + { assign(str.c_str(), str.length()); } + #endif + + #if !wxUSE_UNICODE // ANSI build + // FIXME-UTF8: do this in UTF8 build #if wxUSE_UTF8_LOCALE_ONLY, too + wxString(const std::string& str) : m_impl(str) {} + #else // Unicode + wxString(const std::string& str) + { assign(str.c_str(), str.length()); } + #endif +#endif // wxUSE_STD_STRING - // FIXME: we'll need convertors for both char* and wchar_t* and NONE - // for wxChar*, but that's after completing the transition to - // "smart" wxUniChar class. For now, just have conversion to - // char* in ANSI build and wchar_t in Unicode build. -#if wxUSE_UNICODE - const wchar_t* AsWChar() const; - operator const wchar_t*() const { return AsWChar(); } + // Unlike ctor from std::string, we provide conversion to std::string only + // if wxUSE_STL and not merely wxUSE_STD_STRING (which is on by default), + // because it conflicts with operator const char/wchar_t*: +#if wxUSE_STL + #if wxUSE_UNICODE_WCHAR && wxUSE_STL_BASED_WXSTRING + // wxStringImpl is std::string in the encoding we want + operator const wxStdWideString&() const { return m_impl; } + #else + // wxStringImpl is either not std::string or needs conversion + operator wxStdWideString() const + // FIXME-UTF8: broken for embedded NULs + { return wxStdWideString(wc_str()); } + #endif + + #if (!wxUSE_UNICODE || wxUSE_UTF8_LOCALE_ONLY) && wxUSE_STL_BASED_WXSTRING + // wxStringImpl is std::string in the encoding we want + operator const std::string&() const { return m_impl; } + #else + // wxStringImpl is either not std::string or needs conversion + operator std::string() const + // FIXME-UTF8: broken for embedded NULs + { return std::string(mb_str()); } + #endif +#endif // wxUSE_STL + + // first valid index position + const_iterator begin() const { return const_iterator(this, m_impl.begin()); } + iterator begin() { return iterator(this, m_impl.begin()); } + // position one after the last valid one + const_iterator end() const { return const_iterator(this, m_impl.end()); } + iterator end() { return iterator(this, m_impl.end()); } + + // first element of the reversed string + const_reverse_iterator rbegin() const + { return const_reverse_iterator(end()); } + reverse_iterator rbegin() + { return reverse_iterator(end()); } + // one beyond the end of the reversed string + const_reverse_iterator rend() const + { return const_reverse_iterator(begin()); } + reverse_iterator rend() + { return reverse_iterator(begin()); } + + // std::string methods: +#if wxUSE_UNICODE_UTF8 + size_t length() const { return end() - begin(); } // FIXME-UTF8: optimize! #else - const char* AsChar() const; - operator const char*() const { return AsChar(); } + size_t length() const { return m_impl.length(); } #endif - wxString AsString() const; - operator wxString() const; - - // allow expressions like "c_str()[0]": - wxUniChar operator[](int n) const { return operator[](size_t(n)); } - wxUniChar operator[](size_t n) const; -#ifndef wxSIZE_T_IS_UINT - wxUniChar operator[](unsigned int n) const { return operator[](size_t(n)); } -#endif // size_t != unsigned int - - // these operators are needed to emulate the pointer semantics of c_str(): - // expressions like "wxChar *p = str.c_str() + 1;" should continue to work - // (we need both versions to resolve ambiguities): - wxCStrData operator+(int n) const - { return wxCStrData(m_str, m_offset + n, m_owned); } - wxCStrData operator+(long n) const - { return wxCStrData(m_str, m_offset + n, m_owned); } - wxCStrData operator+(size_t n) const - { return wxCStrData(m_str, m_offset + n, m_owned); } - - // this operator is need to make expressions like "*c_str()" or - // "*(c_str() + 2)" work - wxUniChar operator*() const; - -private: - const wxString *m_str; - size_t m_offset; - bool m_owned; - - friend class WXDLLIMPEXP_BASE wxString; -}; - -// ---------------------------------------------------------------------------- -// wxStringPrintfMixin -// --------------------------------------------------------------------------- - -// NB: VC6 has a bug that causes linker errors if you have template methods -// in a class using __declspec(dllimport). The solution is to split such -// class into two classes, one that contains the template methods and does -// *not* use WXDLLIMPEXP_BASE and another class that contains the rest -// (with DLL linkage). -// -// We only do this for VC6 here, because the code is less efficient -// (Printf() has to use dynamic_cast<>) and because OpenWatcom compiler -// cannot compile this code. + size_type size() const { return length(); } + size_type max_size() const { return npos; } -#if defined(__VISUALC__) && __VISUALC__ < 1300 - #define wxNEEDS_WXSTRING_PRINTF_MIXIN -#endif + bool empty() const { return m_impl.empty(); } -#ifdef wxNEEDS_WXSTRING_PRINTF_MIXIN -// this class contains implementation of wxString's vararg methods, it's -// exported from wxBase DLL -class WXDLLIMPEXP_BASE wxStringPrintfMixinBase -{ -protected: - wxStringPrintfMixinBase() {} + size_type capacity() const { return m_impl.capacity(); } // FIXME-UTF8 + void reserve(size_t sz) { m_impl.reserve(sz); } // FIXME-UTF8 - int DoPrintf(const wxChar *format, ...) ATTRIBUTE_PRINTF_2; - static wxString DoFormat(const wxChar *format, ...) ATTRIBUTE_PRINTF_1; -}; + void resize(size_t nSize, wxUniChar ch = wxT('\0')) + { + const size_t len = length(); + if ( nSize == len) + return; -// this class contains template wrappers for wxString's vararg methods, it's -// intentionally *not* exported from the DLL in order to fix the VC6 bug -// described above -class wxStringPrintfMixin : public wxStringPrintfMixinBase -{ -private: - // to further complicate things, we can't return wxString from - // wxStringPrintfMixin::Format() because wxString is not yet declared at - // this point; the solution is to use this fake type trait template - this - // way the compiler won't know the return type until Format() is used - // (this doesn't compile with Watcom, but VC6 compiles it just fine): - template struct StringReturnType +#if wxUSE_UNICODE_UTF8 + if ( nSize < len ) { - typedef wxString type; - }; - -public: - // these are duplicated wxString methods, they're also declared below - // if !wxNEEDS_WXSTRING_PRINTF_MIXIN: - - // int Printf(const wxChar *pszFormat, ...); - WX_DEFINE_VARARG_FUNC(int, Printf, DoPrintf) - // static wxString Format(const wxChar *pszFormat, ...) ATTRIBUTE_PRINTF_1; - WX_DEFINE_VARARG_FUNC(static typename StringReturnType::type, - Format, DoFormat) - // int sprintf(const wxChar *pszFormat, ...) ATTRIBUTE_PRINTF_2; - WX_DEFINE_VARARG_FUNC(int, sprintf, DoPrintf) - -protected: - wxStringPrintfMixin() : wxStringPrintfMixinBase() {} -}; -#endif // wxNEEDS_WXSTRING_PRINTF_MIXIN - - -// ---------------------------------------------------------------------------- -// wxString: string class trying to be compatible with std::string, MFC -// CString and wxWindows 1.x wxString all at once -// --------------------------------------------------------------------------- - -#ifdef wxNEEDS_WXSTRING_PRINTF_MIXIN - // "non dll-interface class 'wxStringPrintfMixin' used as base interface - // for dll-interface class 'wxString'" -- this is OK in our case - #pragma warning (disable:4275) -#endif - -class WXDLLIMPEXP_BASE wxString : public wxStringBase -#ifdef wxNEEDS_WXSTRING_PRINTF_MIXIN - ,public wxStringPrintfMixin -#endif -{ - // NB: special care was taken in arranging the member functions in such order - // that all inline functions can be effectively inlined, verify that all - // performance critical functions are still inlined if you change order! -private: - // if we hadn't made these operators private, it would be possible to - // compile "wxString s; s = 17;" without any warnings as 17 is implicitly - // converted to char in C and we do have operator=(char) - // - // NB: we don't need other versions (short/long and unsigned) as attempt - // to assign another numeric type to wxString will now result in - // ambiguity between operator=(char) and operator=(int) - wxString& operator=(int); - - // these methods are not implemented - there is _no_ conversion from int to - // string, you're doing something wrong if the compiler wants to call it! - // - // try `s << i' or `s.Printf("%d", i)' instead - wxString(int); - -public: - // constructors and destructor - // ctor for an empty string - wxString() : wxStringBase() { } - // copy ctor - wxString(const wxStringBase& stringSrc) : wxStringBase(stringSrc) { } - wxString(const wxString& stringSrc) : wxStringBase(stringSrc) { } - // string containing nRepeat copies of ch - wxString(wxUniChar ch, size_t nRepeat = 1) - : wxStringBase(nRepeat, ch) { } - wxString(size_t nRepeat, wxUniChar ch) - : wxStringBase(nRepeat, ch) { } - wxString(wxUniCharRef ch, size_t nRepeat = 1) - : wxStringBase(nRepeat, ch) { } - wxString(size_t nRepeat, wxUniCharRef ch) - : wxStringBase(nRepeat, ch) { } - wxString(char ch, size_t nRepeat = 1) - : wxStringBase(nRepeat, ch) { } - wxString(size_t nRepeat, char ch) - : wxStringBase(nRepeat, ch) { } - wxString(wchar_t ch, size_t nRepeat = 1) - : wxStringBase(nRepeat, ch) { } - wxString(size_t nRepeat, wchar_t ch) - : wxStringBase(nRepeat, ch) { } - // ctor takes first nLength characters from C string - // (default value of npos means take all the string) - wxString(const wxChar *psz) - : wxStringBase(psz ? psz : wxT("")) { } - wxString(const wxChar *psz, size_t nLength) - : wxStringBase(psz, nLength) { } - wxString(const wxChar *psz, - const wxMBConv& WXUNUSED(conv), - size_t nLength = npos) - : wxStringBase(psz, nLength == npos ? wxStrlen(psz) : nLength) { } - - // even if we're not built with wxUSE_STL == 1 it is very convenient to allow - // implicit conversions from std::string to wxString as this allows to use - // the same strings in non-GUI and GUI code, however we don't want to - // unconditionally add this ctor as it would make wx lib dependent on - // libstdc++ on some Linux versions which is bad, so instead we ask the - // client code to define this wxUSE_STD_STRING symbol if they need it -#if wxUSE_STD_STRING - wxString(const wxStdString& s) - : wxStringBase(s.c_str()) { } -#endif // wxUSE_STD_STRING - -#if wxUSE_UNICODE - // from multibyte string - wxString(const char *psz, - const wxMBConv& conv = wxConvLibc, - size_t nLength = npos); - // from wxWCharBuffer (i.e. return from wxGetString) - wxString(const wxWCharBuffer& psz) : wxStringBase(psz.data()) { } -#else // ANSI - // from C string (for compilers using unsigned char) - wxString(const unsigned char* psz) - : wxStringBase((const char*)psz) { } - // from part of C string (for compilers using unsigned char) - wxString(const unsigned char* psz, size_t nLength) - : wxStringBase((const char*)psz, nLength) { } + // we can't use wxStringImpl::resize() for truncating the string as it + // counts in bytes, not characters + erase(nSize); + return; + } -#if wxUSE_WCHAR_T - // from wide (Unicode) string - wxString(const wchar_t *pwz, - const wxMBConv& conv = wxConvLibc, - size_t nLength = npos); -#endif // !wxUSE_WCHAR_T + // we also can't use (presumably more efficient) resize() if we have to + // append characters taking more than one byte + if ( !ch.IsAscii() ) + append(nSize - len, ch); + else +#endif // wxUSE_UNICODE_UTF8 + m_impl.resize(nSize, (wxStringCharType)ch); + } - // from wxCharBuffer - wxString(const wxCharBuffer& psz) - : wxStringBase(psz) { } -#endif // Unicode/ANSI + wxString substr(size_t nStart = 0, size_t nLen = npos) const + { + size_t pos, len; + PosLenToImpl(nStart, nLen, &pos, &len); + return FromImpl(m_impl.substr(pos, len)); + } // generic attributes & operations // as standard strlen() @@ -1016,9 +1150,13 @@ public: // data access (all indexes are 0 based) // read access + wxUniChar at(size_t n) const + { return *(begin() + n); } // FIXME-UTF8: optimize? wxUniChar GetChar(size_t n) const { return at(n); } // read/write access + wxUniCharRef at(size_t n) + { return *(begin() + n); } // FIXME-UTF8: optimize? wxUniCharRef GetWritableChar(size_t n) { return at(n); } // write access @@ -1026,56 +1164,81 @@ public: { at(n) = ch; } // get