Also [pd_lua] and [pd_luax] are good for this sort of thing. Right click on the object to open the script in a text editor. [pd_luax] lets you edit the script without restarting Pd.

Martin

On 2012-08-20 04:57, Bryan Jurish wrote:

moin Fernando, moin list,

[... apologies for double-post; forgot to update my client address at work... shame on me! ]

sounds like [rattstok] might do what you want. it's part of [ratts], whose sources live at:

http://www.ling.uni-potsdam.de/~moocow/projects/pd/ratts-0.08.tar.gz

otherwise, you can write a simple regex tokenizer e.g. in python and wrap that in using [py]/[pyext]. a trickier option would be to use a "smarter" tokenizer for your particular target language(s); see e.g.

http://nltk.org/api/nltk.tokenize.html

... or you could roll your own tokenizer using finite-state machines and talk to it live and stream-wise in pd with [gfsm], but that's beyond the scope of this reply ;-)

marmosets, Bryan

On 2012-08-19 00:06, FernandoG wrote:

...

Hi PD users :)

I am working with text to sound transformation using binfile objet and have a question:

binfile objetc read a file and then output caracter by caracter every bang recived, i am planning to use not a stream of ascii number rather than a list of them. Then the idea is to use words as control signal for sound sinthesis. How can i get variable lists from a stream? wich objetc can help my? its needed to split the stream in to words and recognize puctuation or space as words limits.

Thanks! F