} else {
// TODO Discover if these older documents can ever hold Unicode Strings?
// (We think not, because they seem to lack a Piece table)
// TODO Build the Piece Descriptor properly
// (We have to fake it, as they don't seem to have a proper Piece table)
PieceDescriptor pd = new PieceDescriptor(new byte[] {0,0, 0,0,0,127, 0,0}, 0);
pd.setFilePosition(_fib.getFibBase().getFcMin());
// Generate a single Text Piece Table, with a single Text Piece
// which covers all the (8 bit only) text in the file
tpt = new TextPieceTable();
byte[] textData = new byte[_fib.getFibBase().getFcMac()-_fib.getFibBase().getFcMin()];