last character - wxUniChar Last() const - { - wxASSERT_MSG( !empty(), _T("wxString: index out of bounds") ); - - return at(length() - 1); - } + wxUniChar Last() const + { + wxASSERT_MSG( !empty(), _T("wxString: index out of bounds") ); + return *rbegin(); + } // get writable last character wxUniCharRef Last() - { - wxASSERT_MSG( !empty(), _T("wxString: index out of bounds") ); - return at(length() - 1); - } + { + wxASSERT_MSG( !empty(), _T("wxString: index out of bounds") ); + return *rbegin(); + } /* Note that we we must define all of the overloads below to avoid ambiguity when using str[0]. */ wxUniChar operator[](int n) const - { return wxStringBase::at(n); } + { return at(n); } + wxUniChar operator[](long n) const + { return at(n); } wxUniChar operator[](size_t n) const - { return wxStringBase::at(n); } + { return at(n); } #ifndef wxSIZE_T_IS_UINT wxUniChar operator[](unsigned int n) const - { return wxStringBase::at(n); } + { return at(n); } #endif // size_t != unsigned int // operator versions of GetWriteableChar() wxUniCharRef operator[](int n) - { return wxStringBase::at(n); } + { return at(n); } + wxUniCharRef operator[](long n) + { return at(n); } wxUniCharRef operator[](size_t n) - { return wxStringBase::at(n); } + { return at(n); } #ifndef wxSIZE_T_IS_UINT wxUniCharRef operator[](unsigned int n) - { return wxStringBase::at(n); } + { return at(n); } #endif // size_t != unsigned int // explicit conversion to C string (use this with printf()!) wxCStrData c_str() const { return wxCStrData(this); } + wxCStrData data() const { return c_str(); } // implicit conversion to C string operator wxCStrData() const { return c_str(); } - operator const wxChar*() const { return c_str(); } + + // the first two operators conflict with operators for conversion to + // std::string and they must be disabled in STL build; the next one only + // makes sense if conversions to char* are also defined and not defining it + // in STL build also helps us to get more clear error messages for the code + // which relies on implicit conversion to char* in STL build +#if !wxUSE_STL + operator const char*() const { return c_str(); } + operator const wchar_t*() const { return c_str(); } + + // implicit conversion to untyped pointer for compatibility with previous + // wxWidgets versions: this is the same as conversion to const char * so it + // may fail! + operator const void*() const { return c_str(); } +#endif // wxUSE_STL // identical to c_str(), for MFC compatibility const wxCStrData GetData() const { return c_str(); } // explicit conversion to C string in internal representation (char*, // wchar_t*, UTF-8-encoded char*, depending on the build): - const_pointer wx_str() const { return data(); } + const wxStringCharType *wx_str() const { return m_impl.c_str(); } + + // conversion to *non-const* multibyte or widestring buffer; modifying + // returned buffer won't affect the string, these methods are only useful + // for passing values to const-incorrect functions + wxWritableCharBuffer char_str(const wxMBConv& conv = wxConvLibc) const + { return mb_str(conv); } + wxWritableWCharBuffer wchar_str() const { return wc_str(); } // conversion to/from plain (i.e. 7 bit) ASCII: this is useful for // converting numbers or strings which are certain not to contain special @@ -1084,15 +1247,85 @@ public: // the behaviour of these functions with the strings containing anything // else than 7 bit ASCII characters is undefined, use at your own risk. #if wxUSE_UNICODE - static wxString FromAscii(const char *ascii); // string - static wxString FromAscii(const char ascii); // char + static wxString FromAscii(const char *ascii, size_t len); + static wxString FromAscii(const char *ascii); + static wxString FromAscii(char ascii); const wxCharBuffer ToAscii() const; #else // ANSI static wxString FromAscii(const char *ascii) { return wxString( ascii ); } - static wxString FromAscii(const char ascii) { return wxString( ascii ); } + static wxString FromAscii(const char *ascii, size_t len) + { return wxString( ascii, len ); } + static wxString FromAscii(char ascii) { return wxString( ascii ); } const char *ToAscii() const { return c_str(); } #endif // Unicode/!Unicode + // also provide unsigned char overloads as signed/unsigned doesn't matter + // for 7 bit ASCII characters + static wxString FromAscii(const unsigned char *ascii) + { return FromAscii((const char *)ascii); } + static wxString FromAscii(const unsigned char *ascii, size_t len) + { return FromAscii((const char *)ascii, len); } + + // conversion to/from UTF-8: +#if wxUSE_UNICODE_UTF8 + static wxString FromUTF8(const char *utf8) + { + if ( !utf8 ) + return wxEmptyString; + + wxASSERT( wxStringOperations::IsValidUtf8String(utf8) ); + return FromImpl(wxStringImpl(utf8)); + } + static wxString FromUTF8(const char *utf8, size_t len) + { + if ( !utf8 ) + return wxEmptyString; + if ( len == npos ) + return FromUTF8(utf8); + + wxASSERT( wxStringOperations::IsValidUtf8String(utf8, len) ); + return FromImpl(wxStringImpl(utf8, len)); + } + const char* utf8_str() const { return wx_str(); } + const char* ToUTF8() const { return wx_str(); } +#elif wxUSE_UNICODE_WCHAR + static wxString FromUTF8(const char *utf8) + { return wxString(utf8, wxMBConvUTF8()); } + static wxString FromUTF8(const char *utf8, size_t len) + { return wxString(utf8, wxMBConvUTF8(), len); } + const wxCharBuffer utf8_str() const { return mb_str(wxMBConvUTF8()); } + const wxCharBuffer ToUTF8() const { return utf8_str(); } +#else // ANSI + static wxString FromUTF8(const char *utf8) + { return wxString(wxMBConvUTF8().cMB2WC(utf8)); } + static wxString FromUTF8(const char *utf8, size_t len) + { + size_t wlen; + wxWCharBuffer buf(wxMBConvUTF8().cMB2WC(utf8, len == npos ? wxNO_LEN : len, &wlen)); + return wxString(buf.data(), wlen); + } + const wxCharBuffer utf8_str() const + { return wxMBConvUTF8().cWC2MB(wc_str()); } + const wxCharBuffer ToUTF8() const { return utf8_str(); } +#endif + + // functions for storing binary data in wxString: +#if wxUSE_UNICODE + static wxString From8BitData(const char *data, size_t len) + { return wxString(data, wxConvISO8859_1, len); } + // version for NUL-terminated data: + static wxString From8BitData(const char *data) + { return wxString(data, wxConvISO8859_1); } + const wxCharBuffer To8BitData() const { return mb_str(wxConvISO8859_1); } +#else // ANSI + static wxString From8BitData(const char *data, size_t len) + { return wxString(data, len); } + // version for NUL-terminated data: + static wxString From8BitData(const char *data) + { return wxString(data); } + const char *To8BitData() const { return c_str(); } +#endif // Unicode/ANSI + // conversions with (possible) format conversions: have to return a // buffer with temporary data // @@ -1103,79 +1336,84 @@ public: // type differs because a function may either return pointer to the buffer // directly or have to use intermediate buffer for translation. #if wxUSE_UNICODE + +#if wxUSE_UTF8_LOCALE_ONLY + const char* mb_str() const { return wx_str(); } + const wxCharBuffer mb_str(const wxMBConv& conv) const; +#else const wxCharBuffer mb_str(const wxMBConv& conv = wxConvLibc) const; +#endif const wxWX2MBbuf mbc_str() const { return mb_str(*wxConvCurrent); } - const wxChar* wc_str() const { return c_str(); } - +#if wxUSE_UNICODE_WCHAR + const wxChar* wc_str() const { return wx_str(); } +#elif wxUSE_UNICODE_UTF8 + const wxWCharBuffer wc_str() const; +#endif // for compatibility with !wxUSE_UNICODE version - const wxChar* wc_str(const wxMBConv& WXUNUSED(conv)) const { return c_str(); } + const wxWX2WCbuf wc_str(const wxMBConv& WXUNUSED(conv)) const + { return wc_str(); } #if wxMBFILES const wxCharBuffer fn_str() const { return mb_str(wxConvFile); } #else // !wxMBFILES - const wxChar* fn_str() const { return c_str(); } + const wxWX2WCbuf fn_str() const { return wc_str(); } #endif // wxMBFILES/!wxMBFILES + #else // ANSI - const wxChar* mb_str() const { return c_str(); } + const wxChar* mb_str() const { return wx_str(); } // for compatibility with wxUSE_UNICODE version - const wxChar* mb_str(const wxMBConv& WXUNUSED(conv)) const { return c_str(); } + const wxChar* mb_str(const wxMBConv& WXUNUSED(conv)) const { return wx_str(); } const wxWX2MBbuf mbc_str() const { return mb_str(); } #if wxUSE_WCHAR_T - const wxWCharBuffer wc_str(const wxMBConv& conv) const; + const wxWCharBuffer wc_str(const wxMBConv& conv = wxConvLibc) const; #endif // wxUSE_WCHAR_T -#ifdef __WXOSX__ const wxCharBuffer fn_str() const { return wxConvFile.cWC2WX( wc_str( wxConvLocal ) ); } -#else - const wxChar* fn_str() const { return c_str(); } -#endif #endif // Unicode/ANSI // overloaded assignment // from another wxString - wxString& operator=(const wxStringBase& stringSrc) - { return (wxString&)wxStringBase::operator=(stringSrc); } + wxString& operator=(const wxString& stringSrc) + { m_impl = stringSrc.m_impl; return *this; } + wxString& operator=(const wxCStrData& cstr) + { return *this = cstr.AsString(); } // from a character wxString& operator=(wxUniChar ch) - { return (wxString&)wxStringBase::operator=(ch); } + { m_impl = wxStringOperations::EncodeChar(ch); return *this; } wxString& operator=(wxUniCharRef ch) - { return (wxString&)wxStringBase::operator=((wxUniChar)ch); } + { return operator=((wxUniChar)ch); } wxString& operator=(char ch) - { return (wxString&)wxStringBase::operator=(wxUniChar(ch)); } + { return operator=(wxUniChar(ch)); } + wxString& operator=(unsigned char ch) + { return operator=(wxUniChar(ch)); } wxString& operator=(wchar_t ch) - { return (wxString&)wxStringBase::operator=(wxUniChar(ch)); } + { return operator=(wxUniChar(ch)); } // from a C string - STL probably will crash on NULL, // so we need to compensate in that case -#if wxUSE_STL - wxString& operator=(const wxChar *psz) - { if(psz) wxStringBase::operator=(psz); else Clear(); return *this; } +#if wxUSE_STL_BASED_WXSTRING + wxString& operator=(const char *psz) + { if (psz) m_impl = ImplStr(psz); else Clear(); return *this; } + wxString& operator=(const wchar_t *pwz) + { if (pwz) m_impl = ImplStr(pwz); else Clear(); return *this; } #else - wxString& operator=(const wxChar *psz) - { return (wxString&)wxStringBase::operator=(psz); } + wxString& operator=(const char *psz) + { m_impl = ImplStr(psz); return *this; } + wxString& operator=(const wchar_t *pwz) + { m_impl = ImplStr(pwz); return *this; } #endif + wxString& operator=(const unsigned char *psz) + { return operator=((const char*)psz); } -#if wxUSE_UNICODE // from wxWCharBuffer - wxString& operator=(const wxWCharBuffer& psz) - { (void) operator=((const wchar_t *)psz); return *this; } - // from C string - wxString& operator=(const char* psz) - { return operator=(wxString(psz)); } -#else // ANSI - // from another kind of C string - wxString& operator=(const unsigned char* psz); -#if wxUSE_WCHAR_T - // from a wide string - wxString& operator=(const wchar_t *pwz); -#endif + wxString& operator=(const wxWCharBuffer& s) + { return operator=(s.data()); } // FIXME-UTF8: fix for embedded NULs // from wxCharBuffer - wxString& operator=(const wxCharBuffer& psz) - { (void) operator=((const char *)psz); return *this; } -#endif // Unicode/ANSI + wxString& operator=(const wxCharBuffer& s) + { return operator=(s.data()); } // FIXME-UTF8: fix for embedded NULs // string concatenation // in place concatenation @@ -1187,8 +1425,8 @@ public: // string += string wxString& operator<<(const wxString& s) { -#if WXWIN_COMPATIBILITY_2_8 && !wxUSE_STL - wxASSERT_MSG( s.GetStringData()->IsValid(), +#if WXWIN_COMPATIBILITY_2_8 && !wxUSE_STL_BASED_WXSTRING && !wxUSE_UNICODE_UTF8 + wxASSERT_MSG( s.IsValid(), _T("did you forget to call UngetWriteBuf()?") ); #endif @@ -1196,36 +1434,24 @@ public: return *this; } // string += C string - wxString& operator<<(const wxChar *psz) + wxString& operator<<(const char *psz) { append(psz); return *this; } + wxString& operator<<(const wchar_t *pwz) + { append(pwz); return *this; } wxString& operator<<(const wxCStrData& psz) - { append(psz); return *this; } + { append(psz.AsString()); return *this; } // string += char