Nice! It'd be kind of cool if they where in a format easily machine parsable, so you could easily identify content from each character. I guess it wouldn't be

Nice! It'd be kind of cool if they where in a format easily machine parsable, so you could easily identify content from each character. I guess it wouldn't be impossible though to take the pdfs and use information based on the text location on the page etc. to work out who the speech belongs to.