Fixed context menu in wxRichTextCtrl

[wxWidgets.git] / include / wx / string.h

diff --git a/include/wx/string.h b/include/wx/string.h
index d60073694c437e1321546d234cde87c56ff8a82c..b70306b886a239975a99e78c592e3d45043f8dcf 100644 (file)
--- a/include/wx/string.h
+++ b/include/wx/string.h
@@ -51,14 +51,15 @@
     #include <StringMgr.h>
 #endif
 
     #include <StringMgr.h>
 #endif
 

-#include "wx/wxchar.h"      // for wxChar, wxStrlen() etc.
+#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/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"
 
 #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
 
 // unless this symbol is predefined to disable the compatibility functions, do
 // use them


@@ -77,6 +78,14 @@ class WXDLLIMPEXP_BASE wxString;
 #define   wxMBSTRINGCAST (char *)(const char *)
 #define   wxWCSTRINGCAST (wchar_t *)(const wchar_t *)
 
 #define   wxMBSTRINGCAST (char *)(const char *)
 #define   wxWCSTRINGCAST (wchar_t *)(const wchar_t *)
 

+// 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
 // ----------------------------------------------------------------------------
 // ----------------------------------------------------------------------------
 // constants
 // ----------------------------------------------------------------------------


@@ -185,8 +194,9 @@ private:
 public:
     // Ctor constructs the object from char literal; they are needed to make
     // operator?: compile and they intentionally take char*, not const char*
 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);
+    inline wxCStrData(char *buf);
+    inline wxCStrData(wchar_t *buf);
+    inline wxCStrData(const wxCStrData& data);

 
     inline ~wxCStrData();
 
 
     inline ~wxCStrData();
 


@@ -199,9 +209,7 @@ public:
     const wchar_t* AsWChar() const;
     operator const wchar_t*() const { return AsWChar(); }
 
     const wchar_t* AsWChar() const;
     operator const wchar_t*() const { return AsWChar(); }
 

-    inline operator bool() const;
-
-#if !wxUSE_UNICODE
+#if !wxUSE_UNICODE || wxUSE_UTF8_LOCALE_ONLY

     inline
 #endif
     const char* AsChar() const;
     inline
 #endif
     const char* AsChar() const;


@@ -217,6 +225,10 @@ public:
 
     inline wxString AsString() const;
 
 
     inline wxString AsString() const;
 

+    // returns the value as C string in internal representation (equivalent
+    // to AsString().wx_str(), but more efficient)
+    const wxStringCharType *AsInternal() const;
+

     // allow expressions like "c_str()[0]":
     inline wxUniChar operator[](size_t n) const;
     wxUniChar operator[](int n) const { return operator[](size_t(n)); }
     // allow expressions like "c_str()[0]":
     inline wxUniChar operator[](size_t n) const;
     wxUniChar operator[](int n) const { return operator[](size_t(n)); }


@@ -264,7 +276,7 @@ private:
     size_t m_offset;
     bool m_owned;
 
     size_t m_offset;
     bool m_owned;
 

-    friend class WXDLLIMPEXP_BASE wxString;
+    friend class WXDLLIMPEXP_FWD_BASE wxString;

 };
 
 // ----------------------------------------------------------------------------
 };
 
 // ----------------------------------------------------------------------------


@@ -293,8 +305,14 @@ class WXDLLIMPEXP_BASE wxStringPrintfMixinBase
 protected:
     wxStringPrintfMixinBase() {}
 
 protected:
     wxStringPrintfMixinBase() {}
 

-    int DoPrintf(const wxChar *format, ...) ATTRIBUTE_PRINTF_2;
-    static wxString DoFormat(const wxChar *format, ...) ATTRIBUTE_PRINTF_1;
+#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
 };
 
 // this class contains template wrappers for wxString's vararg methods, it's


@@ -317,13 +335,31 @@ public:
     // these are duplicated wxString methods, they're also declared below
     // if !wxNEEDS_WXSTRING_PRINTF_MIXIN:
 
     // 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<T1>::type,
-                          Format, DoFormat)
-    // int sprintf(const wxChar *pszFormat, ...) ATTRIBUTE_PRINTF_2;
-    WX_DEFINE_VARARG_FUNC(int, sprintf, DoPrintf)
+    // static wxString Format(const wString& format, ...) ATTRIBUTE_PRINTF_1;
+    WX_DEFINE_VARARG_FUNC_SANS_N0(static typename StringReturnType<T1>::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<typename T>
+    inline static typename StringReturnType<T>::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));
+    }
+
+    // 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)

 
 protected:
     wxStringPrintfMixin() : wxStringPrintfMixinBase() {}
 
 protected:
     wxStringPrintfMixin() : wxStringPrintfMixinBase() {}


@@ -342,6 +378,43 @@ protected:
     #pragma warning (disable:4275)
 #endif
 
     #pragma warning (disable:4275)
 #endif
 

+#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 wxString
 #ifdef wxNEEDS_WXSTRING_PRINTF_MIXIN
                                 : public wxStringPrintfMixin
 class WXDLLIMPEXP_BASE wxString
 #ifdef wxNEEDS_WXSTRING_PRINTF_MIXIN
                                 : public wxStringPrintfMixin


@@ -373,7 +446,6 @@ private:
 
   // buffer for holding temporary substring when using any of the methods
   // that take (char*,size_t) or (wchar_t*,size_t) arguments:
 
   // 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<typename T>
   struct SubstrBufFromType
   {
   template<typename T>
   struct SubstrBufFromType
   {


@@ -441,28 +513,6 @@ private:
     { return ConvertStr(str, n, wxConvLibc); }
   #endif
 
