shiftforward / apso

ShiftForward's utilities library

Version Matrix

Apso Build Status Maven Central

Apso is ShiftForward's utilities library. It provides a series of useful methods.


Apso's latest release is built against Scala 2.11 and Scala 2.12.

To use it in an existing SBT project, add the following dependency to your build.sbt:

libraryDependencies += "eu.shiftforward" %% "apso" % "0.12.10"

The TestKit is available under the apso-testkit project. You can include it only for the test configuration:

libraryDependencies += "eu.shiftforward" %% "apso-testkit" % "0.12.10" % "test"

Please take into account that the library is still in an experimental stage and the interfaces might change for subsequent releases.

Table of Contents


The Benchmark object provides an apply method to measure the running time of a block of code.

scala> import eu.shiftforward.apso.Benchmark
import eu.shiftforward.apso.Benchmark

scala> Benchmark("test") { (0l to 100000000).sum }
# Block "test" completed, time taken: 1 ms (0.001 s)
res0: Int = 5000000050000000

It's also possible to customize the method (String => Unit) that is used to print the results of the benchmark (by default it's println). This can be particularly useful if you want to use a logging framework, for example:

scala> def info(s: String) = println(s"[INFO] $s")
info: (s: String)Unit

scala> Benchmark("test", info) { (0l to 100000000).sum }
[INFO] # Block "test" completed, time taken: 0 ms (0.0 s)
res1: Long = 5000000050000000


The CounterPair object provides a method to pack two numbers in the range of an unsigned short in an Int.

scala> import eu.shiftforward.apso.CounterPair
import eu.shiftforward.apso.CounterPair

scala> CounterPair(1, 2)
res0: Int = 131073

scala> CounterPair(a, b) = res0
a: Int = 1
b: Int = 2


The Geo object provides methods to compute distances in kilometers between two points on the planet Earth, calculated using the spherical law of cosines. Coordinates are represented by a pair of Double for latitude and longitude.

scala> import eu.shiftforward.apso.Geo
import eu.shiftforward.apso.Geo

scala> Geo.distance((41.1617609, -8.6024716), (41.1763745, -8.5964861))
res0: Double = 1.7004440762344684

You can also have the distance function curried if you are computing distances from a fixed point:

scala> val distFromOffice = Geo.distanceFrom((41.1617609, -8.6024716))
distToOffice: eu.shiftforward.apso.Geo.Coordinates => Double = <function1>

scala> distFromOffice((41.1763745, -8.5964861))
res1: Double = 1.7004440762344684

scala> distFromOffice((38.7223032, -9.1414664))
res2: Double = 275.118392477037


The JreVersionHelper object provides helper methods to check the two most significant parts of the JRE version at runtime:

scala> import eu.shiftforward.apso.JreVersionHelper
import eu.shiftforward.apso.JreVersionHelper

scala> JreVersionHelper.jreVersion
res0: (Int, Int) = (1,8)


The Logging and StrictLogging traits allows mixing in Log4j2 Logger objects. The difference between the two is that in the former the Logger object is initialized lazily, while in the latter it is initialized strictly:

scala> class A extends Logging {}
defined class A

scala> val a = new A
a: A = A@58af6f21



The OrderingHelper object provides the min and max methods for comparing two instances of the same type:

scala> import eu.shiftforward.apso.OrderingHelper._
import eu.shiftforward.apso.OrderingHelper._

scala> min(2, 3)
res0: Int = 2

scala> max(2, 3)
res1: Int = 3


The ProgressBar represents a widget to print a dynamic progress bar in a console.

scala> import eu.shiftforward.apso.ProgressBar
import eu.shiftforward.apso.ProgressBar

scala> val progress = ProgressBar(100)
progress: eu.shiftforward.apso.ProgressBar = ProgressBar(100)

scala> progress.tick(1)
  1% [>                                                     ] / [ 0.19 ] ops/s

scala> progress.tick(2)
  3% [=>                                                    ] - [ 0.15 ] ops/s

scala> progress.tick(1)
  4% [==>                                                   ] \ [ 0.12 ] ops/s

scala> progress.tick(10)
 14% [=======>                                              ] | [ 0.31 ] ops/s

scala> progress.tick(20)
 34% [==================>                                   ] / [ 0.46 ] ops/s

scala> progress.tick(30)
 64% [=================================>                    ] - [ 0.77 ] ops/s


The Reflect object contains helpers for reflection-related tasks, namely to create an instance of a given class given its fully qualified name and also to access singleton objects:

scala> import eu.shiftforward.apso.Reflect
import eu.shiftforward.apso.Reflect

scala> import eu.shiftforward.apso.collection._
import eu.shiftforward.apso.collection._

scala> Reflect.newInstance[HMap[Nothing]]("eu.shiftforward.apso.collection.HMap")
res0: eu.shiftforward.apso.collection.HMap[Nothing] = HMap()

scala> Reflect.companion[Reflect.type]("eu.shiftforward.apso.Reflect")
res1: eu.shiftforward.apso.Reflect.type = eu.shiftforward.apso.Reflect$@3b1dbca


The Retry object provides a method to retry methods or Futures a given number of times until they succeed or the specified maximum number of retries is reached:

scala> import scala.concurrent.Future
import scala.concurrent.Future
scala> import eu.shiftforward.apso.Retry
import eu.shiftforward.apso.Retry

scala> import

scala> import java.util.concurrent.atomic.AtomicInteger
import java.util.concurrent.atomic.AtomicInteger

scala> val a = new AtomicInteger()
a: java.util.concurrent.atomic.AtomicInteger = 0

scala> def f: Future[Int] = {
     |   Future {
     |     val value = a.getAndAdd(1)
     |     if (value > 5)
     |       value
     |     else {
     |       throw new Exception()
     |     }
     |   }
     | }
f: scala.concurrent.Future[Int]

scala> Retry.retryFuture(10)(f).onComplete(println)

scala> var attempts = 0
var attempts = 0

scala> def m() = {
     |   attempts += 1
     |   if (attempts > 5)
     |     attempts
     |   else
     |     throw new Exception()
     | }
scala> println(Retry.retry(10)(m))


