Clean up string encoding handling

We should handle this in the low-level protocol code as much as possible
to avoid mistakes. This way the rest of the code can assume that strings
are always UTF-8 with \n line endings.

1 files changed, 19 insertions, 4 deletions

diff --git a/common/rfb/CMsgReader.cxx b/common/rfb/CMsgReader.cxx
index c0a96690..006645df 100644
--- a/common/rfb/CMsgReader.cxx
+++ b/common/rfb/CMsgReader.cxx
@@ -76,7 +76,12 @@ bool CMsgReader::readServerInit()
   std::vector<char> name(len + 1);
   is->readBytes((uint8_t*)name.data(), len);
   name[len] = '\0';
-  handler->serverInit(width, height, pf, name.data());
+
+  if (isValidUTF8(name.data()))
+    handler->serverInit(width, height, pf, name.data());
+  else
+    handler->serverInit(width, height, pf,
+                        latin1ToUTF8(name.data()).c_str());
 
   return true;
 }
@@ -275,9 +280,13 @@ bool CMsgReader::readServerCutText()
     vlog.error("cut text too long (%d bytes) - ignoring",len);
     return true;
   }
+
   std::vector<char> ca(len);
   is->readBytes((uint8_t*)ca.data(), len);
-  std::string filtered(convertLF(ca.data(), len));
+
+  std::string utf8(latin1ToUTF8(ca.data(), ca.size()));
+  std::string filtered(convertLF(utf8.data(), utf8.size()));
+
   handler->serverCutText(filtered.c_str());
 
   return true;
@@ -768,10 +777,16 @@ bool CMsgReader::readSetDesktopName(int x, int y, int w, int h)
 
   if (x || y || w || h) {
     vlog.error("Ignoring DesktopName rect with non-zero position/size");
-  } else {
-    handler->setName(name.data());
+    return true;
   }
 
+  if (!isValidUTF8(name.data())) {
+    vlog.error("Ignoring DesktopName rect with invalid UTF-8 sequence");
+    return true;
+  }
+
+  handler->setName(name.data());
+
   return true;
 }