     { 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; }
   // translates position index in wxString to/from index in underlying
   // wxStringImpl:
   static size_t PosToImpl(size_t pos) { return pos; }


@@ -474,55 +524,17 @@ private:
 
 #else // wxUSE_UNICODE_UTF8
 
 
 #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];
-  }
+  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, wxMBConvUTF8()).data; }
+  static SubstrBufFromWC ImplStr(const wchar_t* str, size_t n)
+    { return ConvertStr(str, n, wxMBConvUTF8()); }

 
   size_t PosToImpl(size_t pos) const
   {
 
   size_t PosToImpl(size_t pos) const
   {


@@ -546,23 +558,8 @@ private:
       if ( pos == 0 || pos == npos )
           return pos;
       else
       if ( pos == 0 || pos == npos )
           return pos;
       else

-          return const_iterator(m_impl.begin() + pos) - begin();
+          return const_iterator(this, 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:
 #endif // !wxUSE_UNICODE_UTF8/wxUSE_UNICODE_UTF8
 
 public:


@@ -587,8 +584,7 @@ public:
   #define WX_STR_ITERATOR_TAG void /* dummy type */
 #endif
 
   #define WX_STR_ITERATOR_TAG void /* dummy type */
 #endif
 

-  #define WX_STR_ITERATOR_IMPL(iterator_name, pointer_type,                 \
-                               reference_type, reference_ctor)              \
+  #define WX_STR_ITERATOR_IMPL(iterator_name, pointer_type, reference_type) \

       private:                                                              \
           typedef wxStringImpl::iterator_name underlying_iterator;          \
       public:                                                               \
       private:                                                              \
           typedef wxStringImpl::iterator_name underlying_iterator;          \
       public:                                                               \


@@ -598,37 +594,48 @@ public:
           typedef reference_type reference;                                 \
           typedef pointer_type pointer;                                     \
                                                                             \
           typedef reference_type reference;                                 \
           typedef pointer_type pointer;                                     \
                                                                             \

-          reference operator*() const { return reference_ctor; }            \

           reference operator[](size_t n) const { return *(*this + n); }     \
                                                                             \
           iterator_name& operator++()                                       \
           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--()                                       \
           iterator_name& operator--()                                       \

-            { wxString::DecIter(m_cur); return *this; }                     \
+            { wxStringOperations::DecIter(m_cur); return *this; }           \

           iterator_name operator++(int)                                     \
           {                                                                 \
               iterator_name tmp = *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;                                    \
               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)                                  \
               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)                               \
           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)                                  \
           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)                               \
           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           \
                                                                             \
           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; }                                    \
                                                                             \
           bool operator==(const iterator_name& i) const                     \
             { return m_cur == i.m_cur; }                                    \


@@ -648,70 +655,127 @@ public:
           /* for internal wxString use only: */                             \
           underlying_iterator impl() const { return m_cur; }                \
                                                                             \
           /* for internal wxString use only: */                             \
           underlying_iterator impl() const { return m_cur; }                \
                                                                             \

-          friend class WXDLLIMPEXP_BASE wxString;                           \
-          friend class WXDLLIMPEXP_BASE wxCStrData;                         \
+          friend class wxString;                                            \
+          friend class wxCStrData;                                          \

                                                                             \
       private:                                                              \
           underlying_iterator m_cur
 
                                                                             \
       private:                                                              \
           underlying_iterator m_cur
 

-  class const_iterator;
+  class WXDLLIMPEXP_FWD_BASE const_iterator;

 
 #if wxUSE_UNICODE_UTF8
 
 #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.
+  // 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)

 
 

-      WX_STR_ITERATOR_IMPL(iterator, wxChar*, wxUniCharRef,
-                           wxUniCharRef::CreateForString(m_str, m_cur));
+  class WXDLLIMPEXP_BASE iterator
+  {
+      WX_STR_ITERATOR_IMPL(iterator, wxChar*, wxUniCharRef);

 
   public:
 
   public:

-      iterator(const iterator& i) : m_cur(i.m_cur), m_str(i.m_str) {}
+      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
 
       iterator operator+(int n) const

-        { return iterator(m_str, wxString::AddToIter(m_cur, n)); }
+        { return iterator(str(), wxStringOperations::AddToIter(m_cur, n)); }

       iterator operator+(size_t n) const
       iterator operator+(size_t n) const

-        { return iterator(m_str, wxString::AddToIter(m_cur, (int)n)); }
+        { return iterator(str(), wxStringOperations::AddToIter(m_cur, (int)n)); }

       iterator operator-(int n) const
       iterator operator-(int n) const

-        { return iterator(m_str, wxString::AddToIter(m_cur, -n)); }
+        { return iterator(str(), wxStringOperations::AddToIter(m_cur, -n)); }

       iterator operator-(size_t n) const
       iterator operator-(size_t n) const

-        { return iterator(m_str, wxString::AddToIter(m_cur, -(int)n)); }
+        { return iterator(str(), wxStringOperations::AddToIter(m_cur, -(int)n)); }

 
   private:
       iterator(wxString *str, underlying_iterator ptr)
 
   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) {}
+          : m_cur(ptr), m_node(str, &m_cur) {}

 
 

-      wxStringImpl& m_str;
+      wxString* str() const { return wx_const_cast(wxString*, m_node.m_str); }
+
+      wxStringIteratorNode m_node;

 
       friend class const_iterator;
   };
 
 
       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)); }
+
+  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;
+  };
+

   size_t IterToImplPos(wxString::iterator i) const
     { return wxStringImpl::const_iterator(i.impl()) - m_impl.begin(); }
 