wxString& operator<<(wxUniChar ch) { append(1, ch); return *this; } wxString& operator<<(wxUniCharRef ch) { append(1, ch); return *this; } wxString& operator<<(char ch) { append(1, ch); return *this; } + wxString& operator<<(unsigned char ch) { append(1, ch); return *this; } wxString& operator<<(wchar_t ch) { append(1, ch); return *this; } // string += buffer (i.e. from wxGetString) -#if wxUSE_UNICODE wxString& operator<<(const wxWCharBuffer& s) { return operator<<((const wchar_t *)s); } - wxString& operator+=(const wxWCharBuffer& s) - { return operator<<((const wchar_t *)s); } -#else // !wxUSE_UNICODE wxString& operator<<(const wxCharBuffer& s) { return operator<<((const char *)s); } - wxString& operator+=(const wxCharBuffer& s) - { return operator<<((const char *)s); } -#endif // wxUSE_UNICODE/!wxUSE_UNICODE - -#if wxUSE_UNICODE - // string += C string in Unicode build (with conversion) - wxString& operator<<(const char *s) - { return operator<<(wxString(s)); } - wxString& operator+=(const char *s) - { return operator+=(wxString(s)); } -#endif // wxUSE_UNICODE // string += C string wxString& Append(const wxString& s) @@ -1237,9 +1463,15 @@ public: append(s); return *this; } + wxString& Append(const char* psz) + { append(psz); return *this; } + wxString& Append(const wchar_t* pwz) + { append(pwz); return *this; } wxString& Append(const wxCStrData& psz) { append(psz); return *this; } - wxString& Append(const wxChar* psz) + wxString& Append(const wxCharBuffer& psz) + { append(psz); return *this; } + wxString& Append(const wxWCharBuffer& psz) { append(psz); return *this; } // append count copies of given character wxString& Append(wxUniChar ch, size_t count = 1u) @@ -1248,10 +1480,14 @@ public: { append(count, ch); return *this; } wxString& Append(char ch, size_t count = 1u) { append(count, ch); return *this; } + wxString& Append(unsigned char ch, size_t count = 1u) + { append(count, ch); return *this; } wxString& Append(wchar_t ch, size_t count = 1u) { append(count, ch); return *this; } - wxString& Append(const wxChar* psz, size_t nLen) + wxString& Append(const char* psz, size_t nLen) { append(psz, nLen); return *this; } + wxString& Append(const wchar_t* pwz, size_t nLen) + { append(pwz, nLen); return *this; } // prepend a string, return the string itself wxString& Prepend(const wxString& str) @@ -1267,9 +1503,13 @@ public: friend wxString WXDLLIMPEXP_BASE operator+(wxUniChar ch, const wxString& string); // string with C string friend wxString WXDLLIMPEXP_BASE operator+(const wxString& string, - const wxChar *psz); + const char *psz); + friend wxString WXDLLIMPEXP_BASE operator+(const wxString& string, + const wchar_t *pwz); // C string with string - friend wxString WXDLLIMPEXP_BASE operator+(const wxChar *psz, + friend wxString WXDLLIMPEXP_BASE operator+(const char *psz, + const wxString& string); + friend wxString WXDLLIMPEXP_BASE operator+(const wchar_t *pwz, const wxString& string); // stream-like functions @@ -1308,21 +1548,47 @@ public: // string comparison // case-sensitive comparison (returns a value < 0, = 0 or > 0) - int Cmp(const wxChar *psz) const; - int Cmp(const wxString& s) const; + int Cmp(const char *psz) const + { return compare(psz); } + int Cmp(const wchar_t *pwz) const + { return compare(pwz); } + int Cmp(const wxString& s) const + { return compare(s); } + int Cmp(const wxCStrData& s) const + { return compare(s); } + int Cmp(const wxCharBuffer& s) const + { return compare(s); } + int Cmp(const wxWCharBuffer& s) const + { return compare(s); } // same as Cmp() but not case-sensitive - int CmpNoCase(const wxChar *psz) const; int CmpNoCase(const wxString& s) const; // test for the string equality, either considering case or not // (if compareWithCase then the case matters) - bool IsSameAs(const wxChar *psz, bool compareWithCase = true) const - { return (compareWithCase ? Cmp(psz) : CmpNoCase(psz)) == 0; } + bool IsSameAs(const wxString& str, bool compareWithCase = true) const + { return (compareWithCase ? Cmp(str) : CmpNoCase(str)) == 0; } + bool IsSameAs(const char *str, bool compareWithCase = true) const + { return (compareWithCase ? Cmp(str) : CmpNoCase(str)) == 0; } + bool IsSameAs(const wchar_t *str, bool compareWithCase = true) const + { return (compareWithCase ? Cmp(str) : CmpNoCase(str)) == 0; } + bool IsSameAs(const wxCStrData& str, bool compareWithCase = true) const + { return IsSameAs(str.AsString(), compareWithCase); } + bool IsSameAs(const wxCharBuffer& str, bool compareWithCase = true) const + { return IsSameAs(str.data(), compareWithCase); } + bool IsSameAs(const wxWCharBuffer& str, bool compareWithCase = true) const + { return IsSameAs(str.data(), compareWithCase); } // comparison with a single character: returns true if equal - bool IsSameAs(wxUniChar c, bool compareWithCase = true) const - { - return (length() == 1) && (compareWithCase ? GetChar(0u) == c - : wxToupper(GetChar(0u)) == wxToupper(c)); - } + bool IsSameAs(wxUniChar c, bool compareWithCase = true) const; + // FIXME-UTF8: remove these overloads + bool IsSameAs(wxUniCharRef c, bool compareWithCase = true) const + { return IsSameAs(wxUniChar(c), compareWithCase); } + bool IsSameAs(char c, bool compareWithCase = true) const + { return IsSameAs(wxUniChar(c), compareWithCase); } + bool IsSameAs(unsigned char c, bool compareWithCase = true) const + { return IsSameAs(wxUniChar(c), compareWithCase); } + bool IsSameAs(wchar_t c, bool compareWithCase = true) const + { return IsSameAs(wxUniChar(c), compareWithCase); } + bool IsSameAs(int c, bool compareWithCase = true) const + { return IsSameAs(wxUniChar(c), compareWithCase); } // simple sub-string extraction // return substring starting at nFirst of length nCount (or till the end @@ -1336,11 +1602,11 @@ public: // check if the string starts with the given prefix and return the rest // of the string in the provided pointer if it is not NULL; otherwise // return false - bool StartsWith(const wxChar *prefix, wxString *rest = NULL) const; + bool StartsWith(const wxString& prefix, wxString *rest = NULL) const; // check if the string ends with the given suffix and return the // beginning of the string before the suffix in the provided pointer if // it is not NULL; otherwise return false - bool EndsWith(const wxChar *suffix, wxString *rest = NULL) const; + bool EndsWith(const wxString& suffix, wxString *rest = NULL) const; // get first nCount characters wxString Left(size_t nCount) const; @@ -1383,16 +1649,46 @@ public: // searching and replacing // searching (return starting index, or -1 if not found) int Find(wxUniChar ch, bool bFromEnd = false) const; // like strchr/strrchr + int Find(wxUniCharRef ch, bool bFromEnd = false) const + { return Find(wxUniChar(ch), bFromEnd); } + int Find(char ch, bool bFromEnd = false) const + { return Find(wxUniChar(ch), bFromEnd); } + int Find(unsigned char ch, bool bFromEnd = false) const + { return Find(wxUniChar(ch), bFromEnd); } + int Find(wchar_t ch, bool bFromEnd = false) const + { return Find(wxUniChar(ch), bFromEnd); } // searching (return starting index, or -1 if not found) - int Find(const wxChar *pszSub) const; // like strstr + int Find(const wxString& sub) const // like strstr + { + size_type idx = find(sub); + return (idx == npos) ? wxNOT_FOUND : (int)idx; + } + int Find(const char *sub) const // like strstr + { + size_type idx = find(sub); + return (idx == npos) ? wxNOT_FOUND : (int)idx; + } + int Find(const wchar_t *sub) const // like strstr + { + size_type idx = find(sub); + return (idx == npos) ? wxNOT_FOUND : (int)idx; + } + + int Find(const wxCStrData& sub) const + { return Find(sub.AsString()); } + int Find(const wxCharBuffer& sub) const + { return Find(sub.data()); } + int Find(const wxWCharBuffer& sub) const + { return Find(sub.data()); } + // replace first (or all of bReplaceAll) occurences of substring with // another string, returns the number of replacements made - size_t Replace(const wxChar *szOld, - const wxChar *szNew, + size_t Replace(const wxString& strOld, + const wxString& strNew, bool bReplaceAll = true); // check if the string contents matches a mask containing '*' and '?' - bool Matches(const wxChar *szMask) const; + bool Matches(const wxString& mask) const; // conversion to numbers: all functions return true only if the whole // string is a number and put the value of this number into the pointer @@ -1417,16 +1713,40 @@ public: // formatted input/output // as sprintf(), returns the number of characters written or < 0 on error // (take 'this' into account in attribute parameter count) - // int Printf(const wxChar *pszFormat, ...); - WX_DEFINE_VARARG_FUNC(int, Printf, DoPrintf) + // int Printf(const wxString& format, ...); + WX_DEFINE_VARARG_FUNC(int, Printf, 1, (const wxFormatString&), + DoPrintfWchar, DoPrintfUtf8) +#ifdef __WATCOMC__ + // workaround for http://bugzilla.openwatcom.org/show_bug.cgi?id=351 + WX_VARARG_WATCOM_WORKAROUND(int, Printf, 1, (const wxString&), + (wxFormatString(f1))); + WX_VARARG_WATCOM_WORKAROUND(int, Printf, 1, (const wxCStrData&), + (wxFormatString(f1))); + WX_VARARG_WATCOM_WORKAROUND(int, Printf, 1, (const char*), + (wxFormatString(f1))); + WX_VARARG_WATCOM_WORKAROUND(int, Printf, 1, (const wchar_t*), + (wxFormatString(f1))); +#endif #endif // !wxNEEDS_WXSTRING_PRINTF_MIXIN // as vprintf(), returns the number of characters written or < 0 on error int PrintfV(const wxString& format, va_list argptr); #ifndef wxNEEDS_WXSTRING_PRINTF_MIXIN // returns the string containing the result of Printf() to it - // static wxString Format(const wxChar *pszFormat, ...) ATTRIBUTE_PRINTF_1; - WX_DEFINE_VARARG_FUNC(static wxString, Format, DoFormat) + // static wxString Format(const wxString& format, ...) ATTRIBUTE_PRINTF_1; + WX_DEFINE_VARARG_FUNC(static wxString, Format, 1, (const wxFormatString&), + DoFormatWchar, DoFormatUtf8) +#ifdef __WATCOMC__ + // workaround for http://bugzilla.openwatcom.org/show_bug.cgi?id=351 + WX_VARARG_WATCOM_WORKAROUND(static wxString, Format, 1, (const wxString&), + (wxFormatString(f1))); + WX_VARARG_WATCOM_WORKAROUND(static wxString, Format, 1, (const wxCStrData&), + (wxFormatString(f1))); + WX_VARARG_WATCOM_WORKAROUND(static wxString, Format, 1, (const char*), + (wxFormatString(f1))); + WX_VARARG_WATCOM_WORKAROUND(static wxString, Format, 1, (const wchar_t*), + (wxFormatString(f1))); +#endif #endif // the same as above, but takes a va_list static wxString FormatV(const wxString& format, va_list argptr); @@ -1438,16 +1758,16 @@ public: // minimize the string's memory // only works if the data of this string is not shared bool Shrink(); -#if WXWIN_COMPATIBILITY_2_8 && !wxUSE_STL +#if WXWIN_COMPATIBILITY_2_8 && !wxUSE_STL_BASED_WXSTRING && !wxUSE_UNICODE_UTF8 // These are deprecated, use wxStringBuffer or wxStringBufferLength instead // // get writable buffer of at least nLen bytes. Unget() *must* be called // a.s.a.p. to put string back in a reasonable state! - wxDEPRECATED( wxChar *GetWriteBuf(size_t nLen) ); + wxDEPRECATED( wxStringCharType *GetWriteBuf(size_t nLen) ); // call this immediately after GetWriteBuf() has been used wxDEPRECATED( void UngetWriteBuf() ); wxDEPRECATED( void UngetWriteBuf(size_t nLen) ); -#endif // WXWIN_COMPATIBILITY_2_8 && !wxUSE_STL +#endif // WXWIN_COMPATIBILITY_2_8 && !wxUSE_STL_BASED_WXSTRING && wxUSE_UNICODE_UTF8 // wxWidgets version 1 compatibility functions @@ -1462,15 +1782,27 @@ public: #ifndef wxNEEDS_WXSTRING_PRINTF_MIXIN // use Printf() // (take 'this' into account in attribute parameter count) - // int sprintf(const wxChar *pszFormat, ...) ATTRIBUTE_PRINTF_2; - WX_DEFINE_VARARG_FUNC(int, sprintf, DoPrintf) + // int sprintf(const wxString& format, ...) ATTRIBUTE_PRINTF_2; + WX_DEFINE_VARARG_FUNC(int, sprintf, 1, (const wxFormatString&), + DoPrintfWchar, DoPrintfUtf8) +#ifdef __WATCOMC__ + // workaround for http://bugzilla.openwatcom.org/show_bug.cgi?id=351 + WX_VARARG_WATCOM_WORKAROUND(int, sprintf, 1, (const wxString&), + (wxFormatString(f1))); + WX_VARARG_WATCOM_WORKAROUND(int, sprintf, 1, (const wxCStrData&), + (wxFormatString(f1))); + WX_VARARG_WATCOM_WORKAROUND(int, sprintf, 