Example #13: Explain the concepts
Can you explain the Scala concepts applied in the following code snippets?
|
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 |
package com.myapp object Test2 extends App { trait Wrapper[C[_]] { def aFunction[T,R](f: T => R): C[T] => C[R] } implicit object ListWrapper extends Wrapper[List] { //returns a function: so this is a higher order function override def aFunction[T,R](f: T => R): List[T] => List[R] = { l => println("taking T => R & returning List[T] => List[R]") l.map(f) } } // A function that returns an anonymous Object // Anonymous class Test2$$anon$1 AnyRef is like java.lang.Object private implicit def fnGenAnonymousObj[C[_], T](input: C[T]) = new { // or new AnyRef or new Object //takes a function as an arg: So, this is a higher order function def anotherFunction[R](f: T => R)(implicit wrap: Wrapper[C]): C[R] = { println("In here first.......") val aClosure: C[T] => C[R] = wrap.aFunction(f) // C[T] => C[R] val result: C[R] = aClosure(input) // pass the list to the closure result } } //val result = fnGenAnonymousObj(List(1, 2, 3)).anotherFunction {x => x + 1}(ListWrapper) // without implicits val result = List(1, 2, 3).anotherFunction {x => x + 1} println(result) //val result2 = fnGenAnonymousObj(List("Java", "Scala")).anotherFunction {x => x.length}(ListWrapper) // without implicits val result2 = List("Java", "Scala").anotherFunction {x => x.length} // with implicits println(result2) } |
Output:
|
1 2 3 4 5 6 7 8 |
In here first....... taking T => R & returning List[T] => List[R] List(2, 3, 4) In here first....... taking T => R & returning List[T] => List[R] List(4, 5) |
1. Higher Order Functions: “Line 12” is a higher order function as the function not only takes a function as an argument i.e. f: T => R, but also returns another function i.e. List[T] => List[R] via l => l.map(f).
“Line 23” is a higher order function as the function takes another function as an input argument i.e. T => R.
map, flatMap, filter, etc are higher order functions as they take functions as arguments.
2. Closures: “Line 12” is a closure as it closes over a “list”. The list is supplied in “Line 26” aClosure(input) //pass the list to the closure.
3. Generics: C[_], T, R, and C[T] are Scala generic type parameters. For example, C[T] can be substituted for List[Int] and can take List(1, 2, 3) & C[_] is a higher kinded type that can be substituted for List[List[Int]], List[Option[String]], List[List[List[String]]], etc.
4. Implicits: are used where the compiler rewrites
|
1 2 3 |
val result = List(1, 2, 3).anotherFunction {x => x + 1} |
TO:
|
1 2 3 |
val result = fnGenAnonymousObj(List(1, 2, 3)).anotherFunction {x => x + 1}(ListWrapper) |
The magic happens due to implicit invocations. The ” List(1, 2, 3) anotherFunction {x => x + 1}” invokes the function on “Line 23” inside the “AnyRef” object on “Line 20”, which takes “(input: C[T])” as an argument. “C[T]” can be substituted for the “List[Int]”. Similarly, “implicit wrap: Wrapper[C]” on line 23 invokes “aFunction” via ListWrapper on Line 10, which is of type Wrapper..
5. Partial functions: “Line 25” is a partial function.
How will you rewrite the above code without implicits?
|
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 |
package com.myapp object Test2 extends App { trait Wrapper[C[_]] { def aFunction[T,R](f: T => R): C[T] => C[R] } object ListWrapper extends Wrapper[List] { //returns a function: so this is a higher order function override def aFunction[T,R](f: T => R): List[T] => List[R] = { l => println("taking T => R & returning List[T] => List[R]") l.map(f) } } // A function that returns an anonymous Object // Anonymous class Test2$$anon$1 AnyRef is like java.lang.Object private def fnGenAnonymousObj[C[_], T](input: C[T]) = new { // or new AnyRef or new Object //takes a function as an arg: So, this is a higher order function def anotherFunction[R](f: T => R)(wrap: Wrapper[C]): C[R] = { println("In here first.......") val aClosure: C[T] => C[R] = wrap.aFunction(f) // C[T] => C[R] val result: C[R] = aClosure(input) // pass the list to the closure result } } val result = fnGenAnonymousObj(List(1, 2, 3)).anotherFunction {x => x + 1}(ListWrapper) //val result = List(1, 2, 3).anotherFunction {x => x + 1} // with implicits println(result) val result2 = fnGenAnonymousObj(List("Java", "Scala")).anotherFunction {x => x.length}(ListWrapper) //val result2 = List("Java", "Scala").anotherFunction {x => x.length} // with implicits println(result2) } |
Output:
|
1 2 3 4 5 6 7 8 9 |
In here first....... taking T => R & returning List[T] => List[R] List(2, 3, 4) In here first....... taking T => R & returning List[T] => List[R] List(4, 5) |
What is the benefit of the above code?
You could have just called a list class’s “map” function as in
|
1 2 3 |
List(1, 2, 3) map (x => x + 1) // same as List(1, 2, 3).map (x => x + 1) |
Firstly to demonstrate the Scala concepts with a simple example.