The TryWith object mimics the try-with-resources construct from Java world, or a loan pattern, where a given function can try to use a Closeable resource which shall automatically be disposed off and closed properly afterwards.

scala> import
import eu.shiftforward.apso.TryWith

scala> import eu.shiftforward.apso.TryWith
import eu.shiftforward.apso.TryWith

scala> def buildResource = new Closeable {
     |   override def toString: String = "good resource"
     |   def close(): Unit = {
     |     println("Resource is now Closed")
     |   }
     | }

scala> def goodHandler(resource: Closeable) = {
     |   println(resource)
     | }

scala> def badHandler(resource: Closeable) = {
     |   throw new Exception()
     | }

scala> TryWith(buildResource)(goodHandler)
good resource
Resource is now Closed
res2: scala.util.Try[Unit] = Success(())

scala> TryWith(buildResource)(badHandler)
Resource is now Closed
res3: scala.util.Try[Nothing] = Failure(java.lang.Exception)


The Sampler trait exposes a method to extract the first elements of a sequence given a sampling level. The sampling level is an integer and should map to a percentage.

The following implementations of Sampler are available:


The ExpSampler is a sampler in which sampling level ratios are distributed in an exponential way. Each sampling level corresponds to a sample with 1.0 / pow(base, level) of the original size:

scala> import eu.shiftforward.apso.ExpSampler
import eu.shiftforward.apso.ExpSampler

scala> val s = ExpSampler[Int](2)
s: eu.shiftforward.apso.ExpSampler[Int] = ExpSampler(2.0)

scala> s(0)((0 to 15).toSeq)
res0: Seq[Int] = Range(0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15)

scala> s(1)((0 to 15).toSeq)
res1: Seq[Int] = Range(0, 1, 2, 3, 4, 5, 6, 7)

scala> s(2)((0 to 15).toSeq)
res2: Seq[Int] = Range(0, 1, 2, 3)

scala> s(3)((0 to 15).toSeq)
res3: Seq[Int] = Range(0, 1)

scala> s(4)((0 to 15).toSeq)
res4: Seq[Int] = Range(0)


The ListSampler is a sampler in which the ratios for each sampling level are given explicitly as parameters:

scala> import eu.shiftforward.apso.ListSampler
import eu.shiftforward.apso.ListSampler

scala> val s = ListSampler[Int](0.5, 0.25, 0.125, 0.0625)
s: eu.shiftforward.apso.ListSampler[Int] = ListSampler(WrappedArray(0.5, 0.25, 0.125, 0.0625))

scala> s(0)((0 to 15).toSeq)
res0: Seq[Int] = Range(0, 1, 2, 3, 4, 5, 6, 7)

scala> s(1)((0 to 15).toSeq)
res1: Seq[Int] = Range(0, 1, 2, 3)

scala> s(2)((0 to 15).toSeq)
res2: Seq[Int] = Range(0, 1)

scala> s(3)((0 to 15).toSeq)
res3: Seq[Int] = Range(0)


The FallbackToMinimum trait allows one to set a minimum ratio for any sampling level:

scala> import eu.shiftforward.apso.ExpSampler
import eu.shiftforward.apso.ExpSampler

scala> import eu.shiftforward.apso.FallbackToMinimum
import eu.shiftforward.apso.FallbackToMinimum

scala> val s = new ExpSampler[Int](2) with FallbackToMinimum[Int] { val minSample = 0.5 }
s: eu.shiftforward.apso.ExpSampler[Int] with eu.shiftforward.apso.FallbackToMinimum[Int] = ExpSampler(2.0)

scala> s(0)((0 to 15).toSeq)
res0: Seq[Int] = Range(0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15)

scala> s(1)((0 to 15).toSeq)
res1: Seq[Int] = Range(0, 1, 2, 3, 4, 5, 6, 7)

scala> s(2)((0 to 15).toSeq)
res2: Seq[Int] = Range(0, 1, 2, 3, 4, 5, 6, 7)

scala> s(3)((0 to 15).toSeq)
res3: Seq[Int] = Range(0, 1, 2, 3, 4, 5, 6, 7)


The ShellRun object wraps the Scala's process library, facilitating the launching of shell commands.

scala> ShellRun("ls")
res0: String =

scala> ShellRun("ls", "-l")
res1: String =
"total 248
-rw-r--r--  1 jcazevedo  staff   5190 Oct 30 17:52 CHANGELOG
-rw-r--r--  1 jcazevedo  staff   6579 Oct 31 11:44
drwxr-xr-x  4 jcazevedo  staff    136 Oct 30 18:14 apso
drwxr-xr-x  4 jcazevedo  staff    136 Oct 30 18:15 apso-testkit
-rw-r--r--@ 1 jcazevedo  staff  91715 Oct 30 17:43 apso.png
drwxr-xr-x  6 jcazevedo  staff    204 Oct 30 18:20 project
-rwxr-xr-x  1 jcazevedo  staff  16035 Sep 26 14:04 sbt
drwxr-xr-x  8 jcazevedo  staff    272 Oct 30 18:14 target


The Retrier is a helper class for actors that need to retry some of the messages they send to other actors until a certain acknowledgement message (ACK) is received. Messages can be sent individually or in batches.

This class is instantiated by providing functions that extract an identifier from sent messages and from ACK messages. This can be an arbitrary identifier, as long as it uniquely associates a received ACK with the original sent message. Optional per-message filtering functions can be given, as well as the frequency of the retries and an optional timeout. Finally, the onComplete method, which is executed after a message or group of messages is acknowledged, must be implemented.

