Merge fixes for handling Unicode characters outside of BMP

Closes https://github.com/wxWidgets/wxWidgets/pull/467

src/common/strconv.cpp

@@ -80,27 +80,27 @@ static bool NotAllNULs(const char *p, size_t n)
 
 static size_t encode_utf16(wxUint32 input, wxUint16 *output)
 {
-    if (input <= 0xffff)
+    if (wxUniChar::IsBMP(input))
     {
         if (output)
             *output = (wxUint16) input;
 
         return 1;
     }
-    else if (input >= 0x110000)
-    {
-        return wxCONV_FAILED;
-    }
-    else
+    else if (wxUniChar::IsSupplementary(input))
     {
         if (output)
         {
-            *output++ = (wxUint16) ((input >> 10) + 0xd7c0);
-            *output = (wxUint16) ((input & 0x3ff) + 0xdc00);
+            *output++ = wxUniChar::HighSurrogate(input);
+            *output = wxUniChar::LowSurrogate(input);
         }
 
         return 2;
     }
+    else
+    {
+        return wxCONV_FAILED;
+    }
 }
 
 static size_t decode_utf16(const wxUint16* input, wxUint32& output)

src/common/stringops.cpp

@@ -27,6 +27,68 @@
 // implementation
 // ===========================================================================
 
+#if wxUSE_UNICODE_WCHAR || !wxUSE_UNICODE
+
+#if wxUSE_UNICODE_UTF16
+
+wxStringOperationsWchar::Utf16CharBuffer wxStringOperationsWchar::EncodeChar(const wxUniChar& ch)
+{
+    Utf16CharBuffer buf;
+    if ( ch.IsSupplementary() )
+    {
+        buf.data[0] = (wchar_t)ch.HighSurrogate();
+        buf.data[1] = (wchar_t)ch.LowSurrogate();
+    }
+    else
+    {
+        // Assume ch is a BMP character
+        buf.data[0] = (wchar_t)ch;
+    }
+    return buf;
+}
+
+wxWCharBuffer wxStringOperationsWchar::EncodeNChars(size_t n, const wxUniChar& ch)
+{
+    if ( ch.IsSupplementary() )
+    {
+        wxWCharBuffer buf(n * 2);
+        wchar_t s[2] = {
+            (wchar_t)ch.HighSurrogate(),
+            (wchar_t)ch.LowSurrogate(),
+        };
+        wchar_t *ptr = buf.data();
+        for (size_t i = 0; i < n; i++, ptr += 2)
+        {
+            wmemcpy(ptr, s, 2);
+        }
+        return buf;
+    }
+    else
+    {
+        // Assume ch is a BMP character
+        wxWCharBuffer buf(n);
+        wmemset(buf.data(), (wchar_t)ch, n);
+        return buf;
+    }
+}
+
+#else
+
+wxWxCharBuffer wxStringOperationsWchar::EncodeNChars(size_t n, const wxUniChar& ch)
+{
+    wxWxCharBuffer buf(n);
+#if wxUSE_UNICODE_WCHAR
+    wmemset(buf.data(), (wchar_t)ch, n);
+#else // ANSI
+    memset(buf.data(), (unsigned char)ch, n);
+#endif
+    return buf;
+}
+
+#endif // wxUSE_UNICODE_UTF16
+
+#endif // wxUSE_UNICODE_WCHAR || !wxUSE_UNICODE
+
 #if wxUSE_UNICODE_UTF8
 
 // ---------------------------------------------------------------------------

src/common/ustring.cpp

@@ -502,7 +502,7 @@ wxScopedU16CharBuffer wxUString::utf16_str() const
 
         // TODO: error range checks
 
-        if (code < 0x10000)
+        if (wxUniChar::IsBMP(code))
            utf16_length++;
         else
            utf16_length += 2;
@@ -520,15 +520,15 @@ wxScopedU16CharBuffer wxUString::utf16_str() const
 
         // TODO: error range checks
 
-        if (code < 0x10000)
+        if (wxUniChar::IsBMP(code))
         {
            out[0] = code;
            out++;
         }
         else
         {
-           out[0] = (code - 0x10000) / 0x400 + 0xd800;
-           out[1] = (code - 0x10000) % 0x400 + 0xdc00;
+           out[0] = wxUniChar::HighSurrogate(code);
+           out[1] = wxUniChar::LowSurrogate(code);
            out += 2;
         }
     }