All Downloads are FREE. Search and download functionalities are using the official Maven repository.

jadex.rules.rulesystem.rete.extractors.TupleExtractor Maven / Gradle / Ivy

Go to download

Jadex Rules is a small lightweight rule engine, which currently employs the well-known Rete algorithm for highly efficient rule matching. Jadex rules is therefore similar to other rule engines like JESS and Drools. Despite the similarities there are also important differences between these systems: * Jadex Rules is very small and intended to be used as component of other software. Even though rules can be specified in a Java dialect as well as (a small variation of) the CLIPS language its primary usage is on the API level. Jadex Rules is currently the core component of the Jadex BDI reasoning engine. * Jadex Rules cleanly separates between state and rule representation. This allows the state implementation as well as the matcher to be flexibly exchanged. Some experiments have e.g. been conducted with a Jena representation. Regarding the matcher, it is planned to support also the Treat algorithm, which has a lower memory footprint than Rete. * Jadex Rules pays close attention to rule debugging. The state as well as the rete engine can be observed at runtime. The rule debugger provides functionalities to execute a rule program stepwise and also use rule breakpoints to stop the execution at those points.

There is a newer version: 2.4
Show newest version
package jadex.rules.rulesystem.rete.extractors;

import jadex.commons.SUtil;
import jadex.rules.rulesystem.rete.Tuple;
import jadex.rules.rulesystem.rete.nodes.VirtualFact;
import jadex.rules.state.IOAVState;
import jadex.rules.state.OAVAttributeType;

/**
 *  Extractor for fetching a value from a rete tuple.
 */
public class TupleExtractor implements IValueExtractor
{
	//-------- attributes --------
	
	/** The tuple index. */
	protected int tupleindex;
	
	/** The attribute type. */
	protected OAVAttributeType attr;
	
	/** The key value. */
	protected Object key;
	
	//-------- constructors --------
	
	/**
	 *  Create a new extractor.
	 */
	public TupleExtractor(int tupleindex, OAVAttributeType attr)
	{
		this(tupleindex, attr, null);
	}
	
	/**
	 *  Create a new extractor.
	 */
	public TupleExtractor(int tupleindex, OAVAttributeType attr, Object key)
	{
		this.tupleindex = tupleindex;
		this.attr = attr;
		this.key = key;
	}

	//-------- methods --------
	
	/**
	 *  Get the value of an attribute from an object or tuple.
	 *  @param left The left input tuple. 
	 *  @param right The right input object.
	 *  @param prefix The prefix input object (last value from previous extractor in a chain).
	 *  @param state The working memory.
	 */
	public Object getValue(Tuple left, Object right, Object prefix, IOAVState state)
	{
		// Fetch the object from the tuple
		
		// Fetch the value from the state
		// a) attr == null -> use object
		// b) attr !=null -> use state.getAttributeValue(object, attr) or
		//                       state.getAttributeValues(object, attr);
		
		Object object = left.getObject(tupleindex);
		
		if(object instanceof VirtualFact)
			object = ((VirtualFact)object).getObject();
		
		Object ret;
		if(attr==OAVAttributeType.OBJECTTYPE)
		{
			ret = state.getType(object);
		}
		else
		{
			ret = attr==null?  object: 
				OAVAttributeType.NONE.equals(attr.getMultiplicity())?
					state.getAttributeValue(object, attr):
					key!=null? state.getAttributeValue(object, attr, key):
					state.getAttributeValues(object, attr);
		}
		
		return ret;
	}
	
	/**
	 *  Test if a constraint evaluator is affected from a 
	 *  change of a certain attribute.
	 *  @param tupleindex The tuple index.
	 *  @param attr The attribute.
	 *  @return True, if affected.
	 */
	public boolean isAffected(int tupleindex, OAVAttributeType attr)
	{
		return (this.tupleindex==tupleindex || tupleindex==-1) && SUtil.equals(this.attr, attr);
	}
	
	/**
	 *  Get the attribute.
	 *  @return The attribute.
	 */
	public OAVAttributeType getAttribute()
	{
		return attr;
	}
	
	/**
	 *  Get the set of relevant attribute types.
	 */
	public AttributeSet	getRelevantAttributes()
	{
		AttributeSet ret = new AttributeSet();
		if(attr!=null)
			ret.addAttribute(attr);
		return ret;
	}
	
	/**
	 *  Get the set of indirect attribute types.
	 *  I.e. attributes of objects, which are not part of an object conditions
	 *  (e.g. for chained extractors) 
	 *  @return The relevant attribute types.
	 */
	public AttributeSet	getIndirectAttributes()
	{
		return AttributeSet.EMPTY_ATTRIBUTESET;
	}

	/**
	 *  Get the tuple index.
	 *  @return The tuple index.
	 */
	public int getTupleIndex()
	{
		return tupleindex;
	}
	
	/**
	 *  Get the string representation.
	 *  @return The string representation. 
	 */
	public String toString()
	{
		return "["+tupleindex+"]"+"."+(attr==null? "object": attr.getName())+(key==null? "": "["+key+"]");
	}

	/**
	 *  The hash code.
	 */
	public int hashCode()
	{
		int	result	= 31 + (attr!=null ? attr.hashCode() : 0);
		result	= result*31 + tupleindex;
		return result;
	}

	/**
	 *  Test for equality.
	 */
	public boolean equals(Object obj)
	{
		if(this==obj)
			return true;

		boolean	ret	= false;
		if(obj instanceof TupleExtractor)
		{
			TupleExtractor	other	= (TupleExtractor)obj;
			ret	= SUtil.equals(attr, other.getAttribute())
				&& tupleindex==other.getTupleIndex();
		}
		return ret;
	}
}




© 2015 - 2024 Weber Informatics LLC | Privacy Policy