A Retrier can be used as follows:

case class ChangeData(reqId: Long, data: String)
case class ChangeDataAck(reqId: Long)
case class Replicate(reqId: Long, data: String)
case class ReplicateAck(reqId: Long)

class Master(val replica: ActorRef) extends Actor {
  import Retrier._

  val retrier = new Retrier[(ActorRef, ChangeData), Replicate, ReplicateAck, Long](_.reqId, _.reqId) {
    def onComplete(req: (ActorRef, ChangeData)) = req._1 ! ChangeDataAck(req._2.reqId)

  def receive: Receive = ({
    case msg @ ChangeData(reqId, data) =>
      // change data internally here
      retrier.dispatch((sender, msg), Replicate(reqId, data), replica)
    }: Receive).orRetryWith(retrier)

In the previous example, every time a Master actor receives a ChangeData message, it sends a Replicate message to a replica actor and only responds to the original sender after an acknowledgement from the replica is received. The Replicate message is retried periodically.


Apso provides implicit conversions from String, Seq[_], Map[_, _], Seq[Map[_, _]] and AutoCloseable to extended types that come packed with extended features.

scala> import eu.shiftforward.apso.Implicits._
import eu.shiftforward.apso.Implicits._

scala> "abcd".some
res0: Some[String] = Some(abcd)

scala> "abcd".enumerate(2)
res1: IndexedSeq[String] = Vector(aa, ab, ac, ad, ba, bb, bc, bd, ca, cb, cc, cd, da, db, dc, dd)

scala> "abcd".padLeft(10, '0')
res2: String = 000000abcd

scala> "abcd".getBytesWithNullTerminator
res3: Array[Byte] = Array(97, 98, 99, 100, 0)

scala> Seq(1, 2, 3, 4).split(2)
res4: IndexedSeq[Seq[Int]] = Vector(List(1, 2), List(3, 4))

scala> Seq(1, 2, 3, 4, 5).split(2)
res5: IndexedSeq[Seq[Int]] = Vector(List(1, 2, 3), List(4, 5))

scala> Seq(1, 2, 3, 4).sample(0.8)
res6: Seq[Int] = List(1, 2, 3)

scala> scala.util.Random.shuffle((0 to 15)).takeSmallest(3)
res7: Seq[Int] = List(0, 1, 2)

scala> scala.util.Random.shuffle((0 to 15)).takeLargest(3)
res8: Seq[Int] = List(15, 14, 13)

scala> Seq(1, 3, 5).mergeSorted(Seq(2, 4))
res9: Seq[Int] = List(1, 2, 3, 4, 5)

scala> (0 to 15).average
res10: Int = 7

scala> Iterator(1, 3, 5).buffered.mergeSorted(Iterator(2, 4).buffered).toList
res11: List[Int] = List(1, 2, 3, 4, 5)

scala> val it = Iterator(1, 2, 3, 4, 5, 6, 7, 8).buffered
it: scala.collection.BufferedIterator[Int] = non-empty iterator

scala> it.bufferedTakeWhile(_ < 4).toList
res12: List[Int] = List(1, 2, 3)

scala> it.toList
res13: List[Int] = List(4, 5, 6, 7, 8)

scala> Map(1 -> 2, 3 -> 6).merge(Map(2 -> 4, 3 -> 5)) { (a, b) => b }
res14: Map[Int,Int] = Map(1 -> 2, 3 -> 5)

scala> Map(1 -> 2, 2 -> 4, 3 -> 6).merge(Map(2 -> 2, 3 -> 5)) { (a, b) => b }
res15: Map[Int,Int] = Map(1 -> 2, 2 -> 2, 3 -> 5)

scala> Map(1 -> 2, 3 -> 6).twoWayMerge(Map(2 -> 4, 3 -> 5)) { (a, b) => b }
res16: Map[Int,Int] = Map(2 -> 4, 3 -> 5, 1 -> 2)

scala> Map(1 -> 2, 2 -> 4, 3 -> 6).twoWayMerge(Map(2 -> 2, 3 -> 5)) { (a, b) => b }
res17: Map[Int,Int] = Map(2 -> 2, 3 -> 5, 1 -> 2)

scala> Map(1 -> 2, 2 -> 3).mapKeys(_ + 1)
res18: Map[Int,Int] = Map(2 -> 2, 3 -> 3)

scala> Seq(Map(1 -> 2, 2 -> 4), Map(3 -> 6, 4 -> 8)).sequenceOnMap()
res19: Map[Int,List[Int]] = Map(1 -> List(2), 2 -> List(4), 3 -> List(6), 4 -> List(8))

scala> Seq(Map(1 -> 2, 2 -> 3), Map(1 -> 3), Map(2 -> 4, 3 -> 5)).sequenceOnMap()
res20: Map[Int,List[Int]] = Map(1 -> List(2, 3), 2 -> List(3, 4), 3 -> List(5))

scala> Seq(Map(1 -> 2, 2 -> 4), Map(3 -> 6, 4 -> 8)).sequenceOnMap(zero = Some(0))
res21: Map[Int,List[Int]] = Map(1 -> List(2, 0), 2 -> List(4, 0), 3 -> List(0, 6), 4 -> List(0, 8))

scala> Future.successful[Option[Int]](None).ifNoneOrErrorFallbackTo(Future.successful[Option[Int]](Some(4))).onComplete(println)

scala> scala.util.Random.choose((0 to 15).toSeq)
res22: Option[Int] = Some(15)

scala> scala.util.Random.choose((0 to 15).toSeq)
res23: Option[Int] = Some(12)

scala> scala.util.Random.choose((0 to 15).toSeq)
res24: Option[Int] = Some(9)

scala> scala.util.Random.choose((0 to 15).toSeq)
res25: Option[Int] = Some(2)

scala> scala.util.Random.chooseN((0 to 15).toSeq, 4)
res26: Seq[Int] = List(9, 8, 7, 6)

scala> scala.util.Random.chooseN((0 to 15).toSeq, 4)
res27: Seq[Int] = List(8, 5, 2, 1)

Amazon Web Services

Apso provides a group of classes to ease the interaction with the Amazon Web Services, namely S3 and EC2.