1, (const char*), + (wxFormatString(f1))); + WX_VARARG_WATCOM_WORKAROUND(int, sprintf, 1, (const wchar_t*), + (wxFormatString(f1))); +#endif #endif // wxNEEDS_WXSTRING_PRINTF_MIXIN // use Cmp() inline int CompareTo(const wxChar* psz, caseCompare cmp = exact) const { return cmp == exact ? Cmp(psz) : CmpNoCase(psz); } - // use Len + // use length() size_t Length() const { return length(); } // Count the number of characters int Freq(wxUniChar ch) const; @@ -1492,12 +1824,13 @@ public: { return (wxString&)erase( nStart, nLen ); } // use Find() - int First( const wxUniChar ch ) const { return Find(ch); } + int First( wxUniChar ch ) const { return Find(ch); } + int First( wxUniCharRef ch ) const { return Find(ch); } int First( char ch ) const { return Find(ch); } + int First( unsigned char ch ) const { return Find(ch); } int First( wchar_t ch ) const { return Find(ch); } - int First( const wxChar* psz ) const { return Find(psz); } - int First( const wxString &str ) const { return Find(str); } - int Last( const wxUniChar ch ) const { return Find(ch, true); } + int First( const wxString& str ) const { return Find(str); } + int Last( wxUniChar ch ) const { return Find(ch, true); } bool Contains(const wxString& str) const { return Find(str) != wxNOT_FOUND; } // use empty() @@ -1507,176 +1840,849 @@ public: // take nLen chars starting at nPos wxString(const wxString& str, size_t nPos, size_t nLen) - : wxStringBase(str, nPos, nLen) { } - // take all characters from pStart to pEnd - wxString(const void *pStart, const void *pEnd) - : wxStringBase((const wxChar*)pStart, (const wxChar*)pEnd) { } + { assign(str, nPos, nLen); } + // take all characters from first to last wxString(const_iterator first, const_iterator last) - : wxStringBase(first, last) { } - wxString(iterator first, iterator last) - : wxStringBase(first, last) { } + : m_impl(first.impl(), last.impl()) { } +#if WXWIN_COMPATIBILITY_STRING_PTR_AS_ITER + // the 2 overloads below are for compatibility with the existing code using + // pointers instead of iterators + wxString(const char *first, const char *last) + { + SubstrBufFromMB str(ImplStr(first, last - first)); + m_impl.assign(str.data, str.len); + } + wxString(const wchar_t *first, const wchar_t *last) + { + SubstrBufFromWC str(ImplStr(first, last - first)); + m_impl.assign(str.data, str.len); + } + // and this one is needed to compile code adding offsets to c_str() result + wxString(const wxCStrData& first, const wxCStrData& last) + : m_impl(CreateConstIterator(first).impl(), + CreateConstIterator(last).impl()) + { + wxASSERT_MSG( first.m_str == last.m_str, + _T("pointers must be into the same string") ); + } +#endif // WXWIN_COMPATIBILITY_STRING_PTR_AS_ITER // lib.string.modifiers // append elements str[pos], ..., str[pos+n] wxString& append(const wxString& str, size_t pos, size_t n) - { return (wxString&)wxStringBase::append(str, pos, n); } + { + size_t from, len; + str.PosLenToImpl(pos, n, &from, &len); + m_impl.append(str.m_impl, from, len); + return *this; + } // append a string wxString& append(const wxString& str) - { return (wxString&)wxStringBase::append(str); } - wxString& append(const wxCStrData& str) - { return (wxString&)wxStringBase::append(str.AsString()); } + { m_impl.append(str.m_impl); return *this; } // append first n (or all if n == npos) characters of sz - wxString& append(const wxChar *sz) - { return (wxString&)wxStringBase::append(sz); } - wxString& append(const wxChar *sz, size_t n) - { return (wxString&)wxStringBase::append(sz, n); } + wxString& append(const char *sz) + { m_impl.append(ImplStr(sz)); return *this; } + wxString& append(const wchar_t *sz) + { m_impl.append(ImplStr(sz)); return *this; } + wxString& append(const char *sz, size_t n) + { + SubstrBufFromMB str(ImplStr(sz, n)); + m_impl.append(str.data, str.len); + return *this; + } + wxString& append(const wchar_t *sz, size_t n) + { + SubstrBufFromWC str(ImplStr(sz, n)); + m_impl.append(str.data, str.len); + return *this; + } + + wxString& append(const wxCStrData& str) + { return append(str.AsString()); } + wxString& append(const wxCharBuffer& str) + { return append(str.data()); } + wxString& append(const wxWCharBuffer& str) + { return append(str.data()); } + // append n copies of ch wxString& append(size_t n, wxUniChar ch) - { return (wxString&)wxStringBase::append(n, ch); } + { +#if wxUSE_UNICODE_UTF8 + if ( !ch.IsAscii() ) + m_impl.append(wxStringOperations::EncodeNChars(n, ch)); + else +#endif + m_impl.append(n, (wxStringCharType)ch); + return *this; + } // append from first to last wxString& append(const_iterator first, const_iterator last) - { return (wxString&)wxStringBase::append(first, last); } + { m_impl.append(first.impl(), last.impl()); return *this; } +#if WXWIN_COMPATIBILITY_STRING_PTR_AS_ITER + wxString& append(const char *first, const char *last) + { return append(first, last - first); } + wxString& append(const wchar_t *first, const wchar_t *last) + { return append(first, last - first); } + wxString& append(const wxCStrData& first, const wxCStrData& last) + { return append(CreateConstIterator(first), CreateConstIterator(last)); } +#endif // WXWIN_COMPATIBILITY_STRING_PTR_AS_ITER // same as `this_string = str' wxString& assign(const wxString& str) - { return (wxString&)wxStringBase::assign(str); } + { m_impl = str.m_impl; return *this; } + wxString& assign(const wxString& str, size_t len) + { + m_impl.assign(str.m_impl, 0, str.LenToImpl(len)); + return *this; + } // same as ` = str[pos..pos + n] wxString& assign(const wxString& str, size_t pos, size_t n) - { return (wxString&)wxStringBase::assign(str, pos, n); } + { + size_t from, len; + str.PosLenToImpl(pos, n, &from, &len); + m_impl.assign(str.m_impl, from, len); + return *this; + } // same as `= first n (or all if n == npos) characters of sz' - wxString& assign(const wxChar *sz) - { return (wxString&)wxStringBase::assign(sz); } - wxString& assign(const wxChar *sz, size_t n) - { return (wxString&)wxStringBase::assign(sz, n); } + wxString& assign(const char *sz) + { m_impl.assign(ImplStr(sz)); return *this; } + wxString& assign(const wchar_t *sz) + { m_impl.assign(ImplStr(sz)); return *this; } + wxString& assign(const char *sz, size_t n) + { + SubstrBufFromMB str(ImplStr(sz, n)); + m_impl.assign(str.data, str.len); + return *this; + } + wxString& assign(const wchar_t *sz, size_t n) + { + SubstrBufFromWC str(ImplStr(sz, n)); + m_impl.assign(str.data, str.len); + return *this; + } + + wxString& assign(const wxCStrData& str) + { return assign(str.AsString()); } + wxString& assign(const wxCharBuffer& str) + { return assign(str.data()); } + wxString& assign(const wxWCharBuffer& str) + { return assign(str.data()); } + wxString& assign(const wxCStrData& str, size_t len) + { return assign(str.AsString(), len); } + wxString& assign(const wxCharBuffer& str, size_t len) + { return assign(str.data(), len); } + wxString& assign(const wxWCharBuffer& str, size_t len) + { return assign(str.data(), len); } + // same as `= n copies of ch' wxString& assign(size_t n, wxUniChar ch) - { return (wxString&)wxStringBase::assign(n, ch); } + { +#if wxUSE_UNICODE_UTF8 + if ( !ch.IsAscii() ) + m_impl.assign(wxStringOperations::EncodeNChars(n, ch)); + else +#endif + m_impl.assign(n, (wxStringCharType)ch); + return *this; + } + + wxString& assign(size_t n, wxUniCharRef ch) + { return assign(n, wxUniChar(ch)); } + wxString& assign(size_t n, char ch) + { return assign(n, wxUniChar(ch)); } + wxString& assign(size_t n, unsigned char ch) + { return assign(n, wxUniChar(ch)); } + wxString& assign(size_t n, wchar_t ch) + { return assign(n, wxUniChar(ch)); } + // assign from first to last wxString& assign(const_iterator first, const_iterator last) - { return (wxString&)wxStringBase::assign(first, last); } + { m_impl.assign(first.impl(), last.impl()); return *this; } +#if WXWIN_COMPATIBILITY_STRING_PTR_AS_ITER + wxString& assign(const char *first, const char *last) + { return assign(first, last - first); } + wxString& assign(const wchar_t *first, const wchar_t *last) + { return assign(first, last - first); } + wxString& assign(const wxCStrData& first, const wxCStrData& last) + { return assign(CreateConstIterator(first), CreateConstIterator(last)); } +#endif // WXWIN_COMPATIBILITY_STRING_PTR_AS_ITER // string comparison -#if !defined(HAVE_STD_STRING_COMPARE) - int compare(const wxStringBase& str) const; + int compare(const wxString& str) const; + int compare(const char* sz) const; + int compare(const wchar_t* sz) const; + int compare(const wxCStrData& str) const + { return compare(str.AsString()); } + int compare(const wxCharBuffer& str) const + { return compare(str.data()); } + int compare(const wxWCharBuffer& str) const + { return compare(str.data()); } // comparison with a substring - int compare(size_t nStart, size_t nLen, const wxStringBase& str) const; + int compare(size_t nStart, size_t nLen, const wxString& str) const; // comparison of 2 substrings int compare(size_t nStart, size_t nLen, - const wxStringBase& str, size_t nStart2, size_t nLen2) const; - // just like strcmp() - int compare(const wxChar* sz) const; + const wxString& str, size_t nStart2, size_t nLen2) const; // substring comparison with first nCount characters of sz int compare(size_t nStart, size_t nLen, - const wxChar* sz, size_t nCount = npos) const; -#endif // !defined HAVE_STD_STRING_COMPARE + const char* sz, size_t nCount = npos) const; + int compare(size_t nStart, size_t nLen, + const wchar_t* sz, size_t nCount = npos) const; // insert another string wxString& insert(size_t nPos, const wxString& str) - { return (wxString&)wxStringBase::insert(nPos, str); } + { insert(begin() + nPos, str.begin(), str.end()); return *this; } // insert n chars of str starting at nStart (in str) wxString& insert(size_t nPos, const wxString& str, size_t nStart, size_t n) - { return (wxString&)wxStringBase::insert(nPos, str, nStart, n); } + { + size_t from, len; + str.PosLenToImpl(nStart, n, &from, &len); + m_impl.insert(PosToImpl(nPos), str.m_impl, from, len); + return *this; + } // insert first n (or all if n == npos) characters of sz - wxString& insert(size_t nPos, const wxChar *sz) - { return (wxString&)wxStringBase::insert(nPos, sz); } - wxString& insert(size_t nPos, const wxChar *sz, size_t n) - { return (wxString&)wxStringBase::insert(nPos, sz, n); } + wxString& insert(size_t nPos, const char *sz) + { m_impl.insert(PosToImpl(nPos), ImplStr(sz)); return *this; } + wxString& insert(size_t nPos, const wchar_t *sz) + { m_impl.insert(PosToImpl(nPos), ImplStr(sz)); return *this; } + wxString& insert(size_t nPos, const char *sz, size_t n) + { + SubstrBufFromMB str(ImplStr(sz, n)); + m_impl.insert(PosToImpl(nPos), str.data, str.len); + return *this; + } + wxString& insert(size_t nPos, const wchar_t *sz, size_t n) + { + SubstrBufFromWC str(ImplStr(sz, n)); + m_impl.insert(PosToImpl(nPos), str.data, str.len); + return *this; + } // insert n copies of ch wxString& insert(size_t nPos, size_t n, wxUniChar ch) - { return (wxString&)wxStringBase::insert(nPos, n, ch); } + { +#if wxUSE_UNICODE_UTF8 + if ( !ch.IsAscii() ) + m_impl.insert(PosToImpl(nPos), wxStringOperations::EncodeNChars(n, ch)); + else +#endif + m_impl.insert(PosToImpl(nPos), n, (wxStringCharType)ch); + return *this; + } iterator insert(iterator it, wxUniChar ch) - { return wxStringBase::insert(it, ch); } + { +#if wxUSE_UNICODE_UTF8 + if ( !ch.IsAscii() ) + { + size_t pos = IterToImplPos(it); + m_impl.insert(pos, wxStringOperations::EncodeChar(ch)); + return iterator(this, m_impl.begin() + pos); + } + else +#endif + return iterator(this, m_impl.insert(it.impl(), (wxStringCharType)ch)); + } void insert(iterator it, const_iterator first, const_iterator last) - { wxStringBase::insert(it, first, last); } + { m_impl.insert(it.impl(), first.impl(), last.impl()); } +#if WXWIN_COMPATIBILITY_STRING_PTR_AS_ITER + void insert(iterator it, const char *first, const char *last) + { insert(it - begin(), first, last - first); } + void insert(iterator it, const wchar_t *first, const wchar_t *last) + { insert(it - begin(), first, last - first); } + void