I'm particular interested in large PDF parsing, for those table form PDF content. Biggest problem of PDF parsing is the relativity of content according to content that output not the PDF internal layout box sequence. Sometime parse result had content sequence off.