The ConfigCredentialsProvider is an AWSCredentialsProvider (from AWS SDK for Java) that retrieves credentials from a typesafe configuration, allowing customization of its Config object, as well as the access key and secret key paths:

scala> import

scala> import com.typesafe.config._
import com.typesafe.config._

scala> val confProvider = ConfigCredentialsProvider(
     |   config = ConfigFactory.parseString("""{
     |     aws {
     |       access-key = "<access-key>"
     |       secret-key = "<secret-key>"
     |     }
     |   }"""),
     |   accessKeyPath = "aws.access-key",
     |   secretKeyPath = "aws.secret-key")
confProvider: = ConfigCredentialsProvider(Config(SimpleConfigObject({"aws":{"access-key":"<access-key>","secret-key":"<secret-key>"}})),aws.access-key,aws.secret-key)

scala> val credentials = confProvider.getCredentials
credentials: com.amazonaws.auth.AWSCredentials = com.amazonaws.auth.BasicAWSCredentials@46056cef

scala> credentials.getAWSAccessKeyId
res0: String = <access-key>

scala> credentials.getAWSSecretKey
res1: String = <secret-key>


The CredentialStore object serves as an endpoint for the retrieval of AWS credentials from available configurations. It extends the chain in the DefaultAWSCredentialsProviderChain (from AWS SDK for Java) with the retrieval of AWS credentials through the default typesafe configuration file (typically application.conf).


The EC2 class wraps an instance of AmazonEC2 (from AWS SDK for Java), providing a higher level interface for querying the currently running instances. It provides methods to easily access an instance by its id, as well as listing all currently running instances. A method to easily terminate an instance given its id is also provided. The EC2 object provides an implicit conversion of an Instance to a RichEC2Instance, that enables the usage of methods id (to return the id of an instance) and tagValue(key) (to return the value of a tag).


The ElasticIP class provides a representation of an AWS elastic IP address. It publishes the method associateTo(instance) to allow easier association to an EC2 instance.


The InstanceMetadata object provides utilities for obtaining metadata about the EC2 instance the current process is running on. The methods are not expected to work if the JVM is not running on an EC2 instance.


The S3 class wraps an instance of AmazonS3Client (from AWS SDK for Java), providing a higher level interface for querying information about buckets and their objects. It publishes methods to easily list the buckets under the object's credentials, as well as list the objects in a bucket, filtered by an optional prefix.


The S3Bucket class wraps an instance of AmazonS3Client (from AWS SDK for Java) and exposes a higher level interface for pushing and pulling files to and from a bucket.


The SerializableAWSCredentials class provides a serializable container for AWS credentials, extending the AWSCredentials class (from AWS SDK for Java).


The eu.shiftforward.apso.collection package provides some helpful collections:


The HMap is an implementation of an heterogeneous Map, where you declare the instances for the keys explicitly:

scala> import eu.shiftforward.apso.collection._
import eu.shiftforward.apso.collection._

scala> val Key1 = new HMapKey[Int]
Key1: eu.shiftforward.apso.collection.HMapKey[Int] = eu.shiftforward.apso.collection.HMapKey@4eb14055

scala> val Key2 = new HMapKey[String]
Key2: eu.shiftforward.apso.collection.HMapKey[String] = eu.shiftforward.apso.collection.HMapKey@13590b1e

scala> val Key3 = new HMapKey[List[Boolean]]
Key3: eu.shiftforward.apso.collection.HMapKey[List[Boolean]] = eu.shiftforward.apso.collection.HMapKey@7e384bb6

scala> val map = HMap(Key1 -> 4, Key2 -> "s", Key3 -> List(false, true))
map: eu.shiftforward.apso.collection.HMap[eu.shiftforward.apso.collection.HMapKey] = HMap((eu.shiftforward.apso.collection.HMapKey@4eb14055,4), (eu.shiftforward.apso.collection.HMapKey@13590b1e,s), (eu.shiftforward.apso.collection.HMapKey@7e384bb6,List(false, true)))

scala> map(Key1)
res2: Int = 4

scala> map(Key2)
res3: String = s

scala> map(Key3)
res4: List[Boolean] = List(false, true)


The Trie class is an implementation of an immutable trie. An example usage follows:

scala> import eu.shiftforward.apso.collection._
import eu.shiftforward.apso.collection._

scala> val t = Trie[Char, Int]()
t: eu.shiftforward.apso.collection.Trie[Char,Int] = Trie(None,Map())

scala> val nt = t.set("one", 1).set("two", 2).set("three", 3).set("four", 4)
nt: eu.shiftforward.apso.collection.Trie[Char,Int] = Trie(None,Map(o -> Trie(None,Map(n -> Trie(None,Map(e -> Trie(Some(1),Map()))))), t -> Trie(None,Map(w -> Trie(None,Map(o -> Trie(Some(2),Map()))), h -> Trie(None,Map(r -> Trie(None,Map(e -> Trie(None,Map(e -> Trie(Some(3),Map()))))))))), f -> Trie(None,Map(o -> Trie(None,Map(u -> Trie(None,Map(r -> Trie(Some(4),Map())))))))))

scala> nt.get("one")
res0: Option[Int] = Some(1)

scala> nt.get("two")
res1: Option[Int] = Some(2)

scala> nt.get("five")
res2: Option[Int] = None


The TypedMap is a map that associates types with values. It can be used as follows:

scala> val m = TypedMap("one", 2, 3l)
m: eu.shiftforward.apso.collection.TypedMap[Any] = Map(java.lang.String -> one, Int -> 2, Long -> 3)

scala> m[String]
res0: String = one

scala> m[Int]
res1: Int = 2

scala> m[Long]
res2: Long = 3

scala> m.get[String]
res3: Option[String] = Some(one)

scala> m.get[Int]
res4: Option[Int] = Some(2)

scala> m.get[Long]
res5: Option[Long] = Some(3)

scala> m.get[Char]
res6: Option[Char] = None


Apso provides methods to ease working with Typesafe's config.


