Logged In: YES
user_id=3906

This isnt a bug. The code is encoded in 8 bits, no UTF-8,
no Unicode nor BIG-5.
For this reason I consider it an addition and not a correction.
Please explain what exactly is needed.
IMHO doing it is a huge drawback because I cant *see* if
the tables are properly constructed looking at escape sequences.
Are escape sequences needed for all? Which is the encoding
you are using?
Also look for a command line option to instruct the compiler
to process the files as ISO-8859-1 encoded.