X-Git-Url: https://git.saurik.com/wxWidgets.git/blobdiff_plain/f2a1b1bd23ef03bcfb17c7d2894f4982d8553709..55809d1394516044a90fa34b12070d0f3e9f1439:/include/wx/string.h diff --git a/include/wx/string.h b/include/wx/string.h index dd56e13c7a..07e7e86992 100644 --- a/include/wx/string.h +++ b/include/wx/string.h @@ -56,10 +56,17 @@ #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; +// 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 // --------------------------------------------------------------------------- @@ -169,7 +176,7 @@ inline int Stricmp(const char *psz1, const char *psz2) // Lightweight object returned by wxString::c_str() and implicitly convertible // to either const char* or const wchar_t*. -class wxCStrData +class WXDLLIMPEXP_BASE wxCStrData { private: // Ctors; for internal use by wxString and wxCStrData only @@ -182,13 +189,22 @@ public: wxCStrData(char *buf); wxCStrData(wchar_t *buf); - ~wxCStrData(); + inline ~wxCStrData(); + // 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(); } - operator bool() const; + inline operator bool() const; +#if !wxUSE_UNICODE + inline +#endif const char* AsChar() const; const unsigned char* AsUnsignedChar() const { return (const unsigned char *) AsChar(); } @@ -197,11 +213,14 @@ public: operator const void*() const { return AsChar(); } - wxString AsString() const; + inline const wxCharBuffer AsCharBuf() const; + inline const wxWCharBuffer AsWCharBuf() const; + + inline wxString AsString() const; // 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[](size_t n) const; 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)); } @@ -239,7 +258,7 @@ public: // this operator is needed to make expressions like "*c_str()" or // "*(c_str() + 2)" work - wxUniChar operator*() const; + inline wxUniChar operator*() const; private: const wxString *m_str; @@ -355,20 +374,20 @@ private: // buffer for holding temporary substring when using any of the methods // that take (char*,size_t) or (wchar_t*,size_t) arguments: - // FIXME-UTF8: This will need changes when UTF8 build is introduced template struct SubstrBufFromType { T data; size_t len; - SubstrBufFromType() {} SubstrBufFromType(const T& data_, size_t len_) : data(data_), len(len_) {} }; #if wxUSE_UNICODE_UTF8 - // FIXME-UTF8: this will have to use slightly different type + // 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; @@ -383,8 +402,12 @@ private: // between UTF-8 and wchar_t* representations of the string are mostly // contained here. -#if wxUSE_UNICODE - // FIXME-UTF8: This will need changes when UTF8 build is introduced +#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 @@ -418,27 +441,6 @@ private: { return ConvertStr(str, n, wxConvLibc); } #endif - // moves the iterator to the next Unicode character - static void IncIter(wxStringImpl::iterator& i) { ++i; } - static void IncIter(wxStringImpl::const_iterator& i) { ++i; } - // moves the iterator to the previous Unicode character - static void DecIter(wxStringImpl::iterator& i) { --i; } - static void DecIter(wxStringImpl::const_iterator& i) { --i; } - // moves the iterator by n Unicode characters - static wxStringImpl::iterator AddToIter(wxStringImpl::iterator i, int n) - { return i + n; } - static wxStringImpl::const_iterator AddToIter(wxStringImpl::const_iterator i, int n) - { return i + n; } - // returns distance of the two iterators in Unicode characters - static int DiffIters(wxStringImpl::iterator i1, wxStringImpl::iterator i2) - { return i1 - i2; } - static int DiffIters(wxStringImpl::const_iterator i1, wxStringImpl::const_iterator i2) - { return i1 - i2; } - - // encodes the character to a form used to represent it in internal - // representation (returns a string in UTF8 version) - static wxChar EncodeChar(wxUniChar ch) { return (wxChar)ch; } - // translates position index in wxString to/from index in underlying // wxStringImpl: static size_t PosToImpl(size_t pos) { return pos; } @@ -450,17 +452,36 @@ private: #else // wxUSE_UNICODE_UTF8 - typedef char Utf8CharBuffer[5]; - static Utf8CharBuffer EncodeChar(wxUniChar ch); - // returns n copies of ch encoded in UTF-8 string - static wxCharBuffer EncodeNChars(size_t n, wxUniChar ch); + // FIXME-UTF8: return as-is without copying under UTF8 locale, return + // converted string under other locales - needs wxCharBuffer + // changes + 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, wxConvUTF8).data; } + static SubstrBufFromWC ImplStr(const wchar_t* str, size_t n) + { return ConvertStr(str, n, wxConvUTF8); } size_t PosToImpl(size_t pos) const { if ( pos == 0 || pos == npos ) return pos; else - return wxStringImpl::const_iterator(begin() + pos) - m_impl.begin(); + 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 @@ -470,100 +491,8 @@ private: else return const_iterator(m_impl.begin() + pos) - begin(); } - - // FIXME: return as-is without copying under UTF8 locale, return - // converted string under other locales - needs wxCharBuffer - // changes - static wxCharBuffer ImplStr(const char* str); - - static wxCharBuffer ImplStr(const wchar_t* str) - { return wxConvUTF8.cWC2MB(str); } #endif // !wxUSE_UNICODE_UTF8/wxUSE_UNICODE_UTF8 - -public: - // constructors and destructor - // ctor for an empty string - wxString() {} - - // copy ctor - wxString(const wxStringImpl& stringSrc) : m_impl(stringSrc) { } - wxString(const wxString& stringSrc) : m_impl(stringSrc.m_impl) { } - - // string containing nRepeat copies of ch - 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) - { - SubstrBufFromMB str(ImplStr(psz, nLength, conv)); - m_impl.assign(str.data, str.len); - } - - // 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) - { - SubstrBufFromMB str(ImplStr((const char*)psz, nLength, conv)); - m_impl.assign(str.data, str.len); - } - - // 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) - : m_impl(str.Mid(0, nLength).m_impl) {} - - ~wxString(); - public: // standard types typedef wxUniChar value_type; @@ -576,7 +505,12 @@ public: typedef wxUniChar const_reference; #if wxUSE_STL - #define WX_STR_ITERATOR_TAG std::random_access_iterator_tag + #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 @@ -592,47 +526,49 @@ public: 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++() \ - { wxString::IncIter(m_cur); return *this; } \ + { wxStringOperations::IncIter(m_cur); return *this; } \ iterator_name& operator--() \ - { wxString::DecIter(m_cur); return *this; } \ + { wxStringOperations::DecIter(m_cur); return *this; } \ iterator_name operator++(int) \ { \ iterator_name tmp = *this; \ - wxString::IncIter(m_cur); \ + wxStringOperations::IncIter(m_cur); \ return tmp; \ } \ iterator_name operator--(int) \ { \ iterator_name tmp = *this; \ - wxString::DecIter(m_cur); \ + wxStringOperations::DecIter(m_cur); \ return tmp; \ } \ \ - iterator_name operator+(int n) const \ - { return iterator_name(wxString::AddToIter(m_cur, n)); } \ - iterator_name operator+(size_t n) const \ - { return iterator_name(wxString::AddToIter(m_cur, (int)n)); } \ - iterator_name operator-(int n) const \ - { return iterator_name(wxString::AddToIter(m_cur, -n)); } \ - iterator_name operator-(size_t n) const \ - { return iterator_name(wxString::AddToIter(m_cur, -(int)n)); } \ - iterator_name operator+=(int n) \ - { m_cur = wxString::AddToIter(m_cur, n); return *this; } \ - iterator_name operator+=(size_t n) \ - { m_cur = wxString::AddToIter(m_cur, (int)n); return *this; } \ - iterator_name operator-=(int n) \ - { m_cur = wxString::AddToIter(m_cur, -n); return *this; } \ - iterator_name operator-=(size_t n) \ - { m_cur = wxString::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; \ + } \ + 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 wxString::DiffIters(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 \ { return m_cur == i.m_cur; } \ @@ -650,34 +586,105 @@ public: \ private: \ /* for internal wxString use