The LazyConfigFactory object provides static methods for creating Config instances in a lazy way. The lazy way refers to the variable loading process. The usual process loads variables in config files eagerly (i.e. the path needs to be defined in the same file it is refered to). The loading process provided by LazyConfigFactory loads and merges all configuration files and only then resolves variables. This loading process introduces a third file (beyond the default ones - application.conf and reference.conf): overrides.conf. This file has priority over the application.conf file and can be used to specify keys that should always be overriden, e.g. by environment variables.


The config.Implicits object allows one to deserialize a config to a type which has a ConfigReader implicit in scope. Most of scala's standard library types already have a ConfigReader implemented. You can also implement your own ConfigReaders. See the following for an example usage:

scala> import com.typesafe.config._
import com.typesafe.config._

scala> import eu.shiftforward.apso.config.ConfigReader.BasicConfigReaders._
import eu.shiftforward.apso.config.ConfigReader.BasicConfigReaders._

scala> import eu.shiftforward.apso.config.Implicits._
import eu.shiftforward.apso.config.Implicits._

scala> import eu.shiftforward.apso.config._
import eu.shiftforward.apso.config._

scala> import scala.concurrent.duration._
import scala.concurrent.duration._

scala> val conf = ConfigFactory.parseString("""{
     |   v1 = 2
     |   v2 = 60s
     |   v3 = "test"
     |   v4 {
     |     a = 2
     |     b = 3
     |   }
     | }""")
conf: com.typesafe.config.Config = Config(SimpleConfigObject({"v1":2,"v2":"60s","v3":"test","v4":{"a":2,"b":3}}))

scala> conf.get[Int]("v1")
res0: Int = 2

scala> conf.get[FiniteDuration]("v2")
res1: scala.concurrent.duration.FiniteDuration = 1 minute

scala> conf.get[String]("v3")
res2: String = test

scala> case class Foo(a: Int, b: Int)
defined class Foo

scala> implicit val fooConfigReader = new ConfigReader[Foo] {
     |   def apply(config: Config, key: String): Foo = {
     |     val conf = config.get[Config](key)
     |     Foo(conf.get[Int]("a"), conf.get[Int]("b"))
     |   }
     | }
fooConfigReader: eu.shiftforward.apso.config.ConfigReader[Foo] = <function2>

scala> conf.get[Foo]("v4")
res3: Foo = Foo(2,3)


Apso provides some simple utility classes to deal with encryption and decryption of data, and methods that ease the creation of the underlying Cyphers.

The following shows the creation of Encryptor and Decryptor objects, by loading a KeyStore file holding a symmetric key, and its use to encrypt and decrypt data:

scala> val encryptor = Encryptor("AES", getClass.getResourceAsStream("/keystoreFile.jceks"), "keystorePass", "keyAlias", "keyPass")
encryptor: Option[eu.shiftforward.apso.encryption.Encryptor] = Some(eu.shiftforward.apso.encryption.Encryptor@353912)

scala> val decryptor = Decryptor("AES", getClass.getResourceAsStream("/keystoreFile.jceks"), "keystorePass", "keyAlias", "keyPass")
decryptor: Option[eu.shiftforward.apso.encryption.Decryptor] = Some(eu.shiftforward.apso.encryption.Decryptor@68ccfc03)

scala> val secretData = "secret_info"
secretData: String = secret_info

// encrypt data and encode it in base64; then decrypt it to string
scala> decryptor.get.decryptToString(encryptor.get.encryptToSafeString(secretData).get)
res6: Option[String] = Some(secret_info)


Apso provides utilities for various hashing functions.

scala> import eu.shiftforward.apso.hashing.Implicits._
import eu.shiftforward.apso.hashing.Implicits._

scala> "abcd".md5
res0: String = e2fc714c4727ee9395f324cd2e7f331f

scala> "abcd".murmurHash
res1: Long = 7785666560123423118


Apso provides a tiny wrapper for Dispatch with synchronous operations. It's called W, and the following shows some sample usage:

scala> import eu.shiftforward.apso.http.W
import eu.shiftforward.apso.http.W

scala> W.get("").getStatusCode
res0: Int = 302

scala>"", "").getStatusCode
res1: Int = 405

scala> W.put("", "").getStatusCode
res2: Int = 405

scala> W.delete("").getStatusCode
res3: Int = 405

scala> W.head("").getStatusCode
res4: Int = 302


Apso provides methods to deal with IO-related features in the io package.


Apso introduces the concept of a FileDescriptor: a representation of a file stored in an arbitrary location. A descriptor includes logic to copy files to and from a local filesystem, as well as filesystem navigation logic. The following implementations of FileDescriptor are available:

  • LocalFileDescriptor (for files in the local filesystem);
  • S3FileDescriptor (for files in S3);
  • SftpFileDescriptor (for files served over SFTP).


The ResourceUtil object provides methods to access files available through Java's runtime environment classpath:

scala> import

scala> ResourceUtil.getResourceURL("reference.conf")
res0: String = /Users/jcazevedo/work/apso/apso/target/scala-2.11/classes/reference.conf

scala> ResourceUtil.getResourceStream("reference.conf")
res1: =

scala> ResourceUtil.getResourceAsString("reference.conf")
res2: String =
"apso {
  io {
    file-descriptor {
      sftp.max-connections-per-host = 8
      sftp.max-idle-time = 10s


Apso provides some utility iterators.


