DbfReader header size check fails on some dbf files

[pvlcursus]

DbfReader currently does a check on the header length read from bytes 8-9 in the dbf header. This length should exactly match the header descriptor size + the field descriptor size * number of fields + 1 (for the terminator mark). If the check fails DbfReader throws an exception "Invalid dBASE III file format."

However, I have a number of probably valid dbf files where the terminator mark 0Dh is followed by an extra byte 00h. This extra byte is part of the header and calculated in the header size reported in bytes 8-9. DbfReader throws an exception on these files, but I'm not convinced they violate the dBASE III file format. The same files are correctly processed by NetTopologySuite.IO.DbaseFileHeader and major ESRI geoprocessing software.