only: */ \ - iterator_name(underlying_iterator ptr) : m_cur(ptr) {} \ - operator underlying_iterator() const { return m_cur; } \ + underlying_iterator impl() const { return m_cur; } \ \ friend class WXDLLIMPEXP_BASE wxString; \ friend class WXDLLIMPEXP_BASE wxCStrData; \ \ private: \ - underlying_iterator m_cur; + underlying_iterator m_cur class const_iterator; +#if wxUSE_UNICODE_UTF8 class iterator { + // 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. + WX_STR_ITERATOR_IMPL(iterator, wxChar*, wxUniCharRef, - wxUniCharRef::CreateForString(m_cur)) + wxUniCharRef::CreateForString(m_str, m_cur)); + + public: + iterator(const iterator& i) : m_cur(i.m_cur), m_str(i.m_str) {} + + iterator operator+(int n) const + { return iterator(m_str, wxStringOperations::AddToIter(m_cur, n)); } + iterator operator+(size_t n) const + { return iterator(m_str, wxStringOperations::AddToIter(m_cur, (int)n)); } + iterator operator-(int n) const + { return iterator(m_str, wxStringOperations::AddToIter(m_cur, -n)); } + iterator operator-(size_t n) const + { return iterator(m_str, wxStringOperations::AddToIter(m_cur, -(int)n)); } + + private: + iterator(wxString *str, underlying_iterator ptr) + : m_cur(ptr), m_str(str->m_impl) {} + iterator(wxStringImpl& str, underlying_iterator ptr) + : m_cur(ptr), m_str(str) {} + + wxStringImpl& m_str; friend class const_iterator; }; + size_t IterToImplPos(wxString::iterator i) const + { return wxStringImpl::const_iterator(i.impl()) - m_impl.begin(); } + +#else // !wxUSE_UNICODE_UTF8 + + class iterator + { + WX_STR_ITERATOR_IMPL(iterator, wxChar*, wxUniCharRef, + wxUniCharRef::CreateForString(m_cur)); + + public: + iterator(const iterator& i) : m_cur(i.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; + }; +#endif // wxUSE_UNICODE_UTF8/!wxUSE_UNICODE_UTF8 + class 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)) + wxStringOperations::DecodeChar(m_cur)); public: + const_iterator(const const_iterator& i) : m_cur(i.m_cur) {} const_iterator(const iterator& i) : m_cur(i.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) {} }; #undef WX_STR_ITERATOR_TAG @@ -758,12 +765,167 @@ 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->begin() + data.m_offset); + } + + // 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 + wxString() {} + + // copy ctor + wxString(const wxString& stringSrc) : m_impl(stringSrc.m_impl) { } + + // string containing nRepeat copies of ch + 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) + { + SubstrBufFromMB str(ImplStr(psz, nLength, conv)); + m_impl.assign(str.data, str.len); + } + + // 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) + { + SubstrBufFromMB str(ImplStr((const char*)psz, nLength, conv)); + m_impl.assign(str.data, str.len); + } + + // 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) + : m_impl(str.Mid(0, nLength).m_impl) {} + + // 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 + + #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_STL_BASED_WXSTRING + // FIXME-UTF8: do this in UTF8 build #if wxUSE_UTF8_LOCALE_ONLY, too + // 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_STD_STRING + // first valid index position const_iterator begin() const { return const_iterator(m_impl.begin()); } - iterator begin() { return iterator(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(m_impl.end()); } - iterator end() { return iterator(m_impl.end()); } + iterator end() { return iterator(this, m_impl.end()); } // first element of the reversed string const_reverse_iterator rbegin() const @@ -813,7 +975,7 @@ public: { size_t pos, len; PosLenToImpl(nStart, nLen, &pos, &len); - return m_impl.substr(pos, len); + return FromImpl(m_impl.substr(pos, len)); } // generic attributes & operations @@ -918,12 +1080,13 @@ public: // 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 m_impl.c_str(); } + 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 { return mb_str(); } + 