A mini Scala utility library. Compatible with functional programming beginners. For the JVM and Scala.js.
Currently, Hamsters supports :
- Validation
- Enum typeclass
- Monad transformers
- HLists
- Union types
With SBT :
libraryDependencies ++= Seq(
"io.github.scala-hamsters" %% "hamsters" % "1.4.1"
)
With Maven :
<dependency>
<groupId>io.github.scala-hamsters</groupId>
<artifactId>hamsters_${scala.version}</artifactId>
<version>1.4.1</version>
</dependency>
For Scala.js :
libraryDependencies ++= Seq(
"io.github.scala-hamsters" %%% "hamstersjs" % "1.4.1"
)
Statements can be succesful or fail. Validation helps you to retrieve values or failures for a sequence of statements.
Note : Validation relies on standard Either types.
Left is used for failures, Right for correct results.
You can also use KO as an alias for Left and OK as an alias for Right.
import io.github.hamsters.Validation
import Validation._
val e1 = Right(1)
val e2 = Left("error 1")
val e3 = Left("error 2")
val e4 = Right("4")
Validation.result(e1,e2, e3) // List[String] : Left(List("error 1", "error 2"))
Validation.result(e1, e4) // Right((1, "4"))
Validation.failures(e1,e2, e3) // List[String] : List("error 1", "error 2")
Validation.failures(e1, e4) // Nil
To automatically catch exceptions into a Left or KO object, you can use fromCatchable
.
By default it will give you an error message, but you can specify what to do in error cases :
def compute(x: Int):Int = 2/x
fromCatchable(compute(1)) //Right(2)
fromCatchable(compute(0)) //Left("/ by zero")
fromCatchable(compute(0), (t: Throwable) => t.getClass.getSimpleName) //Left("ArithmeticException")
Note : You can also use Right/Left (or OK/KO) in a monadic way (for example in a for-comprehension) if you want to stop processing at the first encountered error, even with Scala 2.11.
To make it work with Scala 2.11, just import Validation.OKBiasedEither.
Example : combine Future and Option types then make it work in a for comprehension. More information on why it's useful here.
def foa: Future[Option[String]] = Future(Some("a"))
def fob(a: String): Future[Option[String]] = Future(Some(a+"b"))
val composedAB: Future[Option[String]] = for {
a <- FutureOption(foa)
ab <- FutureOption(fob(a))
} yield ab
import io.github.hamsters.Validation._
import io.github.hamsters.{FutureEither, FutureOption}
import io.github.hamsters.MonadTransformers._
def fea: Future[Either[String, Int]] = Future(Right(1))
def feb(a: Int): Future[Either[String, Int]] = Future(Right(a+2))
val composedAB: Future[Either[String, Int]] = for {
a <- FutureEither(fea)
ab <- FutureEither(feb(a))
} yield ab
Note : For Future[Try] composition, see Future.fromTry
to get a single future success or failure.
This typeclass allows to use parse and name methods on enumerable types. It can be very useful if you need to serialize and deserialize your types (in Json, in a database...)
sealed trait Season
case object Winter extends Season
case object Spring extends Season
case object Summer extends Season
case object Fall extends Season
implicit val seasonEnumerable = new Enumerable[Season] {
override def list: List[Season] = List(Winter, Spring, Summer, Fall)
}
Enumeration.name(Winter) // "winter"
Enumeration.parse[Season]("winter") // Some(Winter)
It is also possible to use custom namings :
implicit val seasonEnumerable = new Enumerable[Season] {
override def list = List(Winter, Spring, Summer, Fall)
override def name(s: Season) = {
s match {
case Winter => "WINTER_SEASON"
case other => super.name(other)
}
}
}
Enumeration.name(Winter) // "WINTER_SEASON"
Enumeration.parse[Season]("WINTER_SEASON") // Some(Winter)
HLists can contain heterogeneous data types but are strongly typed. It's like tuples on steroids! When you're manipulating data using tuples, it's common to add or subtrack some elements, but you have to make each element explicit to build a new tuple. HList simplifies this kind of task.
::
is used to append elements at the beginning of an HList, and to construct new HLists+
is used add element at the end of a HList++
is used to concatenate 2 Hlists- HNil is the empty HList
- other operations : filter, map, foldLeft, foreach
import io.github.hamsters.{HList, HCons, HNil}
import HList._
val hlist = 2.0 :: "hi" :: HNil
val hlist1 = 2.0 :: "hi" :: HNil
val hlist2 = 1 :: HNil
val sum = hlist1 + 1 // 2.0 :: "hi" :: 1 :: HNil
val sum2 = hlist1 ++ hlist2 // 2.0 :: "hi" :: 1 : HNil
sum2.tail // hi :: (1 : HNil)
sum2.head // 2.0 (Double)
sum2.tail.head // hi (String)
// Retrieve element by index and type
hlist.get[String](1) // Some("hi")
// Or use apply to avoid Option
hlist[String](1) // "hi"
(2.0 :: "hi" :: HNil).foldLeft("")(_+_) // "2.0hi"
(2.0 :: "hi" :: HNil).map(_.toString) // "2.0" :: "hi" :: HNil
(2.0 :: "hi" :: HNil).filter {
case s: String if s.startsWith("h") => true
case _ => false
} //"hi" :: HNil
You can define functions or methods that are able to return several types, depending on the context.
import io.github.hamsters.{Union3, Union3Type}
//json element can contain a String, a Int or a Double
val jsonUnion = new Union3Type[String, Int, Double]
import jsonUnion._
def jsonElement(x: Int): Union3[String, Int, Double] = {
if(x == 0) "0"
else if (x % 2 == 0) 1
else 2.0
}
Then you can use pattern matching, or ask for a specifc type and retrieve an option :
jsonElement(0).get[String] // Some("0")
jsonElement(1).getOrElse("not found") // get String value or "not found" if get[String] is undefined
See hamsters-extensions for more information.
You can find the API documentation here.