On a recent worksheet I was presented with the question asking what would be the output of the following code:
class A { def m(x:Double) = x+x }
class B[Any] extends A{ def m(x: Any) = print(x) }
class C[Any] { def m (x:Double) = x+x; def m (x: Any) = print(x) }
val obj1 = new B[Int]; val obj2 = new C[Any]
obj1.m(1); obj1.m(2.3); obj2.m(4); obj2.m(5.6)
I'm quite confused as to what having a concrete type in the square brackets after the class name would mean (i.e. class B[Any]). Is the later expression val obj1 = new B[Int] valid because Int <: Any, Int being a subclass of Any?
When later running the code snippet, the result given was simply "1" being printed. This was not what I had expected the call to obj.m(2.3) to resolve at def m(x: any), where it seems in actuality the compiler went up to A and called the m in class A.
The later expressions, obj2.m(4) and obj2.m(5.6) seems to make sense as both 4 and 5.6 would land in the function with def m(x: Double), thus not print anything out.
In what order exactly does the compiler traverse to find what to call? I'd be very grateful if someone could clear up my confusions with how polymorphism is handled here by Scala, thank you very much :)
When you do class B[Any], you define a class with a type parameter called Any. Don't confuse the type parameter name with the actual class Any. You are just shadowing its name.
You could just as fine do this:
class B[Int]
val obj = new B[String]
You may see why it is bad practice to name type parameters after actual types. Usually, people use single letter names for their type parameters, like this:
class B[T] // I just changed the name of the type parameter from "Int" to "T".
val obj = new B[String]
Related
I'd like to enrich a 'graph for scala' graph. For this purpose i've created an implicit value class:
import scalax.collection.mutable
import scalax.collection.edge.DiEdge
...
type Graph = mutable.Graph[Int, DiEdge]
implicit class EnrichGraph(val G: Graph) extends AnyVal {
def roots = G.nodes.filter(!_.hasPredecessors)
...
}
...
The problem lies with the return type of its methods, e.g.:
import ....EnrichGraph
val H: Graph = mutable.Graph[Int,DiEdge]()
val roots1 = H.nodes.filter(!_.hasPredecessors) // type Iterable[H.NodeT]
val roots2 = H.roots // type Iterable[RichGraph#G.NodeT] !!
val subgraph1 = H.filter(H.having(roots1)) // works!
val subgraph2 = H.filter(H.having(roots2)) // type mismatch!
Does the cause lie with fact that 'Graph' has dependent subtypes, e.g. NodeT? Is there a way to make this enrichment work?
What usually works is propagating the singleton type as a type parameter to EnrichGraph. That means a little bit of extra boilerplate since you have to split the implicit class into a class and an implicit def.
class EnrichGraph[G <: Graph](val G: G) extends AnyVal {
def roots: Iterable[G#NodeT] = G.nodes.filter(!_.hasPredecessors)
//...
}
implicit def EnrichGraph(g: Graph): EnrichGraph[g.type] = new EnrichGraph[g.type](g)
The gist here being that G#NodeT =:= H.NodeT if G =:= H.type, or in other words (H.type)#NodeT =:= H.NodeT. (=:= is the type equality operator)
The reason you got that weird type, is that roots has a path type dependent type. And that path contains the value G. So then the type of val roots2 in your program would need to contain a path to G. But since G is bound to an instance of EnrichGraph which is not referenced by any variable, the compiler cannot construct such a path. The "best" thing the compiler can do is construct a type with that part of the path left out: Set[_1.G.NodeT] forSome { val _1: EnrichGraph }. This is the type I actually got with your code; I assume you're using Intellij which is printing this type differently.
As pointed out by #DmytroMitin a version which might work better for you is:
import scala.collection.mutable.Set
class EnrichGraph[G <: Graph](val G: G) extends AnyVal {
def roots: Set[G.NodeT] = G.nodes.filter(!_.hasPredecessors)
//...
}
implicit def EnrichGraph(g: Graph): EnrichGraph[g.type] = new EnrichGraph[g.type](g)
Since the rest of your code actually requires a Set instead of an Iterable.