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 @@ -956,21 +1119,26 @@ public: 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(); } @@ -992,7 +1160,7 @@ public: { return *this = cstr.AsString(); } // from a character wxString& operator=(wxUniChar ch) - { m_impl = EncodeChar(ch); return *this; } + { m_impl = wxStringOperations::EncodeChar(ch); return *this; } wxString& operator=(wxUniCharRef ch) { return operator=((wxUniChar)ch); } wxString& operator=(char ch) @@ -1408,7 +1576,10 @@ public: : m_impl(str.m_impl, nPos, nLen) { } // take all characters from first to last wxString(const_iterator first, const_iterator last) - : m_impl(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)); @@ -1419,6 +1590,15 @@ public: 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] @@ -1456,7 +1636,7 @@ public: { #if wxUSE_UNICODE_UTF8 if ( !ch.IsAscii() ) - m_impl.append(EncodeNChars(n, ch)); + m_impl.append(wxStringOperations::EncodeNChars(n, ch)); else #endif m_impl.append(n, (wxStringCharType)ch); @@ -1464,11 +1644,15 @@ public: } // append from first to last wxString& append(const_iterator first, const_iterator last) - { m_impl.append(first, last); return *this; } + { 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) @@ -1508,7 +1692,7 @@ public: { #if wxUSE_UNICODE_UTF8 if ( !ch.IsAscii() ) - m_impl.assign(EncodeNChars(n, ch)); + m_impl.assign(wxStringOperations::EncodeNChars(n, ch)); else #endif m_impl.assign(n, (wxStringCharType)ch); @@ -1526,11 +1710,15 @@ public: // assign from first to last wxString& assign(const_iterator first, const_iterator last) - { m_impl.assign(first, last); return *this; } + { 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 int compare(const wxString& str) const; @@ -1581,28 +1769,44 @@ public: { #if wxUSE_UNICODE_UTF8 if ( !ch.IsAscii() ) - m_impl.insert(begin() + nPos, EncodeNChars(n, ch)); + m_impl.insert(PosToImpl(nPos), wxStringOperations::EncodeNChars(n, ch)); else #endif - m_impl.insert(begin() + nPos, n, (wxStringCharType)ch); + m_impl.insert(PosToImpl(nPos), n, (wxStringCharType)ch); return *this; } iterator insert(iterator it, wxUniChar ch) - { return iterator(m_impl.insert(it, EncodeChar(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) - { m_impl.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 append(iterator it, const wchar_t *first, const wchar_t *last) + 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) { #if wxUSE_UNICODE_UTF8 if ( !ch.IsAscii() ) - m_impl.insert(it, EncodeNChars(n, ch)); + m_impl.insert(IterToImplPos(it), wxStringOperations::EncodeNChars(n, ch)); else #endif - m_impl.insert(it, n, (wxStringCharType)ch); + m_impl.insert(it.impl(), n, (wxStringCharType)ch); } // delete characters from nStart to nStart + nLen @@ -1615,9 +1819,9 @@ public: } // delete characters from first up to last iterator erase(iterator first, iterator last) - { return iterator(m_impl.erase(first, last)); } + { return iterator(this, m_impl.erase(first.impl(), last.impl())); } iterator erase(iterator first) - { return iterator(m_impl.erase(first)); } + { return iterator(this, m_impl.erase(first.impl())); } #ifdef wxSTRING_BASE_HASNT_CLEAR void clear() { erase(); } @@ -1655,7 +1859,7 @@ public: PosLenToImpl(nStart, nLen, &from, &len); #if wxUSE_UNICODE_UTF8 if ( !ch.IsAscii() ) - m_impl.replace(from, len, EncodeNChars(nCount, ch)); + m_impl.replace(from, len, wxStringOperations::EncodeNChars(nCount, ch)); else #endif m_impl.replace(from, len, nCount, (wxStringCharType)ch); @@ -1707,36 +1911,40 @@ public: } wxString& replace(iterator first, iterator last, const char* s) - { m_impl.replace(first, last, ImplStr(s)); return *this; } + { 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, last, ImplStr(s)); return *this; } + { 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, last, str.data, str.len); + 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, last, str.data, str.len); + 