Fix wxStringOutputStream position when using external string

Initialize m_pos correctly when using an existing, and hence possibly
not empty, string (and not the internal one which is always empty
initially). The old code was totally wrong as it divided the string
length by the size of wxChar instead of multiplying by it, but doing
this could have been wrong too with UTF-16 and surrogates, so use the
conversion object to compute the real length of the string
representation in the corresponding encoding.

Add a simple unit test checking that this works as intended.

Closes #17985.

tests/streams/sstream.cpp

@@ -119,3 +119,32 @@ void strStream::Output_Check()
 
 // Register the stream sub suite, by using some stream helper macro.
 STREAM_TEST_SUBSUITE_NAMED_REGISTRATION(strStream)
+
+TEST_CASE("wxStringOutputStream::Tell", "[stream]")
+{
+    wxStringOutputStream ss;
+    CHECK( ss.TellO() == 0 );
+
+    const char* const s = "Hello world";
+    const wxFileOffset len = strlen(s);
+
+    ss.Write(s, len);
+    CHECK( ss.TellO() == len );
+
+    wxString str(s);
+    CHECK( wxStringOutputStream(&str).TellO() == len );
+
+
+    wxMBConvUTF16 convUTF16;
+    wxStringOutputStream ss16(NULL, convUTF16);
+    CHECK( ss16.TellO() == 0 );
+
+    const wxCharBuffer s16 = convUTF16.cWC2MB(wxWCharBuffer(str.wc_str()));
+    ss16.Write(s16, s16.length());
+    CHECK( ss16.TellO() == 2*len );
+    CHECK( wxStringOutputStream(&str, convUTF16).TellO() == 2*len );
+
+    // The U+2070D character is represented by a surrogate pair in UTF-16.
+    wxString u2070D = wxString::FromUTF8("\xF0\xA0\x9C\x8D");
+    CHECK( wxStringOutputStream(&u2070D, convUTF16).TellO() == 4 );
+}