Package org.apache.tinkerpop.gremlin.process.traversal.strategy.optimization

Class AdjacentToIncidentStrategy

java.lang.Object
org.apache.tinkerpop.gremlin.process.traversal.strategy.AbstractTraversalStrategy<TraversalStrategy.OptimizationStrategy>
org.apache.tinkerpop.gremlin.process.traversal.strategy.optimization.AdjacentToIncidentStrategy
All Implemented Interfaces:
Serializable, Comparable<Class<? extends TraversalStrategy>>, TraversalStrategy<TraversalStrategy.OptimizationStrategy>, TraversalStrategy.OptimizationStrategy
public final class AdjacentToIncidentStrategy extends AbstractTraversalStrategy<TraversalStrategy.OptimizationStrategy> implements TraversalStrategy.OptimizationStrategy
This strategy looks for vertex- and value-emitting steps followed by a CountGlobalStep and replaces the pattern with an edge- or property-emitting step followed by a CountGlobalStep. Furthermore, if a vertex- or value-emitting step is the last step in a .has(traversal), .and(traversal, ...) or .or(traversal, ...) child traversal, it is replaced by an appropriate edge- or property-emitting step. Performing this replacement removes situations where the more expensive trip to an adjacent graph element (e.g. the vertex on the other side of an edge) can be satisfied by trips to incident graph elements (e.g. just the edge itself).
Author:
Daniel Kuppitz (http://gremlin.guru)
See Also:
Example:
 __.out().count()            // is replaced by __.outE().count()
 __.in().limit(3).count()    // is replaced by __.inE().limit(3).count()
 __.values("name").count()   // is replaced by __.properties("name").count()
 __.where(__.out())          // is replaced by __.where(__.outE())
 __.where(__.values())       // is replaced by __.where(__.properties())
 __.and(__.in(), __.out())   // is replaced by __.and(__.inE(), __.outE())