The CircularIterator is an iterator that iterates over its elements in a circular way. See the following for sample usage:

scala> import eu.shiftforward.apso.iterator.CircularIterator
import eu.shiftforward.apso.iterator.CircularIterator

scala> val circularIterator = CircularIterator(List(1, 2, 3).toIterator)
circularIterator: eu.shiftforward.apso.iterator.CircularIterator[Int] = non-empty iterator

scala> circularIterator.take(10).toList
res0: List[Int] = List(1, 2, 3, 1, 2, 3, 1, 2, 3, 1)


The CompositeIterator is an iterator that wraps a list of other iterators and iterates over its elements sequentially. It handles compositions of a large number of iterators in a more efficient way than simply concatenating them, avoiding stack overflows in particular. It supports appending of new iterators while keeping its efficiency. See the following for sample usage:

scala> import eu.shiftforward.apso.iterator.CompositeIterator
import eu.shiftforward.apso.iterator.CompositeIterator

scala> val compositeIterator = CompositeIterator(List(1, 2, 3).toIterator, List(4, 5, 6).toIterator, List(7, 8, 9).toIterator)
compositeIterator: eu.shiftforward.apso.iterator.CompositeIterator[Int] = non-empty iterator

scala> compositeIterator.take(9).toList
res0: List[Int] = List(1, 2, 3, 4, 5, 6, 7, 8, 9)


The ExtendedIterator is a decorator for iterators, adding more control over its lifetime. See the following for sample usage:

scala> import eu.shiftforward.apso.iterator.ExtendedIterator
import eu.shiftforward.apso.iterator.ExtendedIterator

scala> val it = (0 to 15).toIterator
it: Iterator[Int] = non-empty iterator

scala> val extIt = new ExtendedIterator(it)
extIt: eu.shiftforward.apso.iterator.ExtendedIterator[Int] = non-empty iterator

scala> extIt.onEnd(println("finished"))

scala> extIt.length
res1: Int = 16


The MergedBufferedIterator is a collection of sorted BufferedIterators that allows traversing them in order, while also providing a mergeSorted method to merge with another sorted BufferedIterator. See the following for sample usage:

scala> import eu.shiftforward.apso.iterator.MergedBufferedIterator
import eu.shiftforward.apso.iterator.MergedBufferedIterator

scala> val it1 = MergedBufferedIterator(List(
     |   (0 to 3).toIterator.buffered,
     |   (0 to 8).toIterator.buffered,
     |   (0 to 15).toIterator.buffered,
     |   (0 to 11).toIterator.buffered))
it1: eu.shiftforward.apso.iterator.MergedBufferedIterator[Int] = non-empty iterator

scala> it1.toList
res0: List[Int] = List(0, 0, 0, 0, 1, 1, 1, 1, 2, 2, 2, 2, 3, 3, 3, 3, 4, 4, 4, 5, 5, 5, 6, 6, 6, 7, 7, 7, 8, 8, 8, 9, 9, 10, 10, 11, 11, 12, 13, 14, 15)

scala> val it2 = MergedBufferedIterator(List(
     |   Iterator(1, 3, 5).buffered,
     |   Iterator(2).buffered))
it2: eu.shiftforward.apso.iterator.MergedBufferedIterator[Int] = non-empty iterator

scala> it2.mergeSorted(Iterator(4, 6).buffered).toList
res1: List[Int] = List(1, 2, 3, 4, 5, 6)


The RoundRobinIterator is an iterator that wraps an array of other iterators and iterates over its elements in a round-robin way. See the following for sample usage:

scala> import eu.shiftforward.apso.iterator.RoundRobinIterator
import eu.shiftforward.apso.iterator.RoundRobinIterator

scala> val roundRobinIterator = RoundRobinIterator(List(1, 2, 3).toIterator, List(4, 5, 6).toIterator, List(7, 8, 9).toIterator)
roundRobinIterator: eu.shiftforward.apso.iterator.RoundRobinIterator[Int] = non-empty iterator

scala> roundRobinIterator.take(9).toList
res0: List[Int] = List(1, 4, 7, 2, 5, 8, 3, 6, 9)


Apso includes a bunch of utilities to work with JSON serialization and deserialization.


The ExtraJsonProtocol object combines three traits that provide extra JsonFormats (of spray-json) for some relevant types. The JsonFormats that are provided en each trait are the following:

  • ExtraTimeJsonProtocol: JsonFormat[FiniteDuration] and JsonFormat[Interval];
  • ExtraHttpJsonProtocol: JsonFormat[URI];
  • ExtraMiscJsonProtocol: JsonFormat[Config], JsonFormat[DateTime] and JsonFormat[LocalDate]. It also includes the non-implicit method mapJsArrayFormat[K, V] which serializes a map as an array of key-value objects. Note that spray-json's JsonFormat for maps has the same signature, so if you need to use both at the same time, you need to extend the DefaultJsonProtocol trait instead of importing it.


The json package provides some implicits around spray-json's JsValue to unwrap JSON values, merge two JsValues and create JsValues from a sequence of dot-separated paths with the corresponding leaf values. See the following for sample usage:

scala> import eu.shiftforward.apso.json.Implicits._
import eu.shiftforward.apso.json.Implicits._

scala> import spray.json.DefaultJsonProtocol._
import spray.json.DefaultJsonProtocol._

scala> import spray.json._
import spray.json._

scala> "a".toJson.toValue
res0: Any = a

scala> "2".toJson.toValue
res1: Any = 2

scala> val js1 = """{
     |   "a": 2,
     |   "b": 3,
     |   "d": {
     |     "f": 6
     |   }
     | }""".parseJson.asJsObject
js1: spray.json.JsObject = {"a":2,"b":3,"d":{"f":6}}

scala> val js2 = """{
     |   "c": 4,
     |   "d": {
     |     "e": 5
     |   }
     | }""".parseJson.asJsObject
js2: spray.json.JsObject = {"c":4,"d":{"e":5}}

scala> js1.merge(js2).prettyPrint
res2: String =
  "c": 4,
  "d": {
    "e": 5,
    "f": 6
  "a": 2,
  "b": 3

scala> fromFullPaths(Seq(
     |   "a" -> 1.toJson,
     |   "b.c" -> 2.toJson,
     |   "b.d" -> 3.toJson,
     |   "e" -> "xpto".toJson,
     |   "f.g.h" -> 5.toJson)).prettyPrint
res3: String =
  "f": {
    "g": {
      "h": 5
  "e": "xpto",
  "b": {
    "d": 3,
    "c": 2
  "a": 1