The reason why this still works despite reintroducing the path dependent type is quite tricky. Actually now roots2 will receive the type Set[_1.G.NodeT] forSome { val _1: EnrichGraph[H.type] } which looks pretty complex. But the important part is that this type still contains the knowledge that the G in _1.G.NodeT has type H.type because that information is stored in val _1: EnrichGraph[H.type].
With Set you can't use G#NodeT to give you the simpler type signatures, because G.NodeT is a subtype of G#NodeT and Set is unfortunately invariant. In our usage those type will actually always be equivalent (as I explained above), but the compiler cannot know that.
See this implementation follows the upper bound example http://docs.scala-lang.org/tutorials/tour/upper-type-bounds.html
class Fruit(name: String)
class Apple (name: String) extends Fruit(name)
class Orange(name: String) extends Fruit(name)
class BigOrange(name:String) extends Orange(name)
class BigFLOrange(name:String) extends BigOrange(name)
// Straight from the doc
trait Node[+B ] {
def prepend[U >: B ](elem: U)
}
case class ListNode[+B](h: B, t: Node[B]) extends Node[B] {
def prepend[U >:B ](elem: U) = ListNode[U](elem, this)
def head: B = h
def tail = t
}
case class Nil[+B ]() extends Node[B] {
def prepend[U >: B ](elem: U) = ListNode[U](elem, this)
}
But this definition seems to allow multiple unrelated things in the same container
val f = new Fruit("fruit")
val a = new Apple("apple")
val o = new Orange("orange")
val bo = new BigOrange("big orange")
val foo :ListNode[BigOrange] = ListNode[BigOrange](bo, Nil())
foo.prepend(a) // add an apple to BigOrangeList
foo.prepend(o) // add an orange to BigOrangeList
val foo2 : ListNode[Orange] = foo // and still get to assign to OrangeList
So I am not sure this is a great example in the docs. And, question, how doe I modify the constraints so that..this behaves more like a List?
User #gábor-bakos points out that I am confusing invariance with covariance. So I tried the mutable list buffer. It does not later allow apple to be inserted into an Orange list Buffer, but it is not covariant
val ll : ListBuffer[BigOrange]= ListBuffer(bo)
ll += bo //good
ll += a // not allowed
So..can my example above (ListNode) be modified so that
1. it is covariant (it already is)
2 It is mutable, but mutable like the ListBuffer example (will not later allow apples to be inserted into BigOrange list
A mutable list cannot/should not be covariant in its argument type.
Exactly because of the reason you noted.
Suppose, you could have a MutableList[Orange], that was a subclass of a MutableList[Fruit]. Now, there is nothing that would prevent you from making a function:
def putApple(fruits: MutableList[Fruit], idx: Int) =
fruits(idx) = new Apple
You can add Apple to the list of Fruits, because Apple is a Fruit, nothing wrong with this.
But once you have a function like that, there is no reason you can't call it like this:
val oranges = new MutableList[Orange](new Orange, new Orange)
putApple(oranges, 0)
This will compile, since MutableList[Orange] is a subclass of MutableList[Fruit]. But now:
val firstOrange: Orange = oranges(0)
will crash, because the first element of oranges is actually an Apple.
For this reason, mutable collections have to be invariant in the element type (to answer the question you asked in the comments, to make the list invariant remove the + before B, and also get rid of the type parameter in prepend. It should just be def pretend(elem: B)).
How to get around it? The best solution is to simply not use mutable collections. You should not need them in 99% or real life scala code. If you think you need one, there is a 99% you are doing something wrong.
The major thing that you are probably missing is that prepend does not modify a list. In the line val foo2 : ListNode[Orange] = foo the list foo is still of type ListNode[BigOrange] and given covariance of parameters this assignment is valid (and there is nothing particularly awkward about it). Compiler will prevent you from assigning Fruits to Oranges (in this case assigning an Apple to an Orange is hazardous) but you have to save modified lists beforehand:
val foo: ListNode[Fruit] = ListNode[BigOrange](bo, Nil()).prepend(a).prepend(o)
val foo2: ListNode[Orange] = foo // this is not valid
Moreover your Node definition lacks return type for prepend - thus compiler infers wrong return type (Unit instead of ListNode[U]).