insert(iterator it, const wxCStrData& first, const wxCStrData& last) + { insert(it, CreateConstIterator(first), CreateConstIterator(last)); } +#endif // WXWIN_COMPATIBILITY_STRING_PTR_AS_ITER + void insert(iterator it, size_type n, wxUniChar ch) - { wxStringBase::insert(it, n, ch); } + { +#if wxUSE_UNICODE_UTF8 + if ( !ch.IsAscii() ) + m_impl.insert(IterToImplPos(it), wxStringOperations::EncodeNChars(n, ch)); + else +#endif + m_impl.insert(it.impl(), n, (wxStringCharType)ch); + } // delete characters from nStart to nStart + nLen wxString& erase(size_type pos = 0, size_type n = npos) - { return (wxString&)wxStringBase::erase(pos, n); } + { + size_t from, len; + PosLenToImpl(pos, n, &from, &len); + m_impl.erase(from, len); + return *this; + } + // delete characters from first up to last iterator erase(iterator first, iterator last) - { return wxStringBase::erase(first, last); } + { return iterator(this, m_impl.erase(first.impl(), last.impl())); } iterator erase(iterator first) - { return wxStringBase::erase(first); } + { return iterator(this, m_impl.erase(first.impl())); } + +#ifdef wxSTRING_BASE_HASNT_CLEAR + void clear() { erase(); } +#else + void clear() { m_impl.clear(); } +#endif + + // replaces the substring of length nLen starting at nStart + wxString& replace(size_t nStart, size_t nLen, const char* sz) + { + size_t from, len; + PosLenToImpl(nStart, nLen, &from, &len); + m_impl.replace(from, len, ImplStr(sz)); + return *this; + } + wxString& replace(size_t nStart, size_t nLen, const wchar_t* sz) + { + size_t from, len; + PosLenToImpl(nStart, nLen, &from, &len); + m_impl.replace(from, len, ImplStr(sz)); + return *this; + } + // replaces the substring of length nLen starting at nStart + wxString& replace(size_t nStart, size_t nLen, const wxString& str) + { + size_t from, len; + PosLenToImpl(nStart, nLen, &from, &len); + m_impl.replace(from, len, str.m_impl); + return *this; + } + // replaces the substring with nCount copies of ch + wxString& replace(size_t nStart, size_t nLen, size_t nCount, wxUniChar ch) + { + size_t from, len; + PosLenToImpl(nStart, nLen, &from, &len); +#if wxUSE_UNICODE_UTF8 + if ( !ch.IsAscii() ) + m_impl.replace(from, len, wxStringOperations::EncodeNChars(nCount, ch)); + else +#endif + m_impl.replace(from, len, nCount, (wxStringCharType)ch); + return *this; + } + // replaces a substring with another substring + wxString& replace(size_t nStart, size_t nLen, + const wxString& str, size_t nStart2, size_t nLen2) + { + size_t from, len; + PosLenToImpl(nStart, nLen, &from, &len); + + size_t from2, len2; + str.PosLenToImpl(nStart2, nLen2, &from2, &len2); + + m_impl.replace(from, len, str.m_impl, from2, len2); + return *this; + } + // replaces the substring with first nCount chars of sz + wxString& replace(size_t nStart, size_t nLen, + const char* sz, size_t nCount) + { + size_t from, len; + PosLenToImpl(nStart, nLen, &from, &len); + + SubstrBufFromMB str(ImplStr(sz, nCount)); + + m_impl.replace(from, len, str.data, str.len); + return *this; + } + wxString& replace(size_t nStart, size_t nLen, + const wchar_t* sz, size_t nCount) + { + size_t from, len; + PosLenToImpl(nStart, nLen, &from, &len); + + SubstrBufFromWC str(ImplStr(sz, nCount)); + + m_impl.replace(from, len, str.data, str.len); + return *this; + } + wxString& replace(size_t nStart, size_t nLen, + const wxString& s, size_t nCount) + { + size_t from, len; + PosLenToImpl(nStart, nLen, &from, &len); + m_impl.replace(from, len, s.m_impl.c_str(), s.LenToImpl(nCount)); + return *this; + } + + wxString& replace(iterator first, iterator last, const char* s) + { m_impl.replace(first.impl(), last.impl(), ImplStr(s)); return *this; } + wxString& replace(iterator first, iterator last, const wchar_t* s) + { m_impl.replace(first.impl(), last.impl(), ImplStr(s)); return *this; } + wxString& replace(iterator first, iterator last, const char* s, size_type n) + { + SubstrBufFromMB str(ImplStr(s, n)); + m_impl.replace(first.impl(), last.impl(), str.data, str.len); + return *this; + } + wxString& replace(iterator first, iterator last, const wchar_t* s, size_type n) + { + SubstrBufFromWC str(ImplStr(s, n)); + m_impl.replace(first.impl(), last.impl(), str.data, str.len); + return *this; + } + wxString& replace(iterator first, iterator last, const wxString& s) + { m_impl.replace(first.impl(), last.impl(), s.m_impl); return *this; } + wxString& replace(iterator first, iterator last, size_type n, wxUniChar ch) + { +#if wxUSE_UNICODE_UTF8 + if ( !ch.IsAscii() ) + m_impl.replace(first.impl(), last.impl(), + wxStringOperations::EncodeNChars(n, ch)); + else +#endif + m_impl.replace(first.impl(), last.impl(), n, (wxStringCharType)ch); + return *this; + } + wxString& replace(iterator first, iterator last, + const_iterator first1, const_iterator last1) + { + m_impl.replace(first.impl(), last.impl(), first1.impl(), last1.impl()); + return *this; + } + wxString& replace(iterator first, iterator last, + const char *first1, const char *last1) + { replace(first, last, first1, last1 - first1); return *this; } + wxString& replace(iterator first, iterator last, + const wchar_t *first1, const wchar_t *last1) + { replace(first, last, first1, last1 - first1); return *this; } + + // swap two strings + void swap(wxString& str) + { m_impl.swap(str.m_impl); } + + // find a substring + size_t find(const wxString& str, size_t nStart = 0) const + { return PosFromImpl(m_impl.find(str.m_impl, PosToImpl(nStart))); } + + // find first n characters of sz + size_t find(const char* sz, size_t nStart = 0, size_t n = npos) const + { + SubstrBufFromMB str(ImplStr(sz, n)); + return PosFromImpl(m_impl.find(str.data, PosToImpl(nStart), str.len)); + } + size_t find(const wchar_t* sz, size_t nStart = 0, size_t n = npos) const + { + SubstrBufFromWC str(ImplStr(sz, n)); + return PosFromImpl(m_impl.find(str.data, PosToImpl(nStart), str.len)); + } + size_t find(const wxCharBuffer& s, size_t nStart = 0, size_t n = npos) const + { return find(s.data(), nStart, n); } + size_t find(const wxWCharBuffer& s, size_t nStart = 0, size_t n = npos) const + { return find(s.data(), nStart, n); } + size_t find(const wxCStrData& s, size_t nStart = 0, size_t n = npos) const + { return find(s.AsWChar(), nStart, n); } + + // find the first occurence of character ch after nStart + size_t find(wxUniChar ch, size_t nStart = 0) const + { + return PosFromImpl(m_impl.find(wxStringOperations::EncodeChar(ch), + PosToImpl(nStart))); + } + size_t find(wxUniCharRef ch, size_t nStart = 0) const + { return find(wxUniChar(ch), nStart); } + size_t find(char ch, size_t nStart = 0) const + { return find(wxUniChar(ch), nStart); } + size_t find(unsigned char ch, size_t nStart = 0) const + { return find(wxUniChar(ch), nStart); } + size_t find(wchar_t ch, size_t nStart = 0) const + { return find(wxUniChar(ch), nStart); } + + // rfind() family is exactly like find() but works right to left + + // as find, but from the end + size_t rfind(const wxString& str, size_t nStart = npos) const + { return PosFromImpl(m_impl.rfind(str.m_impl, PosToImpl(nStart))); } + + // as find, but from the end + size_t rfind(const char* sz, size_t nStart = npos, size_t n = npos) const + { + SubstrBufFromMB str(ImplStr(sz, n)); + return PosFromImpl(m_impl.rfind(str.data, PosToImpl(nStart), str.len)); + } + size_t rfind(const wchar_t* sz, size_t nStart = npos, size_t n = npos) const + { + SubstrBufFromWC str(ImplStr(sz, n)); + return PosFromImpl(m_impl.rfind(str.data, PosToImpl(nStart), str.len)); + } + size_t rfind(const wxCharBuffer& s, size_t nStart = npos, size_t n = npos) const + { return rfind(s.data(), nStart, n); } + size_t rfind(const wxWCharBuffer& s, size_t nStart = npos, size_t n = npos) const + { return rfind(s.data(), nStart, n); } + size_t rfind(const wxCStrData& s, size_t nStart = npos, size_t n = npos) const + { return rfind(s.AsWChar(), nStart, n); } + // as find, but from the end + size_t rfind(wxUniChar ch, size_t nStart = npos) const + { + return PosFromImpl(m_impl.rfind(wxStringOperations::EncodeChar(ch), + PosToImpl(nStart))); + } + size_t rfind(wxUniCharRef ch, size_t nStart = npos) const + { return rfind(wxUniChar(ch), nStart); } + size_t rfind(char ch, size_t nStart = npos) const + { return rfind(wxUniChar(ch), nStart); } + size_t rfind(unsigned char ch, size_t nStart = npos) const + { return rfind(wxUniChar(ch), nStart); } + size_t rfind(wchar_t ch, size_t nStart = npos) const + { return rfind(wxUniChar(ch), nStart); } + + // find first/last occurence of any character (not) in the set: +#if wxUSE_STL_BASED_WXSTRING && !wxUSE_UNICODE_UTF8 + // FIXME-UTF8: this is not entirely correct, because it doesn't work if + // sizeof(wchar_t)==2 and surrogates are present in the string; + // should we care? Probably not. + size_t find_first_of(const wxString& str, size_t nStart = 0) const + { return m_impl.find_first_of(str.m_impl, nStart); } + size_t find_first_of(const char* sz, size_t nStart = 0) const + { return m_impl.find_first_of(ImplStr(sz), nStart); } + size_t find_first_of(const wchar_t* sz, size_t nStart = 0) const + { return m_impl.find_first_of(ImplStr(sz), nStart); } + size_t find_first_of(const char* sz, size_t nStart, size_t n) const + { return m_impl.find_first_of(ImplStr(sz), nStart, n); } + size_t find_first_of(const wchar_t* sz, size_t nStart, size_t n) const + { return m_impl.find_first_of(ImplStr(sz), nStart, n); } + size_t find_first_of(wxUniChar c, size_t nStart = 0) const + { return m_impl.find_first_of((wxChar)c, nStart); } + + size_t find_last_of(const wxString& str, size_t nStart = npos) const + { return m_impl.find_last_of(str.m_impl, nStart); } + size_t find_last_of(const char* sz, size_t nStart = npos) const + { return m_impl.find_last_of(ImplStr(sz), nStart); } + size_t find_last_of(const wchar_t* sz, size_t nStart = npos) const + { return m_impl.find_last_of(ImplStr(sz), nStart); } + size_t find_last_of(const char* sz, size_t nStart, size_t n) const + { return m_impl.find_last_of(ImplStr(sz), nStart, n); } + size_t find_last_of(const wchar_t* sz, size_t nStart, size_t n) const + { return m_impl.find_last_of(ImplStr(sz), nStart, n); } + size_t find_last_of(wxUniChar c, size_t nStart = npos) const + { return m_impl.find_last_of((wxChar)c, nStart); } + + size_t find_first_not_of(const wxString& str, size_t nStart = 0) const + { return m_impl.find_first_not_of(str.m_impl, nStart); } + size_t find_first_not_of(const char* sz, size_t nStart = 0) const + { return m_impl.find_first_not_of(ImplStr(sz), nStart); } + size_t find_first_not_of(const wchar_t* sz, size_t nStart = 0) const + { return m_impl.find_first_not_of(ImplStr(sz), nStart); } + size_t find_first_not_of(const char* sz, size_t nStart, size_t n) const + { return m_impl.find_first_not_of(ImplStr(sz), nStart, n); } + size_t find_first_not_of(const wchar_t* sz, size_t nStart, size_t n) const + { return m_impl.find_first_not_of(ImplStr(sz), nStart, n); } + size_t find_first_not_of(wxUniChar c, size_t nStart = 0) const + { return m_impl.find_first_not_of((wxChar)c, nStart); } + + size_t find_last_not_of(const wxString& str, size_t nStart = npos) const + { return m_impl.find_last_not_of(str.m_impl, nStart); } + size_t find_last_not_of(const char* sz, size_t nStart = npos) const + { return m_impl.find_last_not_of(ImplStr(sz), nStart); } + size_t find_last_not_of(const wchar_t* sz, size_t nStart = npos) const + { return m_impl.find_last_not_of(ImplStr(sz), nStart); } + size_t find_last_not_of(const char* sz, size_t nStart, size_t n) const + { return m_impl.find_last_not_of(ImplStr(sz), nStart, n); } + size_t find_last_not_of(const wchar_t* sz, size_t nStart, size_t n) const + { return m_impl.find_last_not_of(ImplStr(sz), nStart, n); } + size_t find_last_not_of(wxUniChar c, size_t nStart = npos) const + { return m_impl.find_last_not_of((wxChar)c, nStart); } +#else + // we can't use std::string implementation in UTF-8 build, because the + // character sets would be interpreted wrongly: -#ifdef wxSTRING_BASE_HASNT_CLEAR - void clear() { erase(); } + // as strpbrk() but starts at nStart, returns npos if not found + size_t