The JsonConvert object contains helpers for converting between JSON values and other structures. See the following for sample usage:

scala> import eu.shiftforward.apso.json._
import eu.shiftforward.apso.json._

scala> JsonConvert.toJson("abcd")
res0: spray.json.JsValue = "abcd"

scala> JsonConvert.toJson(1)
res1: spray.json.JsValue = 1

scala> JsonConvert.toJson(Map(1 -> 2, 3 -> 4))
res2: spray.json.JsValue = {"1":2,"3":4}


The JsonFormatBuilder class provides a type-safe way to construct a JsonFormat by incrementally adding, removing or updating fields. See the following for sample usage:

scala> import eu.shiftforward.apso.json._
import eu.shiftforward.apso.json._

scala> import shapeless._
import shapeless._

scala> import spray.json.DefaultJsonProtocol._
import spray.json.DefaultJsonProtocol._

scala> import spray.json._
import spray.json._

scala> case class Test(a: Int, b: List[String], c: Double)
defined class Test

scala> val builder = JsonFormatBuilder().field[Int]("a").field[List[String]]("b")
builder: eu.shiftforward.apso.json.JsonFormatBuilder[shapeless.::[Int,shapeless.::[List[String],shapeless.HNil]],shapeless.::[eu.shiftforward.apso.json.JsonFormatBuilder.Field[Int],shapeless.::[eu.shiftforward.apso.json.JsonFormatBuilder.Field[List[String]],shapeless.HNil]]] = JsonFormatBuilder(Field(a,spray.json.BasicFormats$IntJsonFormat$@50ef5db8,None) :: Field(b,spray.json.CollectionFormats$$anon$1@4bc54f85,None) :: HNil)

scala> val jf1 = builder.jsonFormat[Test](
     |   { case a :: b :: HNil => Test(a, b, -1.0) },
     |   { test => test.a :: test.b :: HNil })
jf1: spray.json.RootJsonFormat[Test] = eu.shiftforward.apso.json.JsonFormatBuilder$$anon$1@70ae62e7

scala> """{ "a": 3, "b": ["x", "y"] }""".parseJson.convertTo[Test](jf1)
res0: Test = Test(3,List(x, y),-1.0)

scala> """{ "a": 3, "b": ["x", "y"], "c": 3.0 }""".parseJson.convertTo[Test](jf1)
res1: Test = Test(3,List(x, y),-1.0)

scala> Test(3, List("x", "y"), -1.0).toJson(jf1)
res2: spray.json.JsValue = {"b":["x","y"],"a":3}

scala> val builder2 = builder.field[Double]("c", 0.0)
builder2: eu.shiftforward.apso.json.JsonFormatBuilder[shapeless.::[Int,shapeless.::[List[String],shapeless.::[Double,shapeless.HNil]]],shapeless.::[eu.shiftforward.apso.json.JsonFormatBuilder.Field[Int],shapeless.::[eu.shiftforward.apso.json.JsonFormatBuilder.Field[List[String]],shapeless.::[eu.shiftforward.apso.json.JsonFormatBuilder.Field[Double],shapeless.HNil]]]] = JsonFormatBuilder(Field(a,spray.json.BasicFormats$IntJsonFormat$@50ef5db8,None) :: Field(b,spray.json.CollectionFormats$$anon$1@4bc54f85,None) :: Field(c,spray.json.BasicFormats$DoubleJsonFormat$@73930900,Some(0.0)) :: HNil)

scala> val jf2 = builder2.jsonFormat[Test](
     |   { case a :: b :: c :: HNil => Test(a, b, c) },
     |   { test => test.a :: test.b :: test.c :: HNil })
jf2: spray.json.RootJsonFormat[Test] = eu.shiftforward.apso.json.JsonFormatBuilder$$anon$1@6ce596d9

scala> """{ "a": 3, "b": ["x", "y"] }""".parseJson.convertTo[Test](jf2)
res3: Test = Test(3,List(x, y),0.0)

scala> """{ "a": 3, "b": ["x", "y"], "c": 3.0 }""".parseJson.convertTo[Test](jf2)
res4: Test = Test(3,List(x, y),3.0)

scala> Test(3, List("x", "y"), 0.0).toJson(jf2)
res5: spray.json.JsValue = {"c":0.0,"b":["x","y"],"a":3}


The JsonHMap defines an heterogeneous map with JSON (de)serialization capabilities. See the following for sample usage:

scala> import spray.json._
import spray.json._

scala> import spray.json.DefaultJsonProtocol._
import spray.json.DefaultJsonProtocol._

scala> import eu.shiftforward.apso.json._
import eu.shiftforward.apso.json._

scala> import eu.shiftforward.apso.json.JsonHMap._
import eu.shiftforward.apso.json.JsonHMap._

scala> import eu.shiftforward.apso.collection._
import eu.shiftforward.apso.collection._

scala> implicit val reg = new JsonKeyRegistry {}
reg: eu.shiftforward.apso.json.JsonKeyRegistry = $anon$1@4213d40

scala> val Key1 = new JsonHMapKey[Int]('key1) {}
Key1: eu.shiftforward.apso.json.JsonHMapKey[Int] = 'key1

scala> val Key2 = new JsonHMapKey[String]('key2) {}
Key2: eu.shiftforward.apso.json.JsonHMapKey[String] = 'key2

scala> val Key3 = new JsonHMapKey[List[Boolean]]('key3) {}
Key3: eu.shiftforward.apso.json.JsonHMapKey[List[Boolean]] = 'key3

scala> val json =
           |  "key1": 4,
           |  "key2": "s",
           |  "key3": [ false, true ]
json: String =
  "key1": 4,
  "key2": "s",
  "key3": [ false, true ]

scala> val map = json.asJson.convertTo[JsonHMap]
map: eu.shiftforward.apso.json.JsonHMap.JsonHMap = HMap(('key3,List(false, true)), ('key2,s), ('key1,4))


The profiling package of apso provides utilities to help with profiling the running process.


The CpuSampler is a lightweight configurable CPU profiler based on call stack sampling. When run as a thread, it periodically captures the call stacks of all live threads and maintains counters for each leaf method. The counters are then dumped to a logger with a given periodicity (most probably greater than the sampling period). Each data row written to the logger contains a timestamp, the method profiled, its location in the source code and the associated absolute counters and relative weight.