Here's fixed version:
trait Node[+B] {
def prepend[U >: B ](elem: U): Node[U]
}
I'm trying to implement something like clever parameters converter function with Scala.
Basically in my program I need to read parameters from a properties file, so obviously they are all strings and I would like then to convert each parameter in a specific type that I pass as parameter.
This is the implementation that I start coding:
def getParam[T](key : String , value : String, paramClass : T): Any = {
value match {
paramClass match {
case i if i == Int => value.trim.toInt
case b if b == Boolean => value.trim.toBoolean
case _ => value.trim
}
}
/* Exception handling is missing at the moment */
}
Usage:
val convertedInt = getParam("some.int.property.key", "10", Int)
val convertedBoolean = getParam("some.boolean.property.key", "true", Boolean)
val plainString = getParam("some.string.property.key", "value",String)
Points to note:
For my program now I need just 3 main type of type: String ,Int and Boolean,
if is possible I would like to extends to more object type
This is not clever, cause I need to explicit the matching against every possibile type to convert, I would like an more reflectional like approach
This code doesn't work, it give me compile error: "object java.lang.String is not a value" when I try to convert( actually no conversion happen because property values came as String).
Can anyone help me? I'm quite newbie in Scala and maybe I missing something
The Scala approach for a problem that you are trying to solve is context bounds. Given a type T you can require an object like ParamMeta[T], which will do all conversions for you. So you can rewrite your code to something like this:
trait ParamMeta[T] {
def apply(v: String): T
}
def getParam[T](key: String, value: String)(implicit meta: ParamMeta[T]): T =
meta(value.trim)
implicit case object IntMeta extends ParamMeta[Int] {
def apply(v: String): Int = v.toInt
}
// and so on
getParam[Int](/* ... */, "127") // = 127
There is even no need to throw exceptions! If you supply an unsupported type as getParam type argument, code will even not compile. You can rewrite signature of getParam using a syntax sugar for context bounds, T: Bound, which will require implicit value Bound[T], and you will need to use implicitly[Bound[T]] to access that values (because there will be no parameter name for it).
Also this code does not use reflection at all, because compiler searches for an implicit value ParamMeta[Int], founds it in object IntMeta and rewrites function call like getParam[Int](..., "127")(IntMeta), so it will get all required values at compile time.
If you feel that writing those case objects is too boilerplate, and you are sure that you will not need another method in these objects in future (for example, to convert T back to String), you can simplify declarations like this:
case class ParamMeta[T](f: String => T) {
def apply(s: String): T = f(s)
}
implicit val stringMeta = ParamMeta(identity)
implicit val intMeta = ParamMeta(_.toInt)
To avoid importing them every time you use getParam you can declare these implicits in a companion object of ParamMeta trait/case class, and Scala will pick them automatically.
As for original match approach, you can pass a implicit ClassTag[T] to your function, so you will be able to match classes. You do not need to create any values for ClassTag, as the compiler will pass it automatically. Here is a simple example how to do class matching:
import scala.reflect.ClassTag
import scala.reflect._
def test[T: ClassTag] = classTag[T].runtimeClass match {
case x if x == classOf[Int] => "I'm an int!"
case x if x == classOf[String] => "I'm a string!"
}
println(test[Int])
println(test[String])
However, this approach is less flexible than ParamMeta one, and ParamMeta should be preferred.
This example code is based on Atmosphere classes, but if someone could give me some insights into what the error means in general, I think I can figure out any Atmosphere-specific solution...
val bc = BroadcasterFactory.getDefault().lookup(_broadcasterId)
bc.broadcast(message)
After the first line, bc should contain a handle to an object whose class definition includes the method broadcast() -- in fact, it contains several overloaded variations. However, the compiler chokes on the second line of code with the following: "value broadcast is not a member of Nothing"
Any ideas/suggestions on what would be causing this?
Thanks.
EDIT: signature for [BroadcasterFactor].lookup :
abstract Broadcaster lookup(Object id)
Note: 1) that is the signature version that I've used in the example, 2) it is the java Inteface signature - whereas the getDefault() hands back an instantiated object that implements that interface.
