On Mon, Jun 11, 2012 at 10:29 PM, Andrew Faraday <jbturgid@hotmail.com> wrote:

I've got an open source project using ruby to parse strings and send commands via TCP to pure data. Which started with some of my earliest non-pd coding. It's not currently set up to read text files, but it'd be a fairly simple mod, so you're welcome to learn ruby and submit a patch.

PDF's are a much more complicated file format, I don't know how you'd go about extracting the text content from them to feed the text-to-music algorithm.
[...]

PDF is indeed complicated, but extracting text can be as simple as a (simple) regular expression. As far as I understand, basically, everything between parentheses "(" and ")" is text (or more rigorously, everything between parentheses between the strings "BT" and "ET" between the strings "obj" and "endobj" is text, but I think it's enough to search for only the parantheses). The escape character is the backslash, and it has only a couple uses:

Sequence | Meaning
---------------------------------------------
\n       | LINE FEED (0Ah) (LF)
\r       | CARRIAGE RETURN (0Dh) (CR) 
\t       | HORIZONTAL TAB (09h) (HT)
\b       | BACKSPACE (08h) (BS)
\f       | FORM FEED (FF)
\(       | LEFT PARENTHESIS (28h)
\)       | RIGHT PARENTHESIS (29h)
\\       | REVERSE SOLIDUS (5Ch) (Backslash)
\ddd     | Character code ddd (octal)

Apart from security settings that may block text extraction, unfortunately, there could be compression applied - but I don't know how that works.

András