X-Git-Url: https://git.saurik.com/wxWidgets.git/blobdiff_plain/681e44128670bd58c09ff9b42af52666635b19a6..55809d1394516044a90fa34b12070d0f3e9f1439:/include/wx/string.h diff --git a/include/wx/string.h b/include/wx/string.h index d60073694c..07e7e86992 100644 --- a/include/wx/string.h +++ b/include/wx/string.h @@ -56,6 +56,7 @@ #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; @@ -373,7 +374,6 @@ 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 { @@ -441,28 +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; } - static wxUniChar DecodeChar(wxStringImpl::const_iterator i) { return *i; } - // translates position index in wxString to/from index in underlying // wxStringImpl: static size_t PosToImpl(size_t pos) { return pos; } @@ -474,55 +452,20 @@ private: #else // wxUSE_UNICODE_UTF8 - // checks correctness of UTF-8 sequence - static bool IsValidUtf8String(const char *c); -#ifdef __WXDEBUG__ - static bool IsValidUtf8LeadByte(unsigned char c); -#endif - - // table of offsets to skip forward when iterating - static unsigned char ms_utf8IterTable[256]; - - static void IncIter(wxStringImpl::iterator& i) - { - wxASSERT( IsValidUtf8LeadByte(*i) ); - i += ms_utf8IterTable[(unsigned char)*i]; - } - static void IncIter(wxStringImpl::const_iterator& i) - { - wxASSERT( IsValidUtf8LeadByte(*i) ); - i += ms_utf8IterTable[(unsigned char)*i]; - } - - static void DecIter(wxStringImpl::iterator& i); - static void DecIter(wxStringImpl::const_iterator& i); - static wxStringImpl::iterator AddToIter(wxStringImpl::iterator i, int n); - static wxStringImpl::const_iterator AddToIter(wxStringImpl::const_iterator i, int n); - static int DiffIters(wxStringImpl::const_iterator i1, wxStringImpl::const_iterator i2); - static int DiffIters(wxStringImpl::iterator i1, wxStringImpl::iterator i2); - - struct Utf8CharBuffer - { - char data[5]; - operator const char*() const { return data; } - }; - static Utf8CharBuffer EncodeChar(wxUniChar ch); - // returns n copies of ch encoded in UTF-8 string - static wxCharBuffer EncodeNChars(size_t n, wxUniChar ch); - - // returns the length of UTF-8 encoding of the character with lead byte 'c' - static size_t GetUtf8CharLength(char c) - { - wxASSERT( IsValidUtf8LeadByte(c) ); - return ms_utf8IterTable[(unsigned char)c]; - } + // 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); } - // decodes single UTF-8 character from UTF-8 string - // FIXME-UTF8: move EncodeChar/DecodeChar and other operations to separate - // class - static wxUniChar DecodeChar(wxStringImpl::const_iterator i) - { return wxUniCharRef::DecodeChar(i); } - friend class WXDLLIMPEXP_BASE wxUniCharRef; + 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 { @@ -548,21 +491,6 @@ private: else return const_iterator(m_impl.begin() + pos) - begin(); } - - // 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); } #endif // !wxUSE_UNICODE_UTF8/wxUSE_UNICODE_UTF8 public: @@ -602,33 +530,45 @@ public: 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) \ - { m_cur = wxString::AddToIter(m_cur, n); return *this; } \ + { \ + m_cur = wxStringOperations::AddToIter(m_cur, n); \ + return *this; \ + } \ iterator_name& operator+=(size_t n) \ - { m_cur = wxString::AddToIter(m_cur, (int)n); return *this; } \ + { \ + m_cur = wxStringOperations::AddToIter(m_cur, (int)n); \ + return *this; \ + } \ iterator_name& operator-=(int n) \ - { m_cur = wxString::AddToIter(m_cur, -n); return *this; } \ + { \ + m_cur = wxStringOperations::AddToIter(m_cur, -n); \ + return *this; \ + } \ iterator_name& operator-=(size_t n) \ - { m_cur = wxString::AddToIter(m_cur, -(int)n); return *this; } \ + { \ + m_cur = wxStringOperations::AddToIter(m_cur, -(int)n); \ + return *this; \ + } \ \ difference_type operator-(const iterator_name& i) const \ - { return wxString::DiffIters(m_cur, i.m_cur); } \ + { return wxStringOperations::DiffIters(m_cur, i.m_cur); } \ \ bool operator==(const iterator_name& i) const \ { return m_cur == i.m_cur; } \ @@ -672,13 +612,13 @@ public: iterator(const iterator& i) : m_cur(i.m_cur), m_str(i.m_str) {} iterator operator+(int n) const - { return iterator(m_str, wxString::AddToIter(m_cur, n)); } + { return iterator(m_str, wxStringOperations::AddToIter(m_cur, n)); } iterator operator+(size_t n) const - { return iterator(m_str, wxString::AddToIter(m_cur, (int)n)); } + { return iterator(m_str, wxStringOperations::AddToIter(m_cur, (int)n)); } iterator operator-(int n) const - { return iterator(m_str, wxString::AddToIter(m_cur, -n)); } + { return iterator(m_str, wxStringOperations::AddToIter(m_cur, -n)); } iterator operator-(size_t n) const - { return iterator(m_str, wxString::AddToIter(m_cur, -(int)n)); } + { return iterator(m_str, wxStringOperations::AddToIter(m_cur, -(int)n)); } private: iterator(wxString *str, underlying_iterator ptr) @@ -705,13 +645,13 @@ public: iterator(const iterator& i) : m_cur(i.m_cur) {} iterator operator+(int n) const - { return iterator(wxString::AddToIter(m_cur, n)); } + { return iterator(wxStringOperations::AddToIter(m_cur, n)); } iterator operator+(size_t n) const - { return iterator(wxString::AddToIter(m_cur, (int)n)); } + { return iterator(wxStringOperations::AddToIter(m_cur, (int)n)); } iterator operator-(int n) const - { return iterator(wxString::AddToIter(m_cur, -n)); } + { return iterator(wxStringOperations::AddToIter(m_cur, -n)); } iterator operator-(size_t n) const - { return iterator(wxString::AddToIter(m_cur, -(int)n)); } + { return iterator(wxStringOperations::AddToIter(m_cur, -(int)n)); } private: // for internal wxString use only: @@ -727,20 +667,20 @@ public: // 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, - wxString::DecodeChar(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(wxString::AddToIter(m_cur, n)); } + { return const_iterator(wxStringOperations::AddToIter(m_cur, n)); } const_iterator operator+(size_t n) const - { return const_iterator(wxString::AddToIter(m_cur, (int)n)); } + { return const_iterator(wxStringOperations::AddToIter(m_cur, (int)n)); } const_iterator operator-(int n) const - { return const_iterator(wxString::AddToIter(m_cur, -n)); } + { return const_iterator(wxStringOperations::AddToIter(m_cur, -n)); } const_iterator operator-(size_t n) const - { return const_iterator(wxString::AddToIter(m_cur, -(int)n)); } + { return const_iterator(wxStringOperations::AddToIter(m_cur, -(int)n)); } private: // for internal wxString use only: @@ -833,15 +773,33 @@ private: 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 - // FIXME-UTF8: this one needs to do UTF-8 conversion in UTF-8 build! - wxString(const wxStringImpl& stringSrc) : m_impl(stringSrc) { } - wxString(const wxString& stringSrc) : m_impl(stringSrc.m_impl) { } // string containing nRepeat copies of ch @@ -917,17 +875,50 @@ public: : 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 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 + // 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 && !wxUSE_STL_BASED_WXSTRING - wxString(const wxStdString& s) - // FIXME-UTF8: this one needs to do UTF-8 conversion in UTF-8 build! - : m_impl(s.c_str()) { } // FIXME-UTF8: this is broken for embedded 0s -#endif // wxUSE_STD_STRING && !wxUSE_STL_BASED_WXSTRING +#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()); } @@ -984,7 +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 @@ -1094,7 +1085,8 @@ public: // 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 @@ -1168,7 +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) @@ -1644,7 +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); @@ -1700,7 +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); @@ -1777,7 +1769,7 @@ public: { #if wxUSE_UNICODE_UTF8 if ( !ch.IsAscii() ) - m_impl.insert(PosToImpl(nPos), EncodeNChars(n, ch)); + m_impl.insert(PosToImpl(nPos), wxStringOperations::EncodeNChars(n, ch)); else #endif m_impl.insert(PosToImpl(nPos), n, (wxStringCharType)ch); @@ -1789,7 +1781,7 @@ public: if ( !ch.IsAscii() ) { size_t pos = IterToImplPos(it); - m_impl.insert(pos, EncodeChar(ch)); + m_impl.insert(pos, wxStringOperations::EncodeChar(ch)); return iterator(this, m_impl.begin() + pos); } else @@ -1811,7 +1803,7 @@ public: { #if wxUSE_UNICODE_UTF8 if ( !ch.IsAscii() ) - m_impl.insert(IterToImplPos(it), EncodeNChars(n, ch)); + m_impl.insert(IterToImplPos(it), wxStringOperations::EncodeNChars(n, ch)); else #endif m_impl.insert(it.impl(), n, (wxStringCharType)ch); @@ -1867,7 +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); @@ -1940,7 +1932,8 @@ public: { #if wxUSE_UNICODE_UTF8 if ( !ch.IsAscii() ) - m_impl.replace(first.impl(), last.impl(), EncodeNChars(n, ch)); + m_impl.replace(first.impl(), last.impl(), + wxStringOperations::EncodeNChars(n, ch)); else #endif m_impl.replace(first.impl(), last.impl(), n, (wxStringCharType)ch); @@ -1981,7 +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 @@ -2010,7 +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 @@ -2190,7 +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); } @@ -2426,7 +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) @@ -2613,7 +2612,9 @@ 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]; } // ----------------------------------------------------------------------------