Solution: force type cast on value:
val bc: Broadcaster = BroadcasterFactory.getDefault().lookup(_broadcasterId)
Nothing is the type name. It's the subtype of all other types. You can't call methods from Nothing itself, you have to specify exact type ((bc: ExactType).broadcast(message)). Nothing has no instances. Method, that returns Nothing will, actually, never return value. It will throw an exception eventually.
Type inference
Definition of lookup:
abstract public <T extends Broadcaster> T lookup(Object id);
in scala this definition looks this way:
def lookup[T <: Broadcaster](Object id): T
There is not specified type parameter in lookup method. In this case compiler will infer this type parameter as the most specific type - Nothing:
scala> def test[T](i: Int): T = ???
test: [T](i: Int)T
scala> lazy val x = test(1)
x: Nothing = <lazy>
scala> lazy val x = test[String](1)
x: String = <lazy>
You could specify type parameter like this:
val bc = BroadcasterFactory.getDefault().lookup[Broadcaster](_broadcasterId)
Draft implementation
In development process lookup can be "implemented" like this:
def lookup(...) = ???
??? returns Nothing.
You should specify either result type of lookup method like this: def lookup(...): <TypeHere> = ... or type of bc: val bc: <TypeHere> =.
I was making my way through the Scala playframework tutorial and I came across this snippet of code which had me puzzled:
def newTask = Action { implicit request =>
taskForm.bindFromRequest.fold(
errors => BadRequest(views.html.index(Task.all(), errors)),
label => {
Task.create(label)
Redirect(routes.Application.tasks())
}
)
}
So I decided to investigate and came across this post.
I still don't get it.
What is the difference between this:
implicit def double2Int(d : Double) : Int = d.toInt
and
def double2IntNonImplicit(d : Double) : Int = d.toInt
other than the obvious fact they have different method names.
When should I use implicit and why?
I'll explain the main use cases of implicits below, but for more detail see the relevant chapter of Programming in Scala.
Implicit parameters
The final parameter list on a method can be marked implicit, which means the values will be taken from the context in which they are called. If there is no implicit value of the right type in scope, it will not compile. Since the implicit value must resolve to a single value and to avoid clashes, it's a good idea to make the type specific to its purpose, e.g. don't require your methods to find an implicit Int!
example:
// probably in a library
class Prefixer(val prefix: String)
def addPrefix(s: String)(implicit p: Prefixer) = p.prefix + s
// then probably in your application
implicit val myImplicitPrefixer = new Prefixer("***")
addPrefix("abc") // returns "***abc"
Implicit conversions
When the compiler finds an expression of the wrong type for the context, it will look for an implicit Function value of a type that will allow it to typecheck. So if an A is required and it finds a B, it will look for an implicit value of type B => A in scope (it also checks some other places like in the B and A companion objects, if they exist). Since defs can be "eta-expanded" into Function objects, an implicit def xyz(arg: B): A will do as well.
So the difference between your methods is that the one marked implicit will be inserted for you by the compiler when a Double is found but an Int is required.
implicit def doubleToInt(d: Double) = d.toInt
val x: Int = 42.0
will work the same as
def doubleToInt(d: Double) = d.toInt
val x: Int = doubleToInt(42.0)
In the second we've inserted the conversion manually; in the first the compiler did the same automatically. The conversion is required because of the type annotation on the left hand side.
Regarding your first snippet from Play:
Actions are explained on this page from the Play documentation (see also API docs). You are using
apply(block: (Request[AnyContent]) ⇒ Result): Action[AnyContent]
on the Action object (which is the companion to the trait of the same name).
So we need to supply a Function as the argument, which can be written as a literal in the form
request => ...
In a function literal, the part before the => is a value declaration, and can be marked implicit if you want, just like in any other val declaration. Here, request doesn't have to be marked implicit for this to type check, but by doing so it will be available as an implicit value for any methods that might need it within the function (and of course, it can be used explicitly as well). In this particular case, this has been done because the bindFromRequest method on the Form class requires an implicit Request argument.