The SimpleJmx trait allows mixing in a simple JMX server. The JMX server is configured through a Config object, where the parameters host and port can be set. When behind a firewall, both the port defined (the RMI registry port) and the port + 1 port (the RMI server port) need to be open. In the event of a binding failure to the defined port, a retry is performed with a random port.


The scalaz package provides implicit methods to convert between Scala's Try and Scalaz's Validation. See the following for a sample usage:

scala> import eu.shiftforward.apso.scalaz.Implicits._
import eu.shiftforward.apso.scalaz.Implicits._

scala> import scala.util._
import scala.util._

scala> import scalaz._
import scalaz._

scala> Try(2): Validation[Throwable, Int]
res0: scalaz.Validation[Throwable,Int] = Success(2)

scala> Try(throw new Exception()): Validation[Throwable, Int]
res1: scalaz.Validation[Throwable,Int] = Failure(java.lang.Exception)


The spray package provides additional directives to be used in spray-routing.


The ClientIPDirectives trait exposes an optionalRawClientIP directive that extracts the raw IP of the client from either the X-Forwarded-For, Remote-Address or X-Real-IP header, in that order of priority.


The ExtraMiscDirectives trait exposes the directives cacheControlMaxAge(maxAgeDuration) and optionalRefererHost to set the cache-control header to the supplied finite duration (the minimum resolution is 1 second) to extract the referer from the HTTP request header, respectively. The ExtraMiscDirectives companion object exposes a cacheControlNoCache directive to reply with the no-cache option in the Cache-Control header.


The Implicits companion object exposes an implicit method that provides a Marshaller for Scalaz's Validation.


The ProxySupport traits adds helper methods to proxy requests to a given uri, either directly (proxyTo), or with the unmatched path and query parameters of the current context (proxyToUnmatchedPath).


The eu.shiftforward.apso.time package provides utilities to work with DateTime and LocalDate. It mainly adds support for better working with intervals. See the following sample usage:

scala> import com.github.nscala_time.time.Imports._
import com.github.nscala_time.time.Imports._

scala> import eu.shiftforward.apso.time._
import eu.shiftforward.apso.time._

scala> import eu.shiftforward.apso.time.Implicits._
import eu.shiftforward.apso.time.Implicits._

scala> (new DateTime("2012-01-01") to new DateTime("2012-01-01")).toList
res0: List[com.github.nscala_time.time.Imports.DateTime] = List(2012-01-01T00:00:00.000Z)

scala> (new DateTime("2012-02-01") until new DateTime("2012-03-01") by
res1: eu.shiftforward.apso.time.IterableInterval = SteppedInterval(2012-02-01T00:00:00.000Z, 2012-02-02T00:00:00.000Z, 2012-02-03T00:00:00.000Z, 2012-02-04T00:00:00.000Z, 2012-02-05T00:00:00.000Z, 2012-02-06T00:00:00.000Z, 2012-02-07T00:00:00.000Z, 2012-02-08T00:00:00.000Z, 2012-02-09T00:00:00.000Z, 2012-02-10T00:00:00.000Z, 2012-02-11T00:00:00.000Z, 2012-02-12T00:00:00.000Z, 2012-02-13T00:00:00.000Z, 2012-02-14T00:00:00.000Z, 2012-02-15T00:00:00.000Z, 2012-02-16T00:00:00.000Z, 2012-02-17T00:00:00.000Z, 2012-02-18T00:00:00.000Z, 2012-02-19T00:00:00.000Z, 2012-02-20T00:00:00.000Z, 2012-02-21T00:00:00.000Z, 2012-02-22T00:00:00.000Z, 2012-02-23T00:00:00.000Z, 2012-02-24T00:00:00.000Z, 2012-02-25T00:00:00.000Z, 2012-02-26T00:00:00.000Z, 2012-02-27T00:00:00.000Z, 2012-02-28T00:00:00.000Z, 20...

scala> (new DateTime("2012-01-01") until new DateTime("2012-02-01") by 2.minutes)
res2: eu.shiftforward.apso.time.IterableInterval = SteppedInterval(2012-01-01T00:00:00.000Z, 2012-01-01T00:02:00.000Z, 2012-01-01T00:04:00.000Z, 2012-01-01T00:06:00.000Z, 2012-01-01T00:08:00.000Z, 2012-01-01T00:10:00.000Z, 2012-01-01T00:12:00.000Z, 2012-01-01T00:14:00.000Z, 2012-01-01T00:16:00.000Z, 2012-01-01T00:18:00.000Z, 2012-01-01T00:20:00.000Z, 2012-01-01T00:22:00.000Z, 2012-01-01T00:24:00.000Z, 2012-01-01T00:26:00.000Z, 2012-01-01T00:28:00.000Z, 2012-01-01T00:30:00.000Z, 2012-01-01T00:32:00.000Z, 2012-01-01T00:34:00.000Z, 2012-01-01T00:36:00.000Z, 2012-01-01T00:38:00.000Z, 2012-01-01T00:40:00.000Z, 2012-01-01T00:42:00.000Z, 2012-01-01T00:44:00.000Z, 2012-01-01T00:46:00.000Z, 2012-01-01T00:48:00.000Z, 2012-01-01T00:50:00.000Z, 2012-01-01T00:52:00.000Z, 2012-01-01T00:54:00.000Z, 20...


Apso comes with a TestKit with extra useful matchers for specs2. The following traits with extra matchers are available:

  • CustomMatchers: provides a matcher to check if an object is serializable and one to check if a file exists;
  • FutureExtraMatchers: provides extra matchers for futures and implicit conversions for awaitables;
  • JreVersionTestHelper: provides a wrapper for AsResult to only run a spec if a specific JRE version is satisfied;
  • TestHelper: provides a helper method to create a temporary directory that is deleted on exit;
  • ActorMatchers: provides various matchers to be used on akka's TestKit probes and check for different behaviours regarding the reception of messages.