pyshapelib unicode saga

[Bram de Greve]

I've made an attempt to add (partial) Unicode support to dbflib.  Partial,
because strings still need to be encoded to a subset.  But this is of course
how the DBF databases are designed.   The language driver ID (LDID) in the
DBF header is used to specify this encoding.  Unfortunately there's not
UTF-8 LDID ...