WARNING: contains sarcasm judiciously! YMMV...
Luigi's answer is complete and correct. This one is only to extend it a bit with an example of how you can gloriously overuse implicits, as it happens quite often in Scala projects. Actually so often, you can probably even find it in one of the "Best Practice" guides.
object HelloWorld {
case class Text(content: String)
case class Prefix(text: String)
implicit def String2Text(content: String)(implicit prefix: Prefix) = {
Text(prefix.text + " " + content)
}
def printText(text: Text): Unit = {
println(text.content)
}
def main(args: Array[String]): Unit = {
printText("World!")
}
// Best to hide this line somewhere below a pile of completely unrelated code.
// Better yet, import its package from another distant place.
implicit val prefixLOL = Prefix("Hello")
}
In scala implicit works as:
Converter
Parameter value injector
Extension method
There are some uses of Implicit
Implicitly type conversion : It converts the error producing assignment into intended type
val x :String = "1"
val y:Int = x
String is not the sub type of Int , so error happens in line 2. To resolve the error the compiler will look for such a method in the scope which has implicit keyword and takes a String as argument and returns an Int .
so
implicit def z(a:String):Int = 2
val x :String = "1"
val y:Int = x // compiler will use z here like val y:Int=z(x)
println(y) // result 2 & no error!
Implicitly receiver conversion: We generally by receiver call object's properties, eg. methods or variables . So to call any property by a receiver the property must be the member of that receiver's class/object.
class Mahadi{
val haveCar:String ="BMW"
}
class Johnny{
val haveTv:String = "Sony"
}
val mahadi = new Mahadi
mahadi.haveTv // Error happening
Here mahadi.haveTv will produce an error. Because scala compiler will first look for the haveTv property to mahadi receiver. It will not find. Second it will look for a method in scope having implicit keyword which take Mahadi object as argument and returns Johnny object. But it does not have here. So it will create error. But the following is okay.
class Mahadi{
val haveCar:String ="BMW"
}
class Johnny{
val haveTv:String = "Sony"
}
val mahadi = new Mahadi
implicit def z(a:Mahadi):Johnny = new Johnny
mahadi.haveTv // compiler will use z here like new Johnny().haveTv
println(mahadi.haveTv)// result Sony & no error
Implicitly parameter injection: If we call a method and do not pass its parameter value, it will cause an error. The scala compiler works like this - first will try to pass value, but it will get no direct value for the parameter.
def x(a:Int)= a
x // ERROR happening
Second if the parameter has any implicit keyword it will look for any val in the scope which have the same type of value. If not get it will cause error.
def x(implicit a:Int)= a
x // error happening here
To slove this problem compiler will look for a implicit val having the type of Int because the parameter a has implicit keyword.
def x(implicit a:Int)=a
implicit val z:Int =10
x // compiler will use implicit like this x(z)
println(x) // will result 10 & no error.
Another example:
def l(implicit b:Int)
def x(implicit a:Int)= l(a)
we can also write it like-
def x(implicit a:Int)= l
Because l has a implicit parameter and in scope of method x's body, there is an implicit local variable(parameters are local variables) a which is the parameter of x, so in the body of x method the method-signature l's implicit argument value is filed by the x method's local implicit variable(parameter) a implicitly.
So
def x(implicit a:Int)= l
will be in compiler like this
def x(implicit a:Int)= l(a)
Another example:
def c(implicit k:Int):String = k.toString
def x(a:Int => String):String =a
x{
x => c
}
it will cause error, because c in x{x=>c} needs explicitly-value-passing in argument or implicit val in scope.
So we can make the function literal's parameter explicitly implicit when we call the method x
x{
implicit x => c // the compiler will set the parameter of c like this c(x)
}
This has been used in action method of Play-Framework
in view folder of app the template is declared like
#()(implicit requestHreader:RequestHeader)
in controller action is like
def index = Action{
implicit request =>
Ok(views.html.formpage())
}
if you do not mention request parameter as implicit explicitly then you must have been written-
def index = Action{
request =>
Ok(views.html.formpage()(request))
}
Extension Method
Think, we want to add new method with Integer object. The name of the method will be meterToCm,
> 1 .meterToCm
res0 100
to do this we need to create an implicit class within a object/class/trait . This class can not be a case class.
object Extensions{
implicit class MeterToCm(meter:Int){
def meterToCm={
meter*100
}
}
}
Note the implicit class will only take one constructor parameter.
