How the runner is used.

[rayk]

Walker Signature

(Graph, Processor) => Future<Graph>

Actors

1. Consumer: Code that async calls the walker and received but <Graph>.
2. Graph: Mutatable data structure containing Typed Vertex and Edges.
3. Processor: Self-contained program that produces a subgraph from the graph that was passed int.
4. Walker: Handles the effective traversal of <Graph> yielding nodes to <Processor>.

Dependency Structure & Importing.

1. Walker depends on Graph & Processor.
2. Consumer depends on Walker, Graph & Processor.
3. Processor depends on Graph

Whilst Graph is read and mutated outside Walker, Processor is only executed by Walker. Yet Consumer may create instances of Processor with particular parameters, or get value from the processor to determine what it does or any state that may be contained within.

Use Case Examples

• let g represent the source graph.
• let r represent the resulting graph.
• let p represent the processor.

( g , p ) => r

A. Graph Visualization

The consumer wants to produce a visualization of the graph that includes every vertex and edge in the graph. The actual visualization is handled by another package like gviz, which requires building a series of nodes with some data point from each of the vertex and edges in g.

(r)

• This a <Graph> produced by p as a result of consuming all the vertex and edges in g.
• Each each vertex and edge in g maps to a vertex or edge in r .
• Each vertex and edge in r contains properties that are intended to be passed writer that calls methods in gviz to produce the visualization.

B. Subgraph Extraction

The consumer wants to operate on a subset of vertex and edges which are contained in g, criteria of which is defined in p. As p receives each vertex and edge from g, it decides if the node is of interest and places it in r.

(p)

• Assumes that it is passed each unique node (Vertex & Edge) only once.
• Assumes that nodes are passed in topological order.

C. Subgraph Visualization

The consumer wants to visualise only a subset of the graph.

1. Consumer receives r by calls the walker with p that extracts a subgraph from g.
2. Consumer calls walker a second time, with the previous r as the g parameter and the visualiser processor from above as the p parameter.

D. Calculation From Node Values.

The consumer whats to a calculation using the values contained in a path of vertex and edges.

1. Consumer calls the walker with p that forms a subgraph of the required nodes.
2. Consumer calls the walker a second time with the above-produced subgraph r with another p that knows how to perform the calculation and produces r.
3. Consumer uses r to decide what to do next.