Secondly, the above code can be extended to be an abstraction (i.e. a functor) and handle other C[_] types like “Option[T]” as shown below:
|
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 |
package com.myapp object Test2 extends App { trait Wrapper[C[_]] { def aFunction[T,R](f: T => R): C[T] => C[R] } implicit object ListWrapper extends Wrapper[List] { //returns a function: so this is a higher order function override def aFunction[T,R](f: T => R): List[T] => List[R] = { l => println("taking T => R & returning List[T] => List[R]") l.map(f) } } implicit object OptionWrapper extends Wrapper[Option] { //returns a function: so this is a higher order function override def aFunction[T,R](f: T => R): Option[T] => Option[R] = { l => println("taking T => R & returning Option[T] => Option[R]") l.map(f) } } // A function that returns an anonymous Object // Anonymous class Test2$$anon$1 AnyRef is like java.lang.Object private implicit def fnGenAnonymousObj[C[_], T](input: C[T]) = new { // or new AnyRef or new Object //takes a function as an arg: So, this is a higher order function def anotherFunction[R](f: T => R)(implicit wrap: Wrapper[C]): C[R] = { println("In here first.......") val aClosure: C[T] => C[R] = wrap.aFunction(f) // C[T] => C[R] val result: C[R] = aClosure(input) // pass the list to the closure result } } val result = List(1, 2, 3) anotherFunction (x => x + 1) println(result); val result2 = Option(5) anotherFunction (_ + 1) anotherFunction (_ + 2) println(result2); } |
The Output: will be
|
1 2 3 4 5 6 7 8 9 10 |
In here first....... taking T => R & returning List[T] => List[R] List(2, 3, 4) In here first....... taking T => R & returning Option[T] => Option[R] In here first....... taking T => R & returning Option[T] => Option[R] Some(8) |
Q. What if you want to use Some(5) & None instead of “Option”?
A. You can define two functions that returns a type Option.
|
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 |
package com.myapp object Test2 extends App { trait Wrapper[C[_]] { def aFunction[T,R](f: T => R): C[T] => C[R] } implicit object ListWrapper extends Wrapper[List] { //returns a function: so this is a higher order function override def aFunction[T,R](f: T => R): List[T] => List[R] = { l => println("taking T => R & returning List[T] => List[R]") l.map(f) } } implicit object OptionWrapper extends Wrapper[Option] { //returns a function: so this is a higher order function override def aFunction[T,R](f: T => R): Option[T] => Option[R] = { l => println("taking T => R & returning Option[T] => Option[R]") l.map(f) } } // A function that returns an anonymous Object // Anonymous class Test2$$anon$1 AnyRef is like java.lang.Object private implicit def fnGenAnonymousObj[C[_], T](input: C[T]) = new { // or new AnyRef or new Object //takes a function as an arg: So, this is a higher order function def anotherFunction[R](f: T => R)(implicit wrap: Wrapper[C]): C[R] = { println("In here first.......") val aClosure: C[T] => C[R] = wrap.aFunction(f) // C[T] => C[R] val result: C[R] = aClosure(input) // pass the list to the closure result } } val result2 = some(5) anotherFunction (_ + 1) anotherFunction (_ + 2) println(result2); val result3 = none[Int] anotherFunction (_ + 1) println(result3); def some[T](input: T): Option[T] = Some(input) def none[T]: Option[T] = None } |
Pay attention to: lines 45 & 47 where new functions were added to handle Some(x) and None respectively via functions. Lines 39 to 42 make use of these functions.
Output:
|
1 2 3 4 5 6 7 8 9 10 |
In here first....... taking T => R & returning Option[T] => Option[R] In here first....... taking T => R & returning Option[T] => Option[R] Some(8) In here first....... taking T => R & returning Option[T] => Option[R] None |
Popular posts:
Java 8 String streams and finding the first non repeated character with functional programming Q1.Find the first non repeated character in a given string input using Java 8 or later? A1.Extends Find the first… (2,218)
18 Java scenarios based interview Q&As for the experienced – Part 1 Let's look at scenarios or problem statements & how would you go about handling those scenarios in Java. These scenarios… (1,244)
40+ Apache Spark best practices & optimisation interview FAQs – Part-1 There are so many different ways to solve the big data problems at hand in Spark, but some approaches can… (734)
30+ Java Code Review Checklist Items This Java code review checklist is not only useful during code reviews, but also to answer an important Java job… (683)
01: 50+ SQL scenarios based interview Q&As on identifying & deleting duplicate records 50+ SQL interview questions and answers to solve real business scenarios. SQL is widely used in building microservices & Big… (497)
How to become a Java architect? In industry specific forums, I often see questions like “what certification do I need to do to become an architect?”… (453)
6 Key considerations in processing large files in Java Q1. What are the key considerations in processing large files? A1. Before jumping into coding, get the requirements. #1 Trade-offs… (386)
24: PySpark with Hierarchical Data on Databricks This tutorial extends Getting started with Databricks. Currently spark does not support recursion like you can use in SQL via… (367)
Membership Levels All memberships have access to the entire site. Only the duration is limited by the levels. It is important to… (349)
05: Finding the 2nd highest number in an array in Java Requirements gathering Does the array allow duplicates? If duplicates are allowed, then do you need to report duplicates? For example,… (334)
01: 13 Java low latency interview questions & answers Have you seen job advertisements requiring Java candidates to work in low latency, high throughput, real-time and distributed systems with… (296)
00: Top 50+ Core Java interview questions & answers for 1 to 3 years experience Top 50 core Java interview questions covering Java core concepts with diagrams, code, examples, and scenarios. If you don't get… (280)- Unit Testing Servlet filter with Mockito This assumes that you have gone through the basic JUnit & Mockito tutorials. Here is test class to "Unit Testing… (277)
- Loading a YAML file with Jackson into a Map & flatten it YAML files are often used application configuration. We have already looked at YAML interview Questions & Answers. There are multiple… (223)