Implement SCI_TARGETASUTF8 and SCI_ENCODEDFROMUTF8 on Cocoa.

3 files changed, 94 insertions, 1 deletions

diff --git a/cocoa/ScintillaCocoa.h b/cocoa/ScintillaCocoa.h
index 384282270..45d04e564 100644
--- a/cocoa/ScintillaCocoa.h
+++ b/cocoa/ScintillaCocoa.h
@@ -105,6 +105,8 @@ private:
 
   bool GetPasteboardData(NSPasteboard* board, SelectionText* selectedText);
   void SetPasteboardData(NSPasteboard* board, const SelectionText& selectedText);
+  int TargetAsUTF8(char *text);
+  int EncodedFromUTF8(char *utf8, char *encoded) const;
 
   int scrollSpeed;
   int scrollTicks;
diff --git a/cocoa/ScintillaCocoa.mm b/cocoa/ScintillaCocoa.mm
index 45ce0cc86..f248f3bb0 100644
--- a/cocoa/ScintillaCocoa.mm
+++ b/cocoa/ScintillaCocoa.mm
@@ -526,6 +526,26 @@ static char *EncodedBytes(CFStringRef cfsRef, CFStringEncoding encoding) {
 //--------------------------------------------------------------------------------------------------
 
 /**
+ * Convert a core foundation string into a std::string in a particular encoding
+ */
+
+static std::string EncodedBytesString(CFStringRef cfsRef, CFStringEncoding encoding) {
+	const CFRange rangeAll = {0, CFStringGetLength(cfsRef)};
+	CFIndex usedLen = 0;
+	CFStringGetBytes(cfsRef, rangeAll, encoding, '?', false,
+                         NULL, 0, &usedLen);
+	
+	std::string buffer(usedLen, '\0');
+	if (usedLen > 0) {
+		CFStringGetBytes(cfsRef, rangeAll, encoding, '?', false,
+                                 reinterpret_cast<UInt8 *>(&buffer[0]), usedLen, NULL);
+	}
+	return buffer;
+}
+
+//--------------------------------------------------------------------------------------------------
+
+/**
  * Case folders.
  */
 
@@ -825,6 +845,13 @@ sptr_t ScintillaCocoa::WndProc(unsigned int iMessage, uptr_t wParam, sptr_t lPar
     case SCI_GETDIRECTPOINTER:
       return reinterpret_cast<sptr_t>(this);
 
+    case SCI_TARGETASUTF8:
+      return TargetAsUTF8(reinterpret_cast<char*>(lParam));
+
+    case SCI_ENCODEDFROMUTF8:
+      return EncodedFromUTF8(reinterpret_cast<char*>(wParam),
+                             reinterpret_cast<char*>(lParam));
+
     case SCI_SETIMEINTERACTION:
       // Only inline IME supported on Cocoa
       break;
@@ -1543,6 +1570,68 @@ bool ScintillaCocoa::GetPasteboardData(NSPasteboard* board, SelectionText* selec
 
 //--------------------------------------------------------------------------------------------------
 
+// Returns the target converted to UTF8.
+// Return the length in bytes.
+int ScintillaCocoa::TargetAsUTF8(char *text)
+{
+  const int targetLength = targetEnd - targetStart;
+  if (IsUnicodeMode())
+  {
+    if (text)
+      pdoc->GetCharRange(text, targetStart, targetLength);
+  }
+  else
+  {
+    // Need to convert
+    const CFStringEncoding encoding = EncodingFromCharacterSet(IsUnicodeMode(),
+                                                         vs.styles[STYLE_DEFAULT].characterSet);
+    const std::string s = RangeText(targetStart, targetEnd);
+    CFStringRef cfsVal = CFStringCreateWithBytes(kCFAllocatorDefault,
+                                                 reinterpret_cast<const UInt8 *>(s.c_str()),
+                                                 s.length(), encoding, false);
+	  
+    const std::string tmputf = EncodedBytesString(cfsVal, kCFStringEncodingUTF8);
+    
+    if (text)
+      memcpy(text, tmputf.c_str(), tmputf.length());
+    CFRelease(cfsVal);
+    return tmputf.length();
+  }
+  return targetLength;
+}
+
+//--------------------------------------------------------------------------------------------------
+
+// Translates a UTF8 string into the document encoding.
+// Return the length of the result in bytes.
+int ScintillaCocoa::EncodedFromUTF8(char *utf8, char *encoded) const
+{
+  const int inputLength = (lengthForEncode >= 0) ? lengthForEncode : strlen(utf8);
+  if (IsUnicodeMode())
+  {
+    if (encoded)
+      memcpy(encoded, utf8, inputLength);
+    return inputLength;
+  }
+  else
+  {
+    // Need to convert
+    const CFStringEncoding encoding = EncodingFromCharacterSet(IsUnicodeMode(),
+                                                         vs.styles[STYLE_DEFAULT].characterSet);
+    
+    CFStringRef cfsVal = CFStringCreateWithBytes(kCFAllocatorDefault,
+                                                 reinterpret_cast<const UInt8 *>(utf8),
+                                                 inputLength, kCFStringEncodingUTF8, false);
+    const std::string sEncoded = EncodedBytesString(cfsVal, encoding);
+    if (encoded)
+      memcpy(encoded, sEncoded.c_str(), sEncoded.length());
+    CFRelease(cfsVal);
+    return sEncoded.length();
+  }
+}
+
+//--------------------------------------------------------------------------------------------------
+
 void ScintillaCocoa::SetMouseCapture(bool on)
 {
   capturedMouse = on;
diff --git a/doc/ScintillaDoc.html b/doc/ScintillaDoc.html
index 46039ed54..1749fa1f9 100644
--- a/doc/ScintillaDoc.html
+++ b/doc/ScintillaDoc.html
@@ -585,12 +585,14 @@ struct Sci_TextRange {
 };
 </pre>
 
-    <h3 id="EncodedAccess">GTK+-specific: Access to encoded text</h3>
+    <h3 id="EncodedAccess">Specific to GTK+ and Cocoa only: Access to encoded text</h3>
 
     <p><b id="SCI_TARGETASUTF8">SCI_TARGETASUTF8(&lt;unused&gt;, char *s)</b><br />
      This method retrieves the value of the target encoded as UTF-8 which is the default
      encoding of GTK+ so is useful for retrieving text for use in other parts of the user interface,
      such as find and replace dialogs. The length of the encoded text in bytes is returned.
+     Cocoa uses UTF-16 which is easily converted from UTF-8 so this method can be used to perform the
+     more complex work of transcoding from the various of encodings supported.
     </p>
 
     <p><b id="SCI_ENCODEDFROMUTF8">SCI_ENCODEDFROMUTF8(const char *utf8, char *encoded)</b><br />