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, last, s.m_impl); return *this; } + { 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, last, EncodeNChars(n, ch)); + m_impl.replace(first.impl(), last.impl(), + wxStringOperations::EncodeNChars(n, ch)); else #endif - m_impl.replace(first, last, n, (wxStringCharType)ch); + 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, last, first1, last1); return *this; } + { + 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; } @@ -1766,7 +1974,10 @@ public: // find the first occurence of character ch after nStart size_t find(wxUniChar ch, size_t nStart = 0) const - { return PosFromImpl(m_impl.find(EncodeChar(ch), PosToImpl(nStart))); } + { + 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 @@ -1795,7 +2006,10 @@ public: } // as find, but from the end size_t rfind(wxUniChar ch, size_t nStart = npos) const - { return PosFromImpl(m_impl.rfind(EncodeChar(ch), PosToImpl(nStart))); } + { + 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 @@ -1867,7 +2081,11 @@ public: // 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.mb_str().data(), nStart); } +#else { return find_first_of((const wxChar*)str.c_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; @@ -1878,7 +2096,11 @@ public: { 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.mb_str().data(), nStart); } +#else { return find_last_of((const wxChar*)str.c_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; @@ -1892,7 +2114,11 @@ public: // 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.mb_str().data(), nStart); } +#else { return find_first_not_of((const wxChar*)str.c_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; @@ -1902,7 +2128,11 @@ public: 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.mb_str().data(), nStart); } +#else { return find_last_not_of((const wxChar*)str.c_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; @@ -1959,7 +2189,7 @@ public: { m_impl += s.AsString().m_impl; return *this; } // string += char wxString& operator+=(wxUniChar ch) - { m_impl += EncodeChar(ch); return *this; } + { 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); } @@ -1992,6 +2222,35 @@ private: private: wxStringImpl m_impl; + + // 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 + ConvertedBuffer m_convertedToChar; +#endif +#if !wxUSE_UNICODE_WCHAR + ConvertedBuffer m_convertedToWChar; +#endif + friend class WXDLLIMPEXP_BASE wxCStrData; }; #ifdef wxNEEDS_WXSTRING_PRINTF_MIXIN @@ -2166,7 +2425,7 @@ private: // wxString comparison functions: operator versions are always case sensitive // --------------------------------------------------------------------------- -#define wxCMP_WXCHAR_STRING(p, s, op) s.Cmp(p) op 0 +#define wxCMP_WXCHAR_STRING(p, s, op) 0 op s.Cmp(p) wxDEFINE_ALL_COMPARISONS(const wxChar *, const wxString&, wxCMP_WXCHAR_STRING) @@ -2299,7 +2558,7 @@ inline wxCStrData::~wxCStrData() inline wxCStrData::operator bool() const { return !m_str->empty(); -}; +} // simple cases for AsChar() and AsWChar(), the complicated ones are // in string.cpp @@ -2317,6 +2576,24 @@ inline const char* wxCStrData::AsChar() const } #endif // !wxUSE_UNICODE +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 wxUSE_UNICODE_WCHAR + return wxWCharBuffer::CreateNonOwned(AsWChar()); +#else + return AsString().wc_str(); +#endif +} + inline wxString wxCStrData::AsString() const { if ( m_offset == 0 ) @@ -2335,7 +2612,25 @@ 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(); } // ---------------------------------------------------------------------------- @@ -2344,12 +2639,12 @@ inline wxUniChar wxCStrData::operator[](size_t n) const // FIXME-UTF8: move this to buffer.h inline wxCharBuffer::wxCharBuffer(const wxCStrData& cstr) - : wxCharTypeBufferBase(cstr.AsChar()) + : wxCharTypeBufferBase(cstr.AsCharBuf()) { } inline wxWCharBuffer::wxWCharBuffer(const wxCStrData& cstr) - : wxCharTypeBufferBase(cstr.AsWChar()) + : wxCharTypeBufferBase(cstr.AsWCharBuf()) { }