Fix wrong resulting string length in UTF-16 to wchar_t conversion.

Don't optimize the returned length for surrogate-less case, this does save a
pass of the string but at the price of returning a wrong result, which is not
worth it, just compute the really required length exactly.

Closes #16298.

git-svn-id: https://svn.wxwidgets.org/svn/wx/wxWidgets/branches/WX_3_0_BRANCH@76618 c3d73ce0-8a6f-49c7-b76d-6d57e0e08775

docs/changes.txt

@@ -580,6 +580,7 @@ Major new features in this release
 All:
 
 - Fix wxMBConvUTF8::ToWChar() for non-NUL-terminated strings (andyr).
+- Fix length returned from wxMBConvUTF16::ToWChar(NULL) (Thomas Goyne).
 - Fix wxSocket::WaitForAccept() in non-main thread (Hajo Kirchhoff).
 - Fix memory overallocation in wxVector::reserve() (Nigel Paton).
 - Fix `wx-config --libs` in monolithic build.

src/common/strconv.cpp

@@ -1638,14 +1638,6 @@ wxMBConvUTF16straight::ToWChar(wchar_t *dst, size_t dstLen,
         return wxCONV_FAILED;
 
     const size_t inLen = srcLen / BYTES_PER_CHAR;
-    if ( !dst )
-    {
-        // optimization: return maximal space which could be needed for this
-        // string even if the real size could be smaller if the buffer contains
-        // any surrogates
-        return inLen;
-    }
-
     size_t outLen = 0;
     const wxUint16 *inBuff = reinterpret_cast<const wxUint16 *>(src);
     for ( const wxUint16 * const inEnd = inBuff + inLen; inBuff < inEnd; )
@@ -1654,10 +1646,15 @@ wxMBConvUTF16straight::ToWChar(wchar_t *dst, size_t dstLen,
         if ( !inBuff )
             return wxCONV_FAILED;
 
-        if ( ++outLen > dstLen )
-            return wxCONV_FAILED;
+        outLen++;
 
-        *dst++ = ch;
+        if ( dst )
+        {
+            if ( outLen > dstLen )
+                return wxCONV_FAILED;
+
+            *dst++ = ch;
+        }
     }
 
 
@@ -1711,14 +1708,6 @@ wxMBConvUTF16swap::ToWChar(wchar_t *dst, size_t dstLen,
         return wxCONV_FAILED;
 
     const size_t inLen = srcLen / BYTES_PER_CHAR;
-    if ( !dst )
-    {
-        // optimization: return maximal space which could be needed for this
-        // string even if the real size could be smaller if the buffer contains
-        // any surrogates
-        return inLen;
-    }
-
     size_t outLen = 0;
     const wxUint16 *inBuff = reinterpret_cast<const wxUint16 *>(src);
     for ( const wxUint16 * const inEnd = inBuff + inLen; inBuff < inEnd; )
@@ -1737,10 +1726,15 @@ wxMBConvUTF16swap::ToWChar(wchar_t *dst, size_t dstLen,
         if ( numChars == 2 )
             inBuff++;
 
-        if ( ++outLen > dstLen )
-            return wxCONV_FAILED;
+        outLen++;
 
-        *dst++ = ch;
+        if ( dst )
+        {
+            if ( outLen > dstLen )
+                return wxCONV_FAILED;
+
+            *dst++ = ch;
+        }
     }
 
 

tests/strings/unicode.cpp

@@ -389,6 +389,11 @@ void UnicodeTestCase::ConversionUTF16()
     size_t len;
     conv.cMB2WC("\x01\0\0B\0C" /* A macron BC */, 6, &len);
     CPPUNIT_ASSERT_EQUAL( 3, len );
+
+    // Another one: verify that the length of the resulting string is computed
+    // correctly when there is a surrogate in the input.
+    wxMBConvUTF16BE().cMB2WC("\xd8\x03\xdc\x01" /* OLD TURKIC LETTER YENISEI A */, wxNO_LEN, &len);
+    CPPUNIT_ASSERT_EQUAL( 1, len );
 }
 
 void UnicodeTestCase::ConversionUTF32()