This approach is quite general and now I was wondering which functionality I could insert into this library. Also how I can deal with this general flow of data to make a usable library for all kinds of data.. The data from Yahoo Pipes can be news items (title, description, postdate, images, links,..) Yahoo Pipes also has functionality for parsing CSV files, so "data tables" are also possible.
Also I was thinking about the structure of the library, what the best way is to implement it. In general the library needs to load in the data from the pipe (JSON string) and pass it to a visualization library like Google Chart or D3.js (formerly Protovis)
Thanks for the replies, any advice is welcome!
2) Download and use Douglas Crockford's JSON library