Possible Duplicate:
How to convert PDF to (La)TeX?
Has anyone written a piece of code that (at least) tries to reconstruct (standard) LaTeX from a compiled PDF file? I mean, identifying indices and exponents, common fonts, footnotes, linebreaks should mostly be algorithmizable, if somewhat error-prone. Basically I am looking for an approximative section of the map pdflatex: tex → pdf which does most of the work, leaving maybe a few minor corrections (but not the whole design) to be done at hand.
No, I am not asking for an auto-plagiarizer tool. I am sure I am not the only one here with data loss issues...