user defined RecordReader or Writer using pipes

[Youjie Zhou]

Hi, is there any specific steps on running wordcount-nopipe example?

or some way to control java RecordReader and RecordWriter when using hadoop 
pipes?

Thank you!