Now import the implicit class in the scope you are wanting to use
import Extensions._
2.meterToCm // result 200
Why and when you should mark the request parameter as implicit:
Some methods that you will make use of in the body of your action have an implicit parameter list like, for example, Form.scala defines a method:
def bindFromRequest()(implicit request: play.api.mvc.Request[_]): Form[T] = { ... }
You don't necessarily notice this as you would just call myForm.bindFromRequest() You don't have to provide the implicit arguments explicitly. No, you leave the compiler to look for any valid candidate object to pass in every time it comes across a method call that requires an instance of the request. Since you do have a request available, all you need to do is to mark it as implicit.
You explicitly mark it as available for implicit use.
You hint the compiler that it's "OK" to use the request object sent in by the Play framework (that we gave the name "request" but could have used just "r" or "req") wherever required, "on the sly".
myForm.bindFromRequest()
see it? it's not there, but it is there!
It just happens without your having to slot it in manually in every place it's needed (but you can pass it explicitly, if you so wish, no matter if it's marked implicit or not):
myForm.bindFromRequest()(request)
Without marking it as implicit, you would have to do the above. Marking it as implicit you don't have to.
When should you mark the request as implicit? You only really need to if you are making use of methods that declare an implicit parameter list expecting an instance of the Request. But to keep it simple, you could just get into the habit of marking the request implicit always. That way you can just write beautiful terse code.
Also, in the above case there should be only one implicit function whose type is double => Int. Otherwise, the compiler gets confused and won't compile properly.
//this won't compile
implicit def doubleToInt(d: Double) = d.toInt
implicit def doubleToIntSecond(d: Double) = d.toInt
val x: Int = 42.0
I had the exact same question as you had and I think I should share how I started to understand it by a few really simple examples (note that it only covers the common use cases).
There are two common use cases in Scala using implicit.
Using it on a variable
Using it on a function
Examples are as follows
Using it on a variable. As you can see, if the implicit keyword is used in the last parameter list, then the closest variable will be used.
// Here I define a class and initiated an instance of this class
case class Person(val name: String)
val charles: Person = Person("Charles")
// Here I define a function
def greeting(words: String)(implicit person: Person) = person match {
case Person(name: String) if name != "" => s"$name, $words"
case _ => "$words"
}
greeting("Good morning") // Charles, Good moring
val charles: Person = Person("")
greeting("Good morning") // Good moring
Using it on a function. As you can see, if the implicit is used on the function, then the closest type conversion method will be used.
val num = 10 // num: Int (of course)
// Here I define a implicit function
implicit def intToString(num: Int) = s"$num -- I am a String now!"
val num = 10 // num: Int (of course). Nothing happens yet.. Compiler believes you want 10 to be an Int
// Util...
val num: String = 10 // Compiler trust you first, and it thinks you have `implicitly` told it that you had a way to covert the type from Int to String, which the function `intToString` can do!
// So num is now actually "10 -- I am a String now!"
// console will print this -> val num: String = 10 -- I am a String now!
Hope this can help.
A very basic example of Implicits in scala.
Implicit parameters:
val value = 10
implicit val multiplier = 3
def multiply(implicit by: Int) = value * by
val result = multiply // implicit parameter wiil be passed here
println(result) // It will print 30 as a result
Note: Here multiplier will be implicitly passed into the function multiply. Missing parameters to the function call are looked up by type in the current scope meaning that code will not compile if there is no implicit variable of type Int in the scope.
Implicit conversions:
implicit def convert(a: Double): Int = a.toInt
val res = multiply(2.0) // Type conversions with implicit functions
println(res) // It will print 20 as a result
Note: When we call multiply function passing a double value, the compiler will try to find the conversion implicit function in the current scope, which converts Int to Double (As function multiply accept Int parameter). If there is no implicit convert function then the compiler will not compile the code.