find_first_of(const wxString& str, size_t nStart = 0) const +#if wxUSE_UNICODE // FIXME-UTF8: temporary + { return find_first_of(str.wc_str(), nStart); } +#else + { return find_first_of(str.mb_str(), nStart); } +#endif + // same as above + size_t find_first_of(const char* sz, size_t nStart = 0) const; + size_t find_first_of(const wchar_t* sz, size_t nStart = 0) const; + size_t find_first_of(const char* sz, size_t nStart, size_t n) const; + size_t find_first_of(const wchar_t* sz, size_t nStart, size_t n) const; + // same as find(char, size_t) + size_t find_first_of(wxUniChar c, size_t nStart = 0) const + { return find(c, nStart); } + // find the last (starting from nStart) char from str in this string + size_t find_last_of (const wxString& str, size_t nStart = npos) const +#if wxUSE_UNICODE // FIXME-UTF8: temporary + { return find_last_of(str.wc_str(), nStart); } +#else + { return find_last_of(str.mb_str(), nStart); } #endif + // same as above + size_t find_last_of (const char* sz, size_t nStart = npos) const; + size_t find_last_of (const wchar_t* sz, size_t nStart = npos) const; + size_t find_last_of(const char* sz, size_t nStart, size_t n) const; + size_t find_last_of(const wchar_t* sz, size_t nStart, size_t n) const; + // same as above + size_t find_last_of(wxUniChar c, size_t nStart = npos) const + { return rfind(c, nStart); } - // replaces the substring of length nLen starting at nStart - wxString& replace(size_t nStart, size_t nLen, const wxChar* sz) - { return (wxString&)wxStringBase::replace(nStart, nLen, sz); } - // replaces the substring of length nLen starting at nStart - wxString& replace(size_t nStart, size_t nLen, const wxString& str) - { return (wxString&)wxStringBase::replace(nStart, nLen, str); } - // replaces the substring with nCount copies of ch - wxString& replace(size_t nStart, size_t nLen, size_t nCount, wxUniChar ch) - { return (wxString&)wxStringBase::replace(nStart, nLen, nCount, ch); } - // replaces a substring with another substring - wxString& replace(size_t nStart, size_t nLen, - const wxString& str, size_t nStart2, size_t nLen2) - { return (wxString&)wxStringBase::replace(nStart, nLen, str, - nStart2, nLen2); } - // replaces the substring with first nCount chars of sz - wxString& replace(size_t nStart, size_t nLen, - const wxChar* sz, size_t nCount) - { return (wxString&)wxStringBase::replace(nStart, nLen, sz, nCount); } - wxString& replace(iterator first, iterator last, const_pointer s) - { return (wxString&)wxStringBase::replace(first, last, s); } - wxString& replace(iterator first, iterator last, const_pointer s, - size_type n) - { return (wxString&)wxStringBase::replace(first, last, s, n); } - wxString& replace(iterator first, iterator last, const wxString& s) - { return (wxString&)wxStringBase::replace(first, last, s); } - wxString& replace(iterator first, iterator last, size_type n, wxUniChar c) - { return (wxString&)wxStringBase::replace(first, last, n, c); } - wxString& replace(iterator first, iterator last, - const_iterator first1, const_iterator last1) - { return (wxString&)wxStringBase::replace(first, last, first1, last1); } + // find first/last occurence of any character not in the set + + // as strspn() (starting from nStart), returns npos on failure + size_t find_first_not_of(const wxString& str, size_t nStart = 0) const +#if wxUSE_UNICODE // FIXME-UTF8: temporary + { return find_first_not_of(str.wc_str(), nStart); } +#else + { return find_first_not_of(str.mb_str(), nStart); } +#endif + // same as above + size_t find_first_not_of(const char* sz, size_t nStart = 0) const; + size_t find_first_not_of(const wchar_t* sz, size_t nStart = 0) const; + size_t find_first_not_of(const char* sz, size_t nStart, size_t n) const; + size_t find_first_not_of(const wchar_t* sz, size_t nStart, size_t n) const; + // same as above + size_t find_first_not_of(wxUniChar ch, size_t nStart = 0) const; + // as strcspn() + size_t find_last_not_of(const wxString& str, size_t nStart = npos) const +#if wxUSE_UNICODE // FIXME-UTF8: temporary + { return find_last_not_of(str.wc_str(), nStart); } +#else + { return find_last_not_of(str.mb_str(), nStart); } +#endif + // same as above + size_t find_last_not_of(const char* sz, size_t nStart = npos) const; + size_t find_last_not_of(const wchar_t* sz, size_t nStart = npos) const; + size_t find_last_not_of(const char* sz, size_t nStart, size_t n) const; + size_t find_last_not_of(const wchar_t* sz, size_t nStart, size_t n) const; + // same as above + size_t find_last_not_of(wxUniChar ch, size_t nStart = npos) const; +#endif // wxUSE_STL_BASED_WXSTRING && !wxUSE_UNICODE_UTF8 or not + + // provide char/wchar_t/wxUniCharRef overloads for char-finding functions + // above to resolve ambiguities: + size_t find_first_of(wxUniCharRef ch, size_t nStart = 0) const + { return find_first_of(wxUniChar(ch), nStart); } + size_t find_first_of(char ch, size_t nStart = 0) const + { return find_first_of(wxUniChar(ch), nStart); } + size_t find_first_of(unsigned char ch, size_t nStart = 0) const + { return find_first_of(wxUniChar(ch), nStart); } + size_t find_first_of(wchar_t ch, size_t nStart = 0) const + { return find_first_of(wxUniChar(ch), nStart); } + size_t find_last_of(wxUniCharRef ch, size_t nStart = npos) const + { return find_last_of(wxUniChar(ch), nStart); } + size_t find_last_of(char ch, size_t nStart = npos) const + { return find_last_of(wxUniChar(ch), nStart); } + size_t find_last_of(unsigned char ch, size_t nStart = npos) const + { return find_last_of(wxUniChar(ch), nStart); } + size_t find_last_of(wchar_t ch, size_t nStart = npos) const + { return find_last_of(wxUniChar(ch), nStart); } + size_t find_first_not_of(wxUniCharRef ch, size_t nStart = 0) const + { return find_first_not_of(wxUniChar(ch), nStart); } + size_t find_first_not_of(char ch, size_t nStart = 0) const + { return find_first_not_of(wxUniChar(ch), nStart); } + size_t find_first_not_of(unsigned char ch, size_t nStart = 0) const + { return find_first_not_of(wxUniChar(ch), nStart); } + size_t find_first_not_of(wchar_t ch, size_t nStart = 0) const + { return find_first_not_of(wxUniChar(ch), nStart); } + size_t find_last_not_of(wxUniCharRef ch, size_t nStart = npos) const + { return find_last_not_of(wxUniChar(ch), nStart); } + size_t find_last_not_of(char ch, size_t nStart = npos) const + { return find_last_not_of(wxUniChar(ch), nStart); } + size_t find_last_not_of(unsigned char ch, size_t nStart = npos) const + { return find_last_not_of(wxUniChar(ch), nStart); } + size_t find_last_not_of(wchar_t ch, size_t nStart = npos) const + { return find_last_not_of(wxUniChar(ch), nStart); } + + // and additional overloads for the versions taking strings: + size_t find_first_of(const wxCStrData& sz, size_t nStart = 0) const + { return find_first_of(sz.AsString(), nStart); } + size_t find_first_of(const wxCharBuffer& sz, size_t nStart = 0) const + { return find_first_of(sz.data(), nStart); } + size_t find_first_of(const wxWCharBuffer& sz, size_t nStart = 0) const + { return find_first_of(sz.data(), nStart); } + size_t find_first_of(const wxCStrData& sz, size_t nStart, size_t n) const + { return find_first_of(sz.AsWChar(), nStart, n); } + size_t find_first_of(const wxCharBuffer& sz, size_t nStart, size_t n) const + { return find_first_of(sz.data(), nStart, n); } + size_t find_first_of(const wxWCharBuffer& sz, size_t nStart, size_t n) const + { return find_first_of(sz.data(), nStart, n); } + + size_t find_last_of(const wxCStrData& sz, size_t nStart = 0) const + { return find_last_of(sz.AsString(), nStart); } + size_t find_last_of(const wxCharBuffer& sz, size_t nStart = 0) const + { return find_last_of(sz.data(), nStart); } + size_t find_last_of(const wxWCharBuffer& sz, size_t nStart = 0) const + { return find_last_of(sz.data(), nStart); } + size_t find_last_of(const wxCStrData& sz, size_t nStart, size_t n) const + { return find_last_of(sz.AsWChar(), nStart, n); } + size_t find_last_of(const wxCharBuffer& sz, size_t nStart, size_t n) const + { return find_last_of(sz.data(), nStart, n); } + size_t find_last_of(const wxWCharBuffer& sz, size_t nStart, size_t n) const + { return find_last_of(sz.data(), nStart, n); } + + size_t find_first_not_of(const wxCStrData& sz, size_t nStart = 0) const + { return find_first_not_of(sz.AsString(), nStart); } + size_t find_first_not_of(const wxCharBuffer& sz, size_t nStart = 0) const + { return find_first_not_of(sz.data(), nStart); } + size_t find_first_not_of(const wxWCharBuffer& sz, size_t nStart = 0) const + { return find_first_not_of(sz.data(), nStart); } + size_t find_first_not_of(const wxCStrData& sz, size_t nStart, size_t n) const + { return find_first_not_of(sz.AsWChar(), nStart, n); } + size_t find_first_not_of(const wxCharBuffer& sz, size_t nStart, size_t n) const + { return find_first_not_of(sz.data(), nStart, n); } + size_t find_first_not_of(const wxWCharBuffer& sz, size_t nStart, size_t n) const + { return find_first_not_of(sz.data(), nStart, n); } + + size_t find_last_not_of(const wxCStrData& sz, size_t nStart = 0) const + { return find_last_not_of(sz.AsString(), nStart); } + size_t find_last_not_of(const wxCharBuffer& sz, size_t nStart = 0) const + { return find_last_not_of(sz.data(), nStart); } + size_t find_last_not_of(const wxWCharBuffer& sz, size_t nStart = 0) const + { return find_last_not_of(sz.data(), nStart); } + size_t find_last_not_of(const wxCStrData& sz, size_t nStart, size_t n) const + { return find_last_not_of(sz.AsWChar(), nStart, n); } + size_t find_last_not_of(const wxCharBuffer& sz, size_t nStart, size_t n) const + { return find_last_not_of(sz.data(), nStart, n); } + size_t find_last_not_of(const wxWCharBuffer& sz, size_t nStart, size_t n) const + { return find_last_not_of(sz.data(), nStart, n); } // string += string wxString& operator+=(const wxString& s) - { return (wxString&)wxStringBase::operator+=(s); } + { m_impl += s.m_impl; return *this; } // string += C string - wxString& operator+=(const wxChar *psz) - { return (wxString&)wxStringBase::operator+=(psz); } + wxString& operator+=(const char *psz) + { m_impl += ImplStr(psz); return *this; } + wxString& operator+=(const wchar_t *pwz) + { m_impl += ImplStr(pwz); return *this; } wxString& operator+=(const wxCStrData& s) - { return (wxString&)wxStringBase::operator+=(s.AsString()); } + { m_impl += s.AsString().m_impl; return *this; } + wxString& operator+=(const wxCharBuffer& s) + { return operator+=(s.data()); } + wxString& operator+=(const wxWCharBuffer& s) + { return operator+=(s.data()); } // string += char wxString& operator+=(wxUniChar ch) - { return (wxString&)wxStringBase::operator+=(ch); } + { m_impl += wxStringOperations::EncodeChar(ch); return *this; } wxString& operator+=(wxUniCharRef ch) { return *this += wxUniChar(ch); } + wxString& operator+=(int ch) { return *this += wxUniChar(ch); } wxString& operator+=(char ch) { return *this += wxUniChar(ch); } + wxString& operator+=(unsigned char ch) { return *this += wxUniChar(ch); } wxString& operator+=(wchar_t ch) { return *this += wxUniChar(ch); } private: -#if !wxUSE_STL +#if !wxUSE_STL_BASED_WXSTRING // helpers for wxStringBuffer and wxStringBufferLength - wxChar *DoGetWriteBuf(size_t nLen); - void DoUngetWriteBuf(); - void DoUngetWriteBuf(size_t nLen); + wxStringCharType *DoGetWriteBuf(size_t nLen) + { return m_impl.DoGetWriteBuf(nLen); } + void DoUngetWriteBuf() + { m_impl.DoUngetWriteBuf(); } + void DoUngetWriteBuf(size_t nLen) + { m_impl.DoUngetWriteBuf(nLen); } +#endif // !wxUSE_STL_BASED_WXSTRING + +#ifndef wxNEEDS_WXSTRING_PRINTF_MIXIN + #if !wxUSE_UTF8_LOCALE_ONLY + int DoPrintfWchar(const wxChar *format, ...); + static wxString DoFormatWchar(const wxChar *format, ...); + #endif + #if wxUSE_UNICODE_UTF8 + int DoPrintfUtf8(const char *format, ...); + static wxString DoFormatUtf8(const char *format, ...); + #endif +#endif - friend class WXDLLIMPEXP_BASE wxStringBuffer; - friend class WXDLLIMPEXP_BASE wxStringBufferLength; +#if !wxUSE_STL_BASED_WXSTRING + // check string's data validity + bool IsValid() const { return m_impl.GetStringData()->IsValid(); } #endif -#ifndef wxNEEDS_WXSTRING_PRINTF_MIXIN - int DoPrintf(const wxChar *format, ...) ATTRIBUTE_PRINTF_2; - static wxString DoFormat(const wxChar *format, ...) ATTRIBUTE_PRINTF_1; +private: + wxStringImpl m_impl; + +#ifdef __VISUALC__ + // "struct 'ConvertedBuffer' needs to have dll-interface to be used by + // clients of class 'wxString'" - this is private, we don't care + #pragma warning (disable:4251) +#endif + + // buffers for compatibility conversion from (char*)c_str() and + // (wchar_t*)c_str(): + // FIXME-UTF8: bechmark various approaches to keeping compatibility buffers + template + struct ConvertedBuffer + { + ConvertedBuffer() : m_buf(NULL) {} + ~ConvertedBuffer() + { free(m_buf); } + + operator T*() const { return m_buf; } + + ConvertedBuffer& operator=(T *str) + { + free(m_buf); + m_buf = str; + return *this; + } + + T *m_buf; + }; +#if wxUSE_UNICODE && !wxUSE_UTF8_LOCALE_ONLY + ConvertedBuffer m_convertedToChar; +#endif +#if !wxUSE_UNICODE_WCHAR + ConvertedBuffer m_convertedToWChar; +#endif + +#ifdef __VISUALC__ + #pragma warning (default:4251) #endif + +#if wxUSE_UNICODE_UTF8 + // FIXME-UTF8: (try to) move this elsewhere (TLS) or solve differently + // assigning to character pointer to by wxString::interator may + // change the underlying wxStringImpl iterator, so we have to + // keep track of all iterators and update them as necessary: + struct wxStringIteratorNodeHead + { + wxStringIteratorNodeHead() : ptr(NULL) {} + wxStringIteratorNode *ptr; + + // copying is disallowed as it would result in more than one pointer into + // the same linked list + DECLARE_NO_COPY_CLASS(wxStringIteratorNodeHead) + }; + + wxStringIteratorNodeHead m_iterators; + + friend class WXDLLIMPEXP_FWD_BASE wxStringIteratorNode; + friend class WXDLLIMPEXP_FWD_BASE wxUniCharRef; +#endif // wxUSE_UNICODE_UTF8 + + friend class WXDLLIMPEXP_FWD_BASE wxCStrData; + friend class wxStringInternalBuffer; + friend class wxStringInternalBufferLength; }; #ifdef wxNEEDS_WXSTRING_PRINTF_MIXIN #pragma warning (default:4275) #endif +// string iterator operators that satisfy STL Random Access Iterator +// requirements: +inline wxString::iterator operator+(int n, wxString::iterator i) + { return i + n; } +inline wxString::iterator operator+(size_t n, wxString::iterator i) + { return i + n; } +inline wxString::const_iterator operator+(int n, wxString::const_iterator i) + { return i + n; } +inline wxString::const_iterator operator+(size_t n, wxString::const_iterator i) + { return i + n; } +inline wxString::reverse_iterator operator+(int n, wxString::reverse_iterator i) + { return i + n; } +inline wxString::reverse_iterator operator+(size_t n, wxString::reverse_iterator i) + { return i + n; } +inline wxString::const_reverse_iterator operator+(int n, wxString::const_reverse_iterator i) + { return i + n; } +inline wxString::const_reverse_iterator operator+(size_t n, wxString::const_reverse_iterator i) + { return i + n; } + // notice that even though for many compilers the friend declarations above are // enough, from the point of view of C++ standard we must have the declarations // here as friend ones are not injected in the enclosing namespace and without // them the code fails to compile with conforming compilers such as xlC or g++4 wxString WXDLLIMPEXP_BASE operator+(const wxString& string1, const wxString& string2); -wxString WXDLLIMPEXP_BASE operator+(const wxString& string, const wxChar *psz); -wxString WXDLLIMPEXP_BASE operator+(const wxChar *psz, const wxString& string); +wxString WXDLLIMPEXP_BASE operator+(const wxString& string, const char *psz); +wxString WXDLLIMPEXP_BASE operator+(const wxString& string, const wchar_t *pwz); +wxString WXDLLIMPEXP_BASE operator+(const char *psz, const wxString& string); +wxString WXDLLIMPEXP_BASE operator+(const wchar_t *pwz, const wxString& string); wxString WXDLLIMPEXP_BASE operator+(const wxString& string, wxUniChar ch); wxString WXDLLIMPEXP_BASE operator+(wxUniChar ch, const wxString& string); @@ -1695,171 +2701,245 @@ inline wxString operator+(wchar_t ch, const wxString& string) { return wxUniChar(ch) + string; } -#if wxUSE_STL - // return an empty wxString (not very useful with wxUSE_STL == 1) - inline const wxString wxGetEmptyString() { return wxString(); } -#else // !wxUSE_STL - // return an empty wxString (more efficient than wxString() here) - inline const wxString& wxGetEmptyString() - { - return *(wxString *)&wxEmptyString; - } -#endif // wxUSE_STL/!wxUSE_STL +#define wxGetEmptyString() wxString() // ---------------------------------------------------------------------------- // wxStringBuffer: a tiny class allowing to get a writable pointer into string // ---------------------------------------------------------------------------- -#if wxUSE_STL - -class WXDLLIMPEXP_BASE wxStringBuffer +#if !wxUSE_STL_BASED_WXSTRING +// string buffer for direct access to string data in their native +// representation: +class wxStringInternalBuffer { public: - wxStringBuffer(wxString& str, size_t lenWanted = 1024) - : m_str(str), m_buf(lenWanted) - { } + typedef wxStringCharType CharType; - ~wxStringBuffer() { m_str.assign(m_buf.data(), wxStrlen(m_buf.data())); } + wxStringInternalBuffer(wxString& str, size_t lenWanted = 1024) + : m_str(str), m_buf(NULL) + { m_buf = m_str.DoGetWriteBuf(lenWanted); } - operator wxChar*() { return m_buf.data(); } + ~wxStringInternalBuffer() { m_str.DoUngetWriteBuf(); } + + operator wxStringCharType*() const { return m_buf; } private: - wxString& m_str; -#if wxUSE_UNICODE - wxWCharBuffer m_buf; -#else - wxCharBuffer m_buf; -#endif + wxString& m_str; + wxStringCharType *m_buf; - DECLARE_NO_COPY_CLASS(wxStringBuffer) + DECLARE_NO_COPY_CLASS(wxStringInternalBuffer) }; -class WXDLLIMPEXP_BASE wxStringBufferLength +class wxStringInternalBufferLength { public: - wxStringBufferLength(wxString& str, size_t lenWanted = 1024) - : m_str(str), m_buf(lenWanted), m_len(0), m_lenSet(false) - { } + typedef wxStringCharType CharType; + + wxStringInternalBufferLength(wxString& str, size_t lenWanted = 1024) + : m_str(str), m_buf(NULL), m_len(0), m_lenSet(false) + { + m_buf = m_str.DoGetWriteBuf(lenWanted); + wxASSERT(m_buf != NULL); + } - ~wxStringBufferLength() + ~wxStringInternalBufferLength() { wxASSERT(m_lenSet); - m_str.assign(m_buf.data(), m_len); + m_str.DoUngetWriteBuf(m_len); } - operator wxChar*() { return m_buf.data(); } + operator wxStringCharType*() const { return m_buf; } void SetLength(size_t length) { m_len = length; m_lenSet = true; } private: - wxString& m_str; -#if wxUSE_UNICODE - wxWCharBuffer m_buf; -#else - wxCharBuffer m_buf; -#endif - size_t m_len; - bool m_lenSet; + wxString& m_str; + wxStringCharType *m_buf; + size_t m_len; + bool m_lenSet; - DECLARE_NO_COPY_CLASS(wxStringBufferLength) + DECLARE_NO_COPY_CLASS(wxStringInternalBufferLength) }; -#else // if !wxUSE_STL +#endif // !wxUSE_STL_BASED_WXSTRING -class WXDLLIMPEXP_BASE wxStringBuffer +template +class WXDLLIMPEXP_BASE wxStringTypeBufferBase { public: - wxStringBuffer(wxString& str, size_t lenWanted = 1024) - : m_str(str), m_buf(NULL) - { m_buf = m_str.DoGetWriteBuf(lenWanted); } + typedef T CharType; - ~wxStringBuffer() { m_str.DoUngetWriteBuf(); } + wxStringTypeBufferBase(wxString& str, size_t lenWanted = 1024) + : m_str(str), m_buf(lenWanted) + { } - operator wxChar*() const { return m_buf; } -private: + operator CharType*() { return m_buf.data(); } + +protected: + wxString& m_str; + wxCharTypeBuffer m_buf; +}; + +template +class WXDLLIMPEXP_BASE wxStringTypeBufferLengthBase +{ +public: + typedef T CharType; + + wxStringTypeBufferLengthBase(wxString& str, size_t lenWanted = 1024) + : m_str(str), m_buf(lenWanted), m_len(0), m_lenSet(false) + { } + + operator CharType*() { return m_buf.data(); } + void SetLength(size_t length) { m_len = length; m_lenSet = true; } + +protected: wxString& m_str; - wxChar *m_buf; + wxCharTypeBuffer m_buf; + size_t m_len; + bool m_lenSet; +}; + +template +class wxStringTypeBuffer : public wxStringTypeBufferBase +{ +public: + wxStringTypeBuffer(wxString& str, size_t lenWanted = 1024) + : wxStringTypeBufferBase(str, lenWanted) {} + ~wxStringTypeBuffer() + { + this->m_str.assign(this->m_buf.data()); + } - DECLARE_NO_COPY_CLASS(wxStringBuffer) + DECLARE_NO_COPY_CLASS(wxStringTypeBuffer) }; -class WXDLLIMPEXP_BASE wxStringBufferLength +template +class wxStringTypeBufferLength : public wxStringTypeBufferLengthBase { public: - wxStringBufferLength(wxString& str, size_t lenWanted = 1024) - : m_str(str), m_buf(NULL), m_len(0), m_lenSet(false) + wxStringTypeBufferLength(wxString& str, size_t lenWanted = 1024) + : wxStringTypeBufferLengthBase(str, lenWanted) {} + + ~wxStringTypeBufferLength() { - m_buf = m_str.DoGetWriteBuf(lenWanted); - wxASSERT(m_buf != NULL); + wxASSERT(this->m_lenSet); + this->m_str.assign(this->m_buf.data(), this->m_len); } - ~wxStringBufferLength() + DECLARE_NO_COPY_CLASS(wxStringTypeBufferLength) +}; + +#if wxUSE_STL_BASED_WXSTRING + +WXDLLIMPEXP_TEMPLATE_INSTANCE_BASE( wxStringTypeBufferBase ) + +class wxStringInternalBuffer : public wxStringTypeBufferBase +{ +public: + wxStringInternalBuffer(wxString& str, size_t lenWanted = 1024) + : wxStringTypeBufferBase(str, lenWanted) {} + ~wxStringInternalBuffer() + { m_str.m_impl.assign(m_buf.data()); } + + DECLARE_NO_COPY_CLASS(wxStringInternalBuffer) +}; + +WXDLLIMPEXP_TEMPLATE_INSTANCE_BASE( + wxStringTypeBufferLengthBase ) + +class wxStringInternalBufferLength + : public wxStringTypeBufferLengthBase +{ +public: + wxStringInternalBufferLength(wxString& str, size_t lenWanted = 1024) + : wxStringTypeBufferLengthBase(str, lenWanted) {} + + ~wxStringInternalBufferLength() { wxASSERT(m_lenSet); - m_str.DoUngetWriteBuf(m_len); + m_str.m_impl.assign(m_buf.data(), m_len); } - operator wxChar*() const { return m_buf; } - void SetLength(size_t length) { m_len = length; m_lenSet = true; } + DECLARE_NO_COPY_CLASS(wxStringInternalBufferLength) +}; +#endif // wxUSE_STL_BASED_WXSTRING -private: - wxString& m_str; - wxChar *m_buf; - size_t m_len; - bool m_lenSet; - DECLARE_NO_COPY_CLASS(wxStringBufferLength) +#if wxUSE_STL_BASED_WXSTRING || wxUSE_UNICODE_UTF8 +typedef wxStringTypeBuffer wxStringBuffer; +typedef wxStringTypeBufferLength wxStringBufferLength; +#else // if !wxUSE_STL_BASED_WXSTRING && !wxUSE_UNICODE_UTF8 +typedef wxStringInternalBuffer wxStringBuffer; +typedef wxStringInternalBufferLength wxStringBufferLength; +#endif // !wxUSE_STL_BASED_WXSTRING && !wxUSE_UNICODE_UTF8 + +#if wxUSE_UNICODE_UTF8 +typedef wxStringInternalBuffer wxUTF8StringBuffer; +typedef wxStringInternalBufferLength wxUTF8StringBufferLength; +#elif wxUSE_UNICODE_WCHAR + +WXDLLIMPEXP_TEMPLATE_INSTANCE_BASE( wxStringTypeBufferBase ) + +class WXDLLIMPEXP_BASE wxUTF8StringBuffer : public wxStringTypeBufferBase +{ +public: + wxUTF8StringBuffer(wxString& str, size_t lenWanted = 1024) + : wxStringTypeBufferBase(str, lenWanted) {} + ~wxUTF8StringBuffer(); + + DECLARE_NO_COPY_CLASS(wxUTF8StringBuffer) +}; + +WXDLLIMPEXP_TEMPLATE_INSTANCE_BASE( wxStringTypeBufferLengthBase ) + +class WXDLLIMPEXP_BASE wxUTF8StringBufferLength + : public wxStringTypeBufferLengthBase +{ +public: + wxUTF8StringBufferLength(wxString& str, size_t lenWanted = 1024) + : wxStringTypeBufferLengthBase(str, lenWanted) {} + ~wxUTF8StringBufferLength(); + + DECLARE_NO_COPY_CLASS(wxUTF8StringBufferLength) }; +#endif // wxUSE_UNICODE_UTF8/wxUSE_UNICODE_WCHAR -#endif // !wxUSE_STL // --------------------------------------------------------------------------- // wxString comparison functions: operator versions are always case sensitive // --------------------------------------------------------------------------- -// note that when wxUSE_STL == 1 the comparison operators taking std::string -// are used and defining them also for wxString would only result in -// compilation ambiguities when comparing std::string and wxString -#if !wxUSE_STL +#define wxCMP_WXCHAR_STRING(p, s, op) 0 op s.Cmp(p) +wxDEFINE_ALL_COMPARISONS(const wxChar *, const wxString&, wxCMP_WXCHAR_STRING) + +#undef wxCMP_WXCHAR_STRING + +// note that there is an optimization in operator==() and !