[trez]

you can also use pdf2html with the option -x (to get xml). You would also have the position of each text tokens.