DAG is the acronym for Directed Acyclic Graph which is basically a graph without closed loops made of 'single way' directed connections. It has a lot of pratical applications, usually related to shortest-path or topological ordering.

While the basic algorithms and the same SwiftDAG is modelled to provide a simple interface to represent data structure relations in memory, where each Node (or vertex) is an object and each Edge (or connection) is a relation between nodes.

Example:

class UFrame: Node {
    var size: CGRect?
    var origin: CGPoint?
}

class UView: Node {
    lazy var subviews = EdgeArray<UView>(parent: self)
    lazy var frame = Edge<UFrame>(parent: self)
    var backgroundColor = UIColor.black
}

UView is connected to UFrame via the Edge variable frame and to multiple subviews via the EdgeArray variable subviews.

The connections are created ar run-time.

Example, explicit:

    let view = UView()
    let frame = UFrame()
    view.frame.connect(to: frame)

Example, operator:

    let view = UView()
    let frame = UFrame()
    view.frame <= frame

Pratically speaking the base Node class contains the list of connetions and the Edge classes helps to link/unlink the Nodes

By storing the conections from a Node to its siblings it's possible to traverse the tree to create a Topological Order

    print(view.topologicalOrder())

    [UFrame(uuid: 140418024708768), UView(uuid: 140418022602544)]

1. try/catch: exceptions are used to avoid closed loops and integrity checks
2. behind the scens, inner links and outer links
3. direct graph manipulation
4. simple queries
5. serialization/deserialization
6. PRO users: sourcery based API

1. Integration with CocoaPods / Carthage
2. JSON serialization/deserialization using the hash
3. more examples and more tests
4. travis integration

./Sourcery --templates SwiftDAG/Templates --sources . --output SwiftDAGTests --watch --verbose