User Defined Steps

Gremlin provides the ability for a user to define their own step definitions natively in Groovy or in Java. This is very useful when wishing to work with your low-level graph data at a higher level of abstraction. This section will discuss how to write your own step definitions and demonstrate how they are useful for making your Gremlin code more concise and more self-explanatory.

1. Defining a Step in Gremlin
2. Defining an Anonymous Step in Gremlin

Defining a Step in Gremlin

Gremlin comes with a collection of built-in step definitions (see Gremlin Steps). It is possible for developers to create their own step definitions. Simply add a closure that represents the step to the respective classes. The method to use is:

Gremlin.defineStep(String stepName, List<Class> classes, Closure stepClosure);

Gremlin.defineStep('codeveloper',[Vertex,Pipe], {_().as('x').out('created').in('created').filter{v,m -> m.x != v}})

In the code above, the final argument is a closure to create the desired composite step. The step closure says:

• Reference the current vertex by the variable x
• Get the outgoing created vertices of the current vertex.
• Get the incoming created vertices of those previous vertices.
• Exclude the first vertex from the path (a person can not be a codeveloper of themselves)
  • The second argument of the filter is the variable map of the pipeline (references as steps)

Given the graph diagrammed in Defining a Property Graph, we can determine the codevelopers of a particular vertex.

gremlin> g = TinkerGraphFactory.createTinkerGraph()
==>tinkergraph[vertices:6 edges:6]
gremlin> g.v(1).codeveloper
==>v[4]
==>v[6]

Realize that this step definition can be used like any other step definition.

gremlin> g.v(1).codeveloper.name
==>josh
==>peter
gremlin> g.v(1).codeveloper.name.toString() 
==>[StartPipe, [AsPipe(x,StartPipe), OutPipe(created), InPipe(created), FilterFunctionPipe], PropertyPipe(name)]

What step definitions allow you to do is to work with “higher order” relationships in your graph. Thus, instead of working at the level of

_().as('x').out('created').in('created').filter{v,m -> m.x != v}

you can work at the more semantically natural level of

codeveloper

It is also possible to pass in parameters to a step definition. For example, lets generalize codeveloper to simply be co.

Gremlin.defineStep('co',[Vertex,Pipe], {String label -> _().as('x').out(label).in(label).filter{v,m -> m.x != v}})

In the above step definition, upon evaluating the step, the step expects a String parameter to be passed into it. Thus, in order to properly evaluate the co step do the following:

gremlin> g.v(1).co('created')
==>v[4]
==>v[6]

Finally, realize that the parameters passed to the step can be closures. Thus, it is possible to effect dynamic behavior.

Gremlin.defineStep('twoStep', [Pipe, Vertex], { final Object... params ->  _().as('x').out(params[0]).in(params[0]).filter(params[1]) });

gremlin> g.v(1).twoStep('created'){v,m -> v != m.x}
==>v[4]
==>v[6]
gremlin> g.v(1).twoStep('created'){v,m -> v == m.x}
==>v[1]

Defining an Anonymous Step in Gremlin

You can also create an anonymous step (or “lambda step”) using the step step. For example:

gremlin> g = TinkerGraphFactory.createTinkerGraph()    
==>tinkergraph[vertices:6 edges:6]
gremlin> g.v(1).out.step{ it.next().map() }
==>{name=vadas, age=27}
==>{name=lop, lang=java}
==>{name=josh, age=32}

The method it.next() is equivalent to starts.next() which is defined in Pipes and represents the next end of the previous pipe. For those familiar with Pipes, the provided step function becomes the method AbstractPipe.processNextStart().