 #else // !wxUSE_UNICODE_UTF8
 
   size_t IterToImplPos(wxString::iterator i) const
     { return wxStringImpl::const_iterator(i.impl()) - m_impl.begin(); }
 
 #else // !wxUSE_UNICODE_UTF8
 

-  class iterator
+  class WXDLLIMPEXP_BASE iterator

   {
   {

-      WX_STR_ITERATOR_IMPL(iterator, wxChar*, wxUniCharRef,
-                           wxUniCharRef::CreateForString(m_cur));
+      WX_STR_ITERATOR_IMPL(iterator, wxChar*, wxUniCharRef);

 
   public:
 
   public:

+      iterator() {}

       iterator(const iterator& i) : m_cur(i.m_cur) {}
 
       iterator(const iterator& i) : m_cur(i.m_cur) {}
 

+      reference operator*()
+        { return wxUniCharRef::CreateForString(m_cur); }
+

       iterator operator+(int n) const
       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
       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
       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
       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:
 
   private:
       // for internal wxString use only:


@@ -720,32 +784,37 @@ public:
 
       friend class const_iterator;
   };
 
       friend class const_iterator;
   };

-#endif // wxUSE_UNICODE_UTF8/!wxUSE_UNICODE_UTF8

 
 

-  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:
   {
       // 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));
+      WX_STR_ITERATOR_IMPL(const_iterator, const wxChar*, wxUniChar);

 
   public:
 
   public:

+      const_iterator() {}

       const_iterator(const const_iterator& i) : m_cur(i.m_cur) {}
       const_iterator(const iterator& i) : m_cur(i.m_cur) {}
 
       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
       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
       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
       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
       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:
       const_iterator(underlying_iterator ptr) : m_cur(ptr) {}
 
   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_TAG
   #undef WX_STR_ITERATOR_IMPL
 
   #undef WX_STR_ITERATOR_TAG
   #undef WX_STR_ITERATOR_IMPL


@@ -765,6 +834,7 @@ public:
       typedef typename T::reference reference;
       typedef typename T::pointer *pointer;
 
       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) {}
       reverse_iterator_impl(iterator_type i) : m_cur(i) {}
       reverse_iterator_impl(const reverse_iterator_impl& ri)
           : m_cur(ri.m_cur) {}


@@ -830,18 +900,37 @@ private:
   // wxCStrData) to an iterator into the string
   static const_iterator CreateConstIterator(const wxCStrData& data)
   {
   // wxCStrData) to an iterator into the string
   static const_iterator CreateConstIterator(const wxCStrData& data)
   {

-      return const_iterator(data.m_str->begin() + data.m_offset);
+      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
   wxString() {}
 
     // copy ctor
 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
   wxString(const wxString& stringSrc) : m_impl(stringSrc.m_impl) { }
 
     // string containing nRepeat copies of ch


@@ -914,26 +1003,61 @@ public:
     // and because wxString is convertible to wxCStrData and const wxChar *
     // we also need to provide this one
   wxString(const wxString& str, size_t nLength)
     // 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) {}
+    { assign(str, nLength); }

 
   // even if we're not built with wxUSE_STL == 1 it is very convenient to allow
 
   // 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
   // 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
+#endif // wxUSE_STD_STRING
+
+  // 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
 
   // first valid index position

-  const_iterator begin() const { return const_iterator(m_impl.begin()); }
+  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
   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()); }
+  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
   iterator end() { return iterator(this, m_impl.end()); }
 
   // first element of the reversed string


@@ -964,19 +1088,25 @@ public:
 
   void resize(size_t nSize, wxUniChar ch = wxT('\0'))
   {
 
   void resize(size_t nSize, wxUniChar ch = wxT('\0'))
   {

+    const size_t len = length();
+    if ( nSize == len)
+        return;
+

 #if wxUSE_UNICODE_UTF8
 #if wxUSE_UNICODE_UTF8

-    if ( !ch.IsAscii() )
+    if ( nSize < len )

     {
     {

-        size_t len = length();
-        if ( nSize == len)
-            return;
-        else if ( nSize < len )
-            erase(nSize);
-        else
-            append(nSize - len, ch);
+        // we can't use wxStringImpl::resize() for truncating the string as it
+        // counts in bytes, not characters
+        erase(nSize);
+        return;

     }
     }

+
+    // 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
     else

-#endif
+#endif // wxUSE_UNICODE_UTF8

         m_impl.resize(nSize, (wxStringCharType)ch);
   }
 
         m_impl.resize(nSize, (wxStringCharType)ch);
   }
 


@@ -984,7 +1114,7 @@ public:
   {
     size_t pos, len;
     PosLenToImpl(nStart, nLen, &pos, &len);
   {
     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
   }
 
   // generic attributes & operations


@@ -1035,18 +1165,17 @@ public:
 
     // get last character
     wxUniChar Last() const
 
     // get last character
     wxUniChar Last() const

-      {
-          wxASSERT_MSG( !empty(), _T("wxString: index out of bounds") );
-
-          return at(length() - 1);
-      }
+    {
+      wxASSERT_MSG( !empty(), _T("wxString: index out of bounds") );
+      return *rbegin();
+    }

 
     // get writable last character
     wxUniCharRef Last()
 
     // 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
 
     /*
        Note that we we must define all of the overloads below to avoid


@@ -1081,9 +1210,22 @@ public:
 
     // implicit conversion to C string
     operator wxCStrData() const { return c_str(); }
 
     // implicit conversion to C string
     operator wxCStrData() 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(); }
 
     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(); }
 
     // identical to c_str(), for MFC compatibility
     const wxCStrData GetData() const { return c_str(); }
 


@@ -1094,7 +1236,8 @@ 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
     // 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
     wxWritableWCharBuffer wchar_str() const { return wc_str(); }
 
