Jeff, I wonder how an example code of a handful lines in the API doc can have such an error. However, thanks for the correction. I could find this in the IntSumReducer.java file and changed my code. I think it is safe to fall back to the old API but I will give the new one a try for few more days.
Regards, Farhan 2009/12/18 Jeff Zhang <[email protected]> > Farhan, > > In hadoop 0.20.x, using context.write(key,value) instead of > outputcollector.collect(key,value). > And It's the RecordWriter in Context do the write operation. > > Jeff Zhang > > > > -----Original Message----- > From: Farhan Husain [mailto:[email protected]] > Sent: 2009年12月17日 14:48 > To: [email protected] > Subject: context.collect > > The tutorial for the latest version uses MapReduceBase which is deprecated. > I tried to use the latest Mapper and Reducer classes as bases for my > classes. However, the javadoc page for those classes show something like > this: context.collect(word, one) whereas the Context class and none of its > descendants have any method called collect. What should be the correct code > here? > > Thanks, > Farhan > >
