Added a test case that runs almost all the tsheg bars from all

non-reference, publicly available ACIP files (hundreds of megabytes of
them) through the converter.  The frequencies of these tsheg bars in
in the file, too.

source/org/thdl/tib/input/TibetanConverter.java

@@ -245,6 +245,9 @@ public class TibetanConverter implements FontConverterConstants {
         } catch (IOException e) {
             e.printStackTrace(out);
             return 4;
+        } catch (OutOfMemoryError e) {
+            e.printStackTrace(out);
+            throw e;
         }
 	}