     // conversion to/from plain (i.e. 7 bit) ASCII: this is useful for


@@ -1104,15 +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
     // 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 ); }
     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
 
     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
     //
     // conversions with (possible) format conversions: have to return a
     // buffer with temporary data
     //


@@ -1123,7 +1336,13 @@ 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
     // 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;
     const wxCharBuffer mb_str(const wxMBConv& conv = wxConvLibc) const;

+#endif

 
     const wxWX2MBbuf mbc_str() const { return mb_str(*wxConvCurrent); }
 
 
     const wxWX2MBbuf mbc_str() const { return mb_str(*wxConvCurrent); }
 


@@ -1168,7 +1387,7 @@ public:
     { return *this = cstr.AsString(); }
     // from a character
   wxString& operator=(wxUniChar ch)
     { 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)
   wxString& operator=(wxUniCharRef ch)
     { return operator=((wxUniChar)ch); }
   wxString& operator=(char ch)


@@ -1235,13 +1454,8 @@ public:
       // string += buffer (i.e. from wxGetString)
   wxString& operator<<(const wxWCharBuffer& s)
     { return operator<<((const wchar_t *)s); }
       // string += buffer (i.e. from wxGetString)
   wxString& operator<<(const wxWCharBuffer& s)
     { return operator<<((const wchar_t *)s); }

-  wxString& operator+=(const wxWCharBuffer& s)
-    { return operator<<((const wchar_t *)s); }
-

   wxString& operator<<(const wxCharBuffer& s)
     { return operator<<((const char *)s); }
   wxString& operator<<(const wxCharBuffer& s)
     { return operator<<((const char *)s); }

-  wxString& operator+=(const wxCharBuffer& s)
-    { return operator<<((const char *)s); }

 
     // string += C string
   wxString& Append(const wxString& s)
 
     // string += C string
   wxString& Append(const wxString& s)


@@ -1253,12 +1467,16 @@ public:
             append(s);
         return *this;
     }
             append(s);
         return *this;
     }

-  wxString& Append(const wxCStrData& psz)
-    { append(psz); return *this; }

   wxString& Append(const char* psz)
     { append(psz); return *this; }
   wxString& Append(const wchar_t* pwz)
     { append(pwz); 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 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)
     { append(count, ch); return *this; }
     // append count copies of given character
   wxString& Append(wxUniChar ch, size_t count = 1u)
     { append(count, ch); return *this; }


@@ -1340,12 +1558,14 @@ public:
     { return compare(pwz); }
   int Cmp(const wxString& s) const
     { return compare(s); }
     { 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 wxString& s) const;
     // same as Cmp() but not case-sensitive
   int CmpNoCase(const wxString& s) const;

-  int CmpNoCase(const char *psz) const
-    { return CmpNoCase(wxString(psz)); }
-  int CmpNoCase(const wchar_t *pwz) const
-    { return CmpNoCase(wxString(pwz)); }

     // test for the string equality, either considering case or not
     // (if compareWithCase then the case matters)
   bool IsSameAs(const wxString& str, bool compareWithCase = true) const
     // test for the string equality, either considering case or not
     // (if compareWithCase then the case matters)
   bool IsSameAs(const wxString& str, bool compareWithCase = true) const


@@ -1354,12 +1574,14 @@ public:
     { return (compareWithCase ? Cmp(str) : CmpNoCase(str)) == 0; }
   bool IsSameAs(const wchar_t *str, bool compareWithCase = true) const
     { return (compareWithCase ? Cmp(str) : CmpNoCase(str)) == 0; }
     { 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
     // 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); }
   // FIXME-UTF8: remove these overloads
   bool IsSameAs(wxUniCharRef c, bool compareWithCase = true) const
     { return IsSameAs(wxUniChar(c), compareWithCase); }


@@ -1384,11 +1606,11 @@ 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
       // 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
       // 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;
 
       // get first nCount characters
   wxString Left(size_t nCount) const;


@@ -1440,7 +1662,6 @@ public:
   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(wchar_t ch, bool bFromEnd = false) const
     { return Find(wxUniChar(ch), bFromEnd); }
       // searching (return starting index, or -1 if not found)

-  // FIXME-UTF8: keep wxString overload only

   int Find(const wxString& sub) const               // like strstr
   {
     size_type idx = find(sub);
   int Find(const wxString& sub) const               // like strstr
   {
     size_type idx = find(sub);


@@ -1457,6 +1678,13 @@ public:
     return (idx == npos) ? wxNOT_FOUND : (int)idx;
   }
 
     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 wxString& strOld,
       // replace first (or all of bReplaceAll) occurences of substring with
       // another string, returns the number of replacements made
   size_t Replace(const wxString& strOld,


@@ -1489,16 +1717,40 @@ public:
   // formatted input/output
     // as sprintf(), returns the number of characters written or < 0 on error
     // (take 'this' into account in attribute parameter count)
   // 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
 #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);
 #endif
     // the same as above, but takes a va_list
   static wxString FormatV(const wxString& format, va_list argptr);


@@ -1515,7 +1767,7 @@ public:
     //
     // get writable buffer of at least nLen bytes. Unget() *must* be called
     // a.s.a.p. to put string back in a reasonable state!
     //
     // 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) );
     // call this immediately after GetWriteBuf() has been used
   wxDEPRECATED( void UngetWriteBuf() );
   wxDEPRECATED( void UngetWriteBuf(size_t nLen) );


@@ -1534,8 +1786,20 @@ public:
 #ifndef wxNEEDS_WXSTRING_PRINTF_MIXIN
   // use Printf()
   // (take 'this' into account in attribute parameter count)
 #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()
 #endif // wxNEEDS_WXSTRING_PRINTF_MIXIN
 
