camjo

  // Following model based on the description of 
  // Apache Beam in the book Streaming Systems by Reuven Lax; Tyler Akidau; Slava Chernyak

  /**
    * Tables are data at rest, and act as a container for data to accumulate and be observed over time.
    * K = Key
    * V = Value
    * W = Window
    * P = Partition
    */

camjo

// Note that this is written in terms of functions for now. We will rewrite this with arrows etc 
// and work out the reified encoding once we agree on the API

/**
    * At a high level, there are two core concepts here
    * 1. Data
    * 2. Computation
    *
    * The Data can be thought of as Bounded or Unbounded.
    * This generally corresponds to a "batch" source or a "streaming" source.

camjo

import scalaz.effect.IO

trait Source[F[_], A] {
  def iterator(s: F[A]): Iterator[A]
}

object Source {
  implicit def listSource[A]: Source[List, A] = (s: List[A]) => s.toIterator
  implicit def streamSource[A]: Source[Stream, A] = (s: Stream[A]) => s.toIterator
}

camjo

import scalaz.{-\/, Applicative, Coproduct, Free, FreeAp, Inject, Monad, NaturalTransformation, Nondeterminism, \/-, ~>}
import scala.language.{higherKinds, reflectiveCalls}
import scalaz.concurrent.Task
import Task._
import scala.util.Random

case class User(name: String, age: Int)

sealed trait UserOperation[T]
case class CreateUser(name: String, age: Int) extends UserOperation[User]