PandaLex PDF Parser

A siginificantly improved lexer, which successfully parses many more of the regression tests.

A bug in the header event which stopped the extra
header info from being passed through was fixed.
The name of the sample application, pdfinfo, was
changed to pdfdump since it conflicts with the
pdfinfo utility from the xpdf distribution. More
callbacks were implemented.

A massively-improved PDF parsing engine.

Initial freshmeat announcement.