     // use Cmp()


@@ -1569,7 +1833,6 @@ public:
   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( 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( wxUniChar ch ) const { return Find(ch, true); }
   bool Contains(const wxString& str) const { return Find(str) != wxNOT_FOUND; }
   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; }


@@ -1581,7 +1844,7 @@ public:
 
     // take nLen chars starting at nPos
   wxString(const wxString& str, size_t nPos, size_t nLen)
 
     // take nLen chars starting at nPos
   wxString(const wxString& str, size_t nPos, size_t nLen)

-      : m_impl(str.m_impl, nPos, nLen) { }
+      { assign(str, nPos, nLen); }

     // take all characters from first to last
   wxString(const_iterator first, const_iterator last)
       : m_impl(first.impl(), last.impl()) { }
     // take all characters from first to last
   wxString(const_iterator first, const_iterator last)
       : m_impl(first.impl(), last.impl()) { }


@@ -1620,8 +1883,6 @@ public:
     // append a string
   wxString& append(const wxString& str)
     { m_impl.append(str.m_impl); return *this; }
     // append a string
   wxString& append(const wxString& str)
     { m_impl.append(str.m_impl); return *this; }

-  wxString& append(const wxCStrData& str)
-    { m_impl.append(str.AsString().m_impl); return *this; }

     // append first n (or all if n == npos) characters of sz
   wxString& append(const char *sz)
     { m_impl.append(ImplStr(sz)); return *this; }
     // append first n (or all if n == npos) characters of sz
   wxString& append(const char *sz)
     { m_impl.append(ImplStr(sz)); return *this; }


@@ -1639,12 +1900,20 @@ public:
     m_impl.append(str.data, str.len);
     return *this;
   }
     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)
   {
 #if wxUSE_UNICODE_UTF8
     if ( !ch.IsAscii() )
     // append n copies of ch
   wxString& append(size_t n, wxUniChar ch)
   {
 #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);
     else
 #endif
         m_impl.append(n, (wxStringCharType)ch);


@@ -1695,12 +1964,26 @@ public:
     m_impl.assign(str.data, str.len);
     return *this;
   }
     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)
   {
 #if wxUSE_UNICODE_UTF8
     if ( !ch.IsAscii() )
     // same as `= n copies of ch'
   wxString& assign(size_t n, wxUniChar ch)
   {
 #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);
     else
 #endif
         m_impl.assign(n, (wxStringCharType)ch);


@@ -1730,14 +2013,19 @@ public:
 
     // string comparison
   int compare(const wxString& str) const;
 
     // string comparison
   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 wxString& str) const;
     // comparison of 2 substrings
   int compare(size_t nStart, size_t nLen,
               const wxString& str, size_t nStart2, size_t nLen2) const;
     // comparison with a substring
   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 wxString& str, size_t nStart2, size_t nLen2) const;

-    // just like strcmp()
-  int compare(const char* sz) const;
-  int compare(const wchar_t* sz) const;

     // substring comparison with first nCount characters of sz
   int compare(size_t nStart, size_t nLen,
               const char* sz, size_t nCount = npos) const;
     // substring comparison with first nCount characters of sz
   int compare(size_t nStart, size_t nLen,
               const char* sz, size_t nCount = npos) const;


@@ -1777,7 +2065,7 @@ public:
   {
 #if wxUSE_UNICODE_UTF8
     if ( !ch.IsAscii() )
   {
 #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);
     else
 #endif
         m_impl.insert(PosToImpl(nPos), n, (wxStringCharType)ch);


@@ -1789,7 +2077,7 @@ public:
     if ( !ch.IsAscii() )
     {
         size_t pos = IterToImplPos(it);
     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
         return iterator(this, m_impl.begin() + pos);
     }
     else


@@ -1811,7 +2099,7 @@ public:
   {
 #if wxUSE_UNICODE_UTF8
     if ( !ch.IsAscii() )
   {
 #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);
     else
 #endif
         m_impl.insert(it.impl(), n, (wxStringCharType)ch);


@@ -1867,7 +2155,7 @@ public:
     PosLenToImpl(nStart, nLen, &from, &len);
 #if wxUSE_UNICODE_UTF8
     if ( !ch.IsAscii() )
     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);
     else
 #endif
         m_impl.replace(from, len, nCount, (wxStringCharType)ch);


@@ -1940,7 +2228,8 @@ public:
   {
 #if wxUSE_UNICODE_UTF8
     if ( !ch.IsAscii() )
   {
 #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);
     else
 #endif
         m_impl.replace(first.impl(), last.impl(), n, (wxStringCharType)ch);


@@ -1978,10 +2267,19 @@ public:
       SubstrBufFromWC str(ImplStr(sz, n));
       return PosFromImpl(m_impl.find(str.data, PosToImpl(nStart), str.len));
   }
       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
 
     // 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
   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


@@ -2008,9 +2306,18 @@ public:
       SubstrBufFromWC str(ImplStr(sz, n));
       return PosFromImpl(m_impl.rfind(str.data, PosToImpl(nStart), str.len));
   }
       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
     // 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
   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


@@ -2083,9 +2390,9 @@ 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
     // 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); }
+    { return find_first_of(str.wc_str(), nStart); }

 #else
 #else

-    { return find_first_of((const wxChar*)str.c_str(), nStart); }
+    { 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;
 #endif
     // same as above
   size_t find_first_of(const char* sz, size_t nStart = 0) const;


@@ -2098,9 +2405,9 @@ public:
     // 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
     // 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); }
+    { return find_last_of(str.wc_str(), nStart); }

