Using map reduce in a declarative paradigm to do tf-idf

I found this very nice post on how to do a TF-IDF very quicly using a declarative language called PIG. The main benifit is that PIG integrates well with Hadoop and Map-reduce. For any one who knows how mechanically long it is to writing a Java program for map reduce on hadoop would like this.

http://horicky.blogspot.com/2009/01/solving-tf-idf-using-map-reduce.html