Scqla

A fully asynchronous Cassandra CQL driver for Scala using Akka IO.

##Highlights:

1. Directly construct objects from results.

2. Completely stateless, no sessions required. Fire any query from anywhere, anytime.

3. Load balance between cassandra nodes.

4. Register callbacks for events.

###Connect to cassandra cluster.

Scqla.connect

###Import from Scqla object

import Scqla._

When you execute a query you get a future which can have a result or an error ErrorException(error: Error) where Error is

case class Error(code: Int, e: String)

###Create a new keyspace

query("CREATE KEYSPACE demodb WITH REPLICATION = {'class' : 'SimpleStrategy','replication_factor': 1}") onComplete {
  case Success => 
  case Failure =>
}

###Set global keyspace

Note: This only works if you only connect to one node in Cassandra cluster. If that is not the case use full table qualifiers as shown in the examples here.

query("use demodb") onComplete (
  case Failure(f) => println(f.error)
  case Success => 
)

###Create a new table

    query("""CREATE TABLE demodb.emp (
    		empID int,
    		deptID int,
    		alive boolean,
    		id uuid,
    		first_name varchar,
    		last_name varchar,
    		salary double,
    		age bigint,
        PRIMARY KEY (empID, deptID))""") onComplete (
      case Failure(f) => println(f.error)
      case Success =>
    )

###Execute prepared queries

    prepare("INSERT INTO demodb.emp (empID, deptID, alive, id, first_name, last_name, salary, age) VALUES (?, ?, ?, ?, ?, ?, ?, ?)") onComplete (
      case Failure(f) => println(f.error)
      case Success(p) => p.execute(104, 15, true, new java.util.UUID(0, 0), "Hot", "Shot", 10000000.0, 98763L)
    )

###You can directly construct objects from the result.

###Case class list

case class Emp(empId: Int, deptId: Int, alive: Boolean, id: java.util.UUID, first: String, last: String, salary: Double, age: Long)

queryAs[Emp]("select empID, deptID, alive, id, first_name, last_name, salary, age from demodb.emp") onComplete (
  case Failure(f) => println(f.error)
  case Success(empList) => 
    empList.foreach(e => s"First name of employee id ${e.empId} is ${e.first}")
)

###Primitives list

queryAs[Int]("select empid from demodb.emp") onComplete (
  case Failure(f) => println(f.error)
  case Success(idList) => {
    idList.foreach(println)
  }
)

###Strings

queryAs[String]("select first_name from demodb.emp") onComplete (
  case Failure(f) => println(f.error)
  case Success(nameList) => {
    nameList.foreach(println)
  }
)

###Execute prepared queries and get results

val f = prepare("select empID, deptID, alive, id, first_name, last_name, salary, age from demodb.emp where empid = ? and deptid = ?")

val list = for {
  prepared <- f
  empList <- prepared.executeGet[Emp](104, 15)
} yield empList

list foreach {
    e => println(s"First name of employee id ${e.empId} is ${e.first}")
}

###Register a callback for an event

Events.registerDBEvent(CreatedEvent, (a, b) => {
  println(s"Hear hear, $a $b have come to be.")
})

Events:

abstract class Event

abstract class NodeEvent extends Event
case object NewNodeEvent extends NodeEvent
case object RemovedNodeEvent extends NodeEvent
case object NodeUpEvent extends NodeEvent
case object NodeDownEvent extends NodeEvent

abstract class DBEvent extends Event
case object CreatedEvent extends DBEvent
case object UpdatedEvent extends DBEvent
case object DroppedEvent extends DBEvent

NodeEvent handlers are of the type Function[InetAddress, Unit] DBEvent handlers are of the type Function2[String, String, Unit]

###Drop keyspace

query("drop KEYSPACE demodb") onComplete (
  case Failure(f) => println(f.error)
  case Success => {
    //do something
  }
)

###Configuring thread pool for scqla future execution By default Scqla makes use of the default execution context in scala. However you can specify a manual ExecutionContext config by adding something like this to your application.conf

contexts {
  scqla-pool {
    fork-join-executor {
      parallelism-factor = 1.0
      parallelism-max = 4
    }
  }
}