 #else
 #else

-    { return find_last_of((const wxChar*)str.c_str(), nStart); }
+    { 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;
 #endif
     // same as above
   size_t find_last_of (const char* sz, size_t nStart = npos) const;


@@ -2116,9 +2423,9 @@ 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
     // 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); }
+    { return find_first_not_of(str.wc_str(), nStart); }

 #else
 #else

-    { return find_first_not_of((const wxChar*)str.c_str(), nStart); }
+    { 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;
 #endif
     // same as above
   size_t find_first_not_of(const char* sz, size_t nStart = 0) const;


@@ -2130,9 +2437,9 @@ public:
     //  as strcspn()
   size_t find_last_not_of(const wxString& str, size_t nStart = npos) const
 #if wxUSE_UNICODE // FIXME-UTF8: temporary
     //  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); }
+    { return find_last_not_of(str.wc_str(), nStart); }

 #else
 #else

-    { return find_last_not_of((const wxChar*)str.c_str(), nStart); }
+    { 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;
 #endif
     // same as above
   size_t find_last_not_of(const char* sz, size_t nStart = npos) const;


@@ -2178,6 +2485,59 @@ public:
   size_t find_last_not_of(wchar_t 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)
     { m_impl += s.m_impl; return *this; }
       // string += string
   wxString& operator+=(const wxString& s)
     { m_impl += s.m_impl; return *this; }


@@ -2188,9 +2548,13 @@ public:
     { m_impl += ImplStr(pwz); return *this; }
   wxString& operator+=(const wxCStrData& s)
     { m_impl += s.AsString().m_impl; return *this; }
     { m_impl += ImplStr(pwz); return *this; }
   wxString& operator+=(const wxCStrData& s)
     { 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)
       // 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); }
   wxString& operator+=(wxUniCharRef ch) { return *this += wxUniChar(ch); }
   wxString& operator+=(int ch) { return *this += wxUniChar(ch); }
   wxString& operator+=(char ch) { return *this += wxUniChar(ch); }


@@ -2198,7 +2562,7 @@ public:
   wxString& operator+=(wchar_t ch) { return *this += wxUniChar(ch); }
 
 private:
   wxString& operator+=(wchar_t ch) { return *this += wxUniChar(ch); }
 
 private:

-#if !wxUSE_STL_BASED_WXSTRING && !wxUSE_UNICODE_UTF8
+#if !wxUSE_STL_BASED_WXSTRING

   // helpers for wxStringBuffer and wxStringBufferLength
   wxStringCharType *DoGetWriteBuf(size_t nLen)
     { return m_impl.DoGetWriteBuf(nLen); }
   // helpers for wxStringBuffer and wxStringBufferLength
   wxStringCharType *DoGetWriteBuf(size_t nLen)
     { return m_impl.DoGetWriteBuf(nLen); }


@@ -2206,14 +2570,17 @@ private:
     { m_impl.DoUngetWriteBuf(); }
   void DoUngetWriteBuf(size_t nLen)
     { m_impl.DoUngetWriteBuf(nLen); }
     { m_impl.DoUngetWriteBuf(); }
   void DoUngetWriteBuf(size_t nLen)
     { m_impl.DoUngetWriteBuf(nLen); }

-
-  friend class WXDLLIMPEXP_BASE wxStringBuffer;
-  friend class WXDLLIMPEXP_BASE wxStringBufferLength;
-#endif // !wxUSE_STL_BASED_WXSTRING && !wxUSE_UNICODE_UTF8
+#endif // !wxUSE_STL_BASED_WXSTRING

 
 #ifndef wxNEEDS_WXSTRING_PRINTF_MIXIN
 
 #ifndef wxNEEDS_WXSTRING_PRINTF_MIXIN

-  int DoPrintf(const wxChar *format, ...) ATTRIBUTE_PRINTF_2;
-  static wxString DoFormat(const wxChar *format, ...) ATTRIBUTE_PRINTF_1;
+  #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
 
 #if !wxUSE_STL_BASED_WXSTRING
 #endif
 
 #if !wxUSE_STL_BASED_WXSTRING


@@ -2224,6 +2591,12 @@ private:
 private:
   wxStringImpl m_impl;
 
 private:
   wxStringImpl m_impl;
 

+#ifdef __VISUALC__
+    // "struct 'ConvertedBuffer<T>' 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
   // buffers for compatibility conversion from (char*)c_str() and
   // (wchar_t*)c_str():
   // FIXME-UTF8: bechmark various approaches to keeping compatibility buffers


@@ -2245,13 +2618,41 @@ private:
 
       T *m_buf;
   };
 
       T *m_buf;
   };

-#if wxUSE_UNICODE
+#if wxUSE_UNICODE && !wxUSE_UTF8_LOCALE_ONLY

   ConvertedBuffer<char> m_convertedToChar;
 #endif
 #if !wxUSE_UNICODE_WCHAR
   ConvertedBuffer<wchar_t> m_convertedToWChar;
 #endif
   ConvertedBuffer<char> m_convertedToChar;
 #endif
 #if !wxUSE_UNICODE_WCHAR
   ConvertedBuffer<wchar_t> m_convertedToWChar;
 #endif

-  friend class WXDLLIMPEXP_BASE wxCStrData;
+
+#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
 };
 
 #ifdef wxNEEDS_WXSTRING_PRINTF_MIXIN


@@ -2304,129 +2705,216 @@ inline wxString operator+(wchar_t ch, const wxString& string)
     { return wxUniChar(ch) + string; }
 
 
     { return wxUniChar(ch) + string; }
 
 

