skeleton
Class Compute

java.lang.Object
  skeleton.Compute

All Implemented Interfaces:

java.io.Serializable, java.lang.Cloneable

Direct Known Subclasses:

Farm, ParCompute, Pipeline

public abstract class Compute
extends java.lang.Object
implements java.lang.Cloneable, java.io.Serializable

This is the interface that has to be implemented by the sequential portions of code used in the program computation. Each sequential skeleton is an object that implements this interface. We explicitly choosed not to use 1.5 features, at the moment. Therefore the generic function takes Objects and computes Objects.
The typical usage is the following:

• suppose you have to compute in parallel f(x1) ... f(xn) out of x1 ... xn
  
• suppose f takes as input an InputTask task and produces and OutputResult result
• then you'll use a code such as:

   public class myF extends Compute {
      public Object compute(Object task) {
        OutputResult res = f((InputTask) task);
        return(res);
      }
   }
   

  every time you need to use the f function to compute a pipeline stage, a farm worker, etc.

See Also:

Serialized Form

Constructor Summary

Compute()

Method Summary

java.lang.Object	clone() just to be able to clone in case of usage of ObjectStreams
abstract java.lang.Object	compute(java.lang.Object task) the method computes an Object result out of an Object input data set

Methods inherited from class java.lang.Object

equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

Compute

public Compute()

Method Detail

compute

public abstract java.lang.Object compute(java.lang.Object task)

the method computes an Object result out of an Object input data set

Returns:

the result computed

clone

public java.lang.Object clone()

just to be able to clone in case of usage of ObjectStreams

Overrides:

clone in class java.lang.Object

Returns:

a freh copy of the object