=(): we (quickly) +// checks the strings length first, before comparing their data inline bool operator==(const wxString& s1, const wxString& s2) { return (s1.Len() == s2.Len()) && (s1.Cmp(s2) == 0); } -inline bool operator==(const wxString& s1, const wxChar * s2) - { return s1.Cmp(s2) == 0; } -inline bool operator==(const wxChar * s1, const wxString& s2) - { return s2.Cmp(s1) == 0; } inline bool operator!=(const wxString& s1, const wxString& s2) { return (s1.Len() != s2.Len()) || (s1.Cmp(s2) != 0); } -inline bool operator!=(const wxString& s1, const wxChar * s2) - { return s1.Cmp(s2) != 0; } -inline bool operator!=(const wxChar * s1, const wxString& s2) - { return s2.Cmp(s1) != 0; } inline bool operator< (const wxString& s1, const wxString& s2) { return s1.Cmp(s2) < 0; } -inline bool operator< (const wxString& s1, const wxChar * s2) - { return s1.Cmp(s2) < 0; } -inline bool operator< (const wxChar * s1, const wxString& s2) - { return s2.Cmp(s1) > 0; } inline bool operator> (const wxString& s1, const wxString& s2) { return s1.Cmp(s2) > 0; } -inline bool operator> (const wxString& s1, const wxChar * s2) - { return s1.Cmp(s2) > 0; } -inline bool operator> (const wxChar * s1, const wxString& s2) - { return s2.Cmp(s1) < 0; } inline bool operator<=(const wxString& s1, const wxString& s2) { return s1.Cmp(s2) <= 0; } -inline bool operator<=(const wxString& s1, const wxChar * s2) - { return s1.Cmp(s2) <= 0; } -inline bool operator<=(const wxChar * s1, const wxString& s2) - { return s2.Cmp(s1) >= 0; } inline bool operator>=(const wxString& s1, const wxString& s2) { return s1.Cmp(s2) >= 0; } -inline bool operator>=(const wxString& s1, const wxChar * s2) - { return s1.Cmp(s2) >= 0; } -inline bool operator>=(const wxChar * s1, const wxString& s2) - { return s2.Cmp(s1) <= 0; } -#if wxUSE_UNICODE +inline bool operator==(const wxString& s1, const wxCStrData& s2) + { return s1 == s2.AsString(); } +inline bool operator==(const wxCStrData& s1, const wxString& s2) + { return s1.AsString() == s2; } +inline bool operator!=(const wxString& s1, const wxCStrData& s2) + { return s1 != s2.AsString(); } +inline bool operator!=(const wxCStrData& s1, const wxString& s2) + { return s1.AsString() != s2; } + inline bool operator==(const wxString& s1, const wxWCharBuffer& s2) { return (s1.Cmp((const wchar_t *)s2) == 0); } inline bool operator==(const wxWCharBuffer& s1, const wxString& s2) @@ -1868,7 +2948,7 @@ inline bool operator!=(const wxString& s1, const wxWCharBuffer& s2) { return (s1.Cmp((const wchar_t *)s2) != 0); } inline bool operator!=(const wxWCharBuffer& s1, const wxString& s2) { return (s2.Cmp((const wchar_t *)s1) != 0); } -#else // !wxUSE_UNICODE + inline bool operator==(const wxString& s1, const wxCharBuffer& s2) { return (s1.Cmp((const char *)s2) == 0); } inline bool operator==(const wxCharBuffer& s1, const wxString& s2) @@ -1877,24 +2957,18 @@ inline bool operator!=(const wxString& s1, const wxCharBuffer& s2) { return (s1.Cmp((const char *)s2) != 0); } inline bool operator!=(const wxCharBuffer& s1, const wxString& s2) { return (s2.Cmp((const char *)s1) != 0); } -#endif // wxUSE_UNICODE/!wxUSE_UNICODE -#if wxUSE_UNICODE inline wxString operator+(const wxString& string, const wxWCharBuffer& buf) { return string + (const wchar_t *)buf; } inline wxString operator+(const wxWCharBuffer& buf, const wxString& string) { return (const wchar_t *)buf + string; } -#else // !wxUSE_UNICODE + inline wxString operator+(const wxString& string, const wxCharBuffer& buf) { return string + (const char *)buf; } inline wxString operator+(const wxCharBuffer& buf, const wxString& string) { return (const char *)buf + string; } -#endif // wxUSE_UNICODE/!wxUSE_UNICODE -#endif // !wxUSE_STL - -// comparison with char (those are not defined by std::[w]string and so should -// be always available) +// comparison with char inline bool operator==(const wxUniChar& c, const wxString& s) { return s.IsSameAs(c); } inline bool operator==(const wxUniCharRef& c, const wxString& s) { return s.IsSameAs(c); } inline bool operator==(char c, const wxString& s) { return s.IsSameAs(c); } @@ -1916,47 +2990,53 @@ inline bool operator!=(const wxString& s, wchar_t c) { return !s.IsSameAs(c); } // comparison with C string in Unicode build #if wxUSE_UNICODE -inline bool operator==(const wxString& s1, const char* s2) - { return s1 == wxString(s2); } -inline bool operator==(const char* s1, const wxString& s2) - { return wxString(s1) == s2; } -inline bool operator!=(const wxString& s1, const char* s2) - { return s1 != wxString(s2); } -inline bool operator!=(const char* s1, const wxString& s2) - { return wxString(s1) != s2; } -inline bool operator< (const wxString& s1, const char* s2) - { return s1 < wxString(s2); } -inline bool operator< (const char* s1, const wxString& s2) - { return wxString(s1) < s2; } -inline bool operator> (const wxString& s1, const char* s2) - { return s1 > wxString(s2); } -inline bool operator> (const char* s1, const wxString& s2) - { return wxString(s1) > s2; } -inline bool operator<=(const wxString& s1, const char* s2) - { return s1 <= wxString(s2); } -inline bool operator<=(const char* s1, const wxString& s2) - { return wxString(s1) <= s2; } -inline bool operator>=(const wxString& s1, const char* s2) - { return s1 >= wxString(s2); } -inline bool operator>=(const char* s1, const wxString& s2) - { return wxString(s1) >= s2; } + +#define wxCMP_CHAR_STRING(p, s, op) wxString(p) op s + +wxDEFINE_ALL_COMPARISONS(const char *, const wxString&, wxCMP_CHAR_STRING) + +#undef wxCMP_CHAR_STRING + #endif // wxUSE_UNICODE +// we also need to provide the operators for comparison with wxCStrData to +// resolve ambiguity between operator(const wxChar *,const wxString &) and +// operator(const wxChar *, const wxChar *) for "p == s.c_str()" +// +// notice that these are (shallow) pointer comparisons, not (deep) string ones +#define wxCMP_CHAR_CSTRDATA(p, s, op) p op s.AsChar() +#define wxCMP_WCHAR_CSTRDATA(p, s, op) p op s.AsWChar() + +wxDEFINE_ALL_COMPARISONS(const wchar_t *, const wxCStrData&, wxCMP_WCHAR_CSTRDATA) +wxDEFINE_ALL_COMPARISONS(const char *, const wxCStrData&, wxCMP_CHAR_CSTRDATA) + +#undef wxCMP_CHAR_CSTRDATA +#undef wxCMP_WCHAR_CSTRDATA + // --------------------------------------------------------------------------- // Implementation only from here until the end of file // --------------------------------------------------------------------------- -// don't pollute the library user's name space -#undef wxASSERT_VALID_INDEX - #if wxUSE_STD_IOSTREAM #include "wx/iosfwrap.h" WXDLLIMPEXP_BASE wxSTD ostream& operator<<(wxSTD ostream&, const wxString&); WXDLLIMPEXP_BASE wxSTD ostream& operator<<(wxSTD ostream&, const wxCStrData&); +WXDLLIMPEXP_BASE wxSTD ostream& operator<<(wxSTD ostream&, const wxCharBuffer&); +#ifndef __BORLANDC__ +WXDLLIMPEXP_BASE wxSTD ostream& operator<<(wxSTD ostream&, const wxWCharBuffer&); +#endif + +#if wxUSE_UNICODE && defined(HAVE_WOSTREAM) + +WXDLLIMPEXP_BASE wxSTD wostream& operator<<(wxSTD wostream&, const wxString&); +WXDLLIMPEXP_BASE wxSTD wostream& operator<<(wxSTD wostream&, const wxCStrData&); +WXDLLIMPEXP_BASE wxSTD wostream& operator<<(wxSTD wostream&, const wxWCharBuffer&); + +#endif // wxUSE_UNICODE && defined(HAVE_WOSTREAM) -#endif // wxSTD_STRING_COMPATIBILITY +#endif // wxUSE_STD_IOSTREAM // --------------------------------------------------------------------------- // wxCStrData implementation @@ -1967,22 +3047,58 @@ inline wxCStrData::wxCStrData(char *buf) inline wxCStrData::wxCStrData(wchar_t *buf) : m_str(new wxString(buf)), m_offset(0), m_owned(true) {} +inline wxCStrData::wxCStrData(const wxCStrData& data) + : m_str(data.m_owned ? new wxString(*data.m_str) : data.m_str), + m_offset(data.m_offset), + m_owned(data.m_owned) +{ +} + inline wxCStrData::~wxCStrData() { if ( m_owned ) - delete m_str; + delete wx_const_cast(wxString*, m_str); // cast to silence warnings } -#if wxUSE_UNICODE +// simple cases for AsChar() and AsWChar(), the complicated ones are +// in string.cpp +#if wxUSE_UNICODE_WCHAR inline const wchar_t* wxCStrData::AsWChar() const -#else +{ + return m_str->wx_str() + m_offset; +} +#endif // wxUSE_UNICODE_WCHAR + +#if !wxUSE_UNICODE +inline const char* wxCStrData::AsChar() const +{ + return m_str->wx_str() + m_offset; +} +#endif // !wxUSE_UNICODE + +#if wxUSE_UTF8_LOCALE_ONLY inline const char* wxCStrData::AsChar() const +{ + return wxStringOperations::AddToIter(m_str->wx_str(), m_offset); +} +#endif // wxUSE_UTF8_LOCALE_ONLY + +inline const wxCharBuffer wxCStrData::AsCharBuf() const +{ +#if !wxUSE_UNICODE + return wxCharBuffer::CreateNonOwned(AsChar()); +#else + return AsString().mb_str(); #endif +} + +inline const wxWCharBuffer wxCStrData::AsWCharBuf() const { - if ( m_offset == 0 ) - return m_str->wx_str(); // FIXME - else - return (const wxChar*)(m_str->begin() + m_offset); +#if wxUSE_UNICODE_WCHAR + return wxWCharBuffer::CreateNonOwned(AsWChar()); +#else + return AsString().wc_str(); +#endif } inline wxString wxCStrData::AsString() const @@ -1993,7 +3109,14 @@ inline wxString wxCStrData::AsString() const return m_str->Mid(m_offset); } -inline wxCStrData::operator wxString() const { return AsString(); } +inline const wxStringCharType *wxCStrData::AsInternal() const +{ +#if wxUSE_UNICODE_UTF8 + return wxStringOperations::AddToIter(m_str->wx_str(), m_offset); +#else + return m_str->wx_str() + m_offset; +#endif +} inline wxUniChar wxCStrData::operator*() const { @@ -2005,7 +3128,90 @@ inline wxUniChar wxCStrData::operator*() const inline wxUniChar wxCStrData::operator[](size_t n) const { - return m_str->at(m_offset + n); + // NB: we intentionally use operator[] and not at() here because the former + // works for the terminating NUL while the latter does not + return (*m_str)[m_offset + n]; +} + +// ---------------------------------------------------------------------------- +// more wxCStrData operators +// ---------------------------------------------------------------------------- + +// we need to define those to allow "size_t pos = p - s.c_str()" where p is +// some pointer into the string +inline size_t operator-(const char *p, const wxCStrData& cs) +{ + return p - cs.AsChar(); +} + +inline size_t operator-(const wchar_t *p, const wxCStrData& cs) +{ + return p - cs.AsWChar(); +} + +// ---------------------------------------------------------------------------- +// implementation of wx[W]CharBuffer inline methods using wxCStrData +// ---------------------------------------------------------------------------- + +// FIXME-UTF8: move this to buffer.h +inline wxCharBuffer::wxCharBuffer(const wxCStrData& cstr) + : wxCharTypeBufferBase(cstr.AsCharBuf()) +{ +} + +inline wxWCharBuffer::wxWCharBuffer(const wxCStrData& cstr) + : wxCharTypeBufferBase(cstr.AsWCharBuf()) +{ +} + +#if wxUSE_UNICODE_UTF8 +// ---------------------------------------------------------------------------- +// implementation of wxStringIteratorNode inline methods +// ---------------------------------------------------------------------------- + +void wxStringIteratorNode::DoSet(const wxString *str, + wxStringImpl::const_iterator *citer, + wxStringImpl::iterator *iter) +{ + m_prev = NULL; + m_iter = iter; + m_citer = citer; + m_str = str; + if ( str ) + { + m_next = str->m_iterators.ptr; + wx_const_cast(wxString*, m_str)->m_iterators.ptr = this; + if ( m_next ) + m_next->m_prev = this; + } + else + { + m_next = NULL; + } +} + +void wxStringIteratorNode::clear() +{ + if ( m_next ) + m_next->m_prev = m_prev; + if ( m_prev ) + m_prev->m_next = m_next; + else if ( m_str ) // first in the list + wx_const_cast(wxString*, m_str)->m_iterators.ptr = m_next; + + m_next = m_prev = NULL; + m_citer = NULL; + m_iter = NULL; + m_str = NULL; } +#endif // wxUSE_UNICODE_UTF8 + +#if WXWIN_COMPATIBILITY_2_8 + // lot of code out there doesn't explicitly include wx/crt.h, but uses + // CRT wrappers that are now declared in wx/wxcrt.h and wx/wxcrtvararg.h, + // so let's include this header now that wxString is defined and it's safe + // to do it: + #include "wx/crt.h" +#endif -#endif // _WX_WXSTRINGH__ +#endif // _WX_WXSTRING_H_