-#if wxUSE_STL_BASED_WXSTRING
-    // return an empty wxString (not very useful with wxUSE_STL == 1)
-    inline const wxString wxGetEmptyString() { return wxString(); }
-#else // !wxUSE_STL_BASED_WXSTRING
-    // return an empty wxString (more efficient than wxString() here)
-    inline const wxString& wxGetEmptyString()
-    {
-        return *(wxString *)&wxEmptyString;
-    }
-#endif // wxUSE_STL_BASED_WXSTRING/!wxUSE_STL_BASED_WXSTRING
+#define wxGetEmptyString() wxString()

 
 // ----------------------------------------------------------------------------
 // wxStringBuffer: a tiny class allowing to get a writable pointer into string
 // ----------------------------------------------------------------------------
 
 
 // ----------------------------------------------------------------------------
 // wxStringBuffer: a tiny class allowing to get a writable pointer into string
 // ----------------------------------------------------------------------------
 

-#if wxUSE_STL_BASED_WXSTRING || wxUSE_UNICODE_UTF8
-
-class WXDLLIMPEXP_BASE wxStringBuffer
+#if !wxUSE_STL_BASED_WXSTRING
+// string buffer for direct access to string data in their native
+// representation:
+class wxStringInternalBuffer

 {
 public:
 {
 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:
 
 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:
 {
 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);
     {
         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:
     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_BASED_WXSTRING && !wxUSE_UNICODE_UTF8
+#endif // !wxUSE_STL_BASED_WXSTRING

 
 

-class WXDLLIMPEXP_BASE wxStringBuffer
+template<typename T>
+class WXDLLIMPEXP_BASE wxStringTypeBufferBase

 {
 public:
 {
 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;
     wxString& m_str;

-    wxChar   *m_buf;
+    wxCharTypeBuffer<CharType> m_buf;
+};
+
+template<typename T>
+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)
+        { }

 
 

-    DECLARE_NO_COPY_CLASS(wxStringBuffer)
+    operator CharType*() { return m_buf.data(); }
+    void SetLength(size_t length) { m_len = length; m_lenSet = true; }
+
+protected:
+    wxString& m_str;
+    wxCharTypeBuffer<CharType> m_buf;
+    size_t        m_len;
+    bool          m_lenSet;

 };
 
 };
 

-class WXDLLIMPEXP_BASE wxStringBufferLength
+template<typename T>
+class wxStringTypeBuffer : public wxStringTypeBufferBase<T>

 {
 public:
 {
 public:

-    wxStringBufferLength(wxString& str, size_t lenWanted = 1024)
-        : m_str(str), m_buf(NULL), m_len(0), m_lenSet(false)
+    wxStringTypeBuffer(wxString& str, size_t lenWanted = 1024)
+        : wxStringTypeBufferBase<T>(str, lenWanted) {}
+    ~wxStringTypeBuffer()

     {
     {

-        m_buf = m_str.DoGetWriteBuf(lenWanted);
-        wxASSERT(m_buf != NULL);
+        this->m_str.assign(this->m_buf.data());

     }
 
     }
 

-    ~wxStringBufferLength()
+    DECLARE_NO_COPY_CLASS(wxStringTypeBuffer)
+};
+
+template<typename T>
+class wxStringTypeBufferLength : public wxStringTypeBufferLengthBase<T>
+{
+public:
+    wxStringTypeBufferLength(wxString& str, size_t lenWanted = 1024)
+        : wxStringTypeBufferLengthBase<T>(str, lenWanted) {}
+
+    ~wxStringTypeBufferLength()

     {
     {

-        wxASSERT(m_lenSet);
-        m_str.DoUngetWriteBuf(m_len);
+        wxASSERT(this->m_lenSet);
+        this->m_str.assign(this->m_buf.data(), this->m_len);

     }
 
     }
 

-    operator wxChar*() const { return m_buf; }
-    void SetLength(size_t length) { m_len = length; m_lenSet = true; }
+    DECLARE_NO_COPY_CLASS(wxStringTypeBufferLength)
+};

 
 

-private:
-    wxString& m_str;
-    wxChar   *m_buf;
-    size_t    m_len;
-    bool      m_lenSet;
+#if wxUSE_STL_BASED_WXSTRING

 
 

-    DECLARE_NO_COPY_CLASS(wxStringBufferLength)
+WXDLLIMPEXP_TEMPLATE_INSTANCE_BASE( wxStringTypeBufferBase<wxStringCharType> )
+
+class wxStringInternalBuffer : public wxStringTypeBufferBase<wxStringCharType>
+{
+public:
+    wxStringInternalBuffer(wxString& str, size_t lenWanted = 1024)
+        : wxStringTypeBufferBase<wxStringCharType>(str, lenWanted) {}
+    ~wxStringInternalBuffer()
+        { m_str.m_impl.assign(m_buf.data()); }
+
+    DECLARE_NO_COPY_CLASS(wxStringInternalBuffer)

 };
 
 };
 

+WXDLLIMPEXP_TEMPLATE_INSTANCE_BASE(
+    wxStringTypeBufferLengthBase<wxStringCharType> )
+
+class wxStringInternalBufferLength
+    : public wxStringTypeBufferLengthBase<wxStringCharType>
+{
+public:
+    wxStringInternalBufferLength(wxString& str, size_t lenWanted = 1024)
+        : wxStringTypeBufferLengthBase<wxStringCharType>(str, lenWanted) {}
+
+    ~wxStringInternalBufferLength()
+    {
+        wxASSERT(m_lenSet);
+        m_str.m_impl.assign(m_buf.data(), m_len);
+    }
+
+    DECLARE_NO_COPY_CLASS(wxStringInternalBufferLength)
+};
+#endif // wxUSE_STL_BASED_WXSTRING
+
+
+#if wxUSE_STL_BASED_WXSTRING || wxUSE_UNICODE_UTF8
+typedef wxStringTypeBuffer<wxChar>        wxStringBuffer;
+typedef wxStringTypeBufferLength<wxChar>  wxStringBufferLength;
+#else // if !wxUSE_STL_BASED_WXSTRING && !wxUSE_UNICODE_UTF8
+typedef wxStringInternalBuffer                wxStringBuffer;
+typedef wxStringInternalBufferLength          wxStringBufferLength;

 #endif // !wxUSE_STL_BASED_WXSTRING && !wxUSE_UNICODE_UTF8
 
 #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<char> )
+
+class WXDLLIMPEXP_BASE wxUTF8StringBuffer : public wxStringTypeBufferBase<char>
+{
+public:
+    wxUTF8StringBuffer(wxString& str, size_t lenWanted = 1024)
+        : wxStringTypeBufferBase<char>(str, lenWanted) {}
+    ~wxUTF8StringBuffer();
+
+    DECLARE_NO_COPY_CLASS(wxUTF8StringBuffer)
+};
+
+WXDLLIMPEXP_TEMPLATE_INSTANCE_BASE( wxStringTypeBufferLengthBase<char> )
+
+class WXDLLIMPEXP_BASE wxUTF8StringBufferLength
+    : public wxStringTypeBufferLengthBase<char>
+{
+public:
+    wxUTF8StringBufferLength(wxString& str, size_t lenWanted = 1024)
+        : wxStringTypeBufferLengthBase<char>(str, lenWanted) {}
+    ~wxUTF8StringBufferLength();
+
+    DECLARE_NO_COPY_CLASS(wxUTF8StringBufferLength)
+};
+#endif // wxUSE_UNICODE_UTF8/wxUSE_UNICODE_WCHAR
+
+

 // ---------------------------------------------------------------------------
 // wxString comparison functions: operator versions are always case sensitive
 // ---------------------------------------------------------------------------
 
 // ---------------------------------------------------------------------------
 // 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)
 
 
 wxDEFINE_ALL_COMPARISONS(const wxChar *, const wxString&, wxCMP_WXCHAR_STRING)
 


@@ -2447,7 +2935,15 @@ inline bool operator<=(const wxString& s1, const wxString& s2)
 inline bool operator>=(const wxString& s1, const wxString& s2)
     { return s1.Cmp(s2) >= 0; }
 
 inline bool operator>=(const wxString& s1, const wxString& s2)
     { return s1.Cmp(s2) >= 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)
 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)


@@ -2456,7 +2952,7 @@ 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); }
     { 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)
 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)


@@ -2465,19 +2961,16 @@ 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); }
     { 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; }
 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; }
 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

 
 // comparison with char
 inline bool operator==(const wxUniChar& c, const wxString& s) { return s.IsSameAs(c); }
 
 // comparison with char
 inline bool operator==(const wxUniChar& c, const wxString& s) { return s.IsSameAs(c); }


@@ -2539,7 +3032,15 @@ WXDLLIMPEXP_BASE wxSTD ostream& operator<<(wxSTD ostream&, const wxCharBuffer&);
 WXDLLIMPEXP_BASE wxSTD ostream& operator<<(wxSTD ostream&, const wxWCharBuffer&);
 #endif
 
 WXDLLIMPEXP_BASE wxSTD ostream& operator<<(wxSTD ostream&, const wxWCharBuffer&);
 #endif
 

-#endif  // wxSTD_STRING_COMPATIBILITY
+#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  // wxUSE_STD_IOSTREAM

 
 // ---------------------------------------------------------------------------
 // wxCStrData implementation
 
 // ---------------------------------------------------------------------------
 // wxCStrData implementation


@@ -2550,15 +3051,17 @@ 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(wchar_t *buf)
     : m_str(new wxString(buf)), m_offset(0), m_owned(true) {}
 

-inline wxCStrData::~wxCStrData()
+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)

 {
 {

-    if ( m_owned )
-        delete m_str;

 }
 
 }
 

-inline wxCStrData::operator bool() const
+inline wxCStrData::~wxCStrData()

 {
 {

-    return !m_str->empty();
+    if ( m_owned )
+        delete wx_const_cast(wxString*, m_str); // cast to silence warnings

 }
 
 // simple cases for AsChar() and AsWChar(), the complicated ones are
 }
 
 // simple cases for AsChar() and AsWChar(), the complicated ones are


@@ -2577,6 +3080,13 @@ inline const char* wxCStrData::AsChar() const
 }
 #endif // !wxUSE_UNICODE
 
 }
 #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
 inline const wxCharBuffer wxCStrData::AsCharBuf() const
 {
 #if !wxUSE_UNICODE


@@ -2603,6 +3113,15 @@ inline wxString wxCStrData::AsString() const
         return m_str->Mid(m_offset);
 }
 
         return m_str->Mid(m_offset);
 }
 

+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
 {
     if ( m_str->empty() )
 inline wxUniChar wxCStrData::operator*() const
 {
     if ( m_str->empty() )


@@ -2613,7 +3132,9 @@ inline wxUniChar wxCStrData::operator*() const
 
 inline wxUniChar wxCStrData::operator[](size_t n) 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];

 }
 
 // ----------------------------------------------------------------------------
 }
 
 // ----------------------------------------------------------------------------


@@ -2647,4 +3168,54 @@ inline wxWCharBuffer::wxWCharBuffer(const wxCStrData& cstr)
 {
 }
 
 {
 }
 

+#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_WXSTRING_H_
 #endif  // _WX_WXSTRING_H_