progs/lecture2.scala
author Christian Urban <christian.urban@kcl.ac.uk>
Sun, 22 Nov 2020 03:45:22 +0000
changeset 364 f1a6fa599d26
parent 363 e5c1d69cffa4
child 365 fc118ee0fce4
permissions -rw-r--r--
updated

// Scala Lecture 2
//=================
 

// String Interpolations
//=======================


def cube(n: Int) : Int = n * n * n

val n = 3
println("The cube of " + n + " is " + cube(n) + ".")

println(s"The cube of $n is ${cube(n)}.")

// or even

println(s"The cube of $n is ${n * n * n}.")

// helpful for debugging purposes
//
//     "The most effective debugging tool is still careful 
//          thought, coupled with judiciously placed print 
//                                             statements."
//       — Brian W. Kernighan, in Unix for Beginners (1979)


def gcd_db(a: Int, b: Int) : Int = {
  println(s"Function called with $a and $b.")
  if (b == 0) a else gcd_db(b, a % b)
}

gcd_db(48, 18)



// you can also implement your own string interpolations

import scala.language.implicitConversions
import scala.language.reflectiveCalls

implicit def sring_inters(sc: StringContext) = new {
    def i(args: Any*): String = s"\t${sc.s(args:_*)}\n"
    def l(args: Any*): String = s"${sc.s(args:_*)}:\n"
}

// this allows you to write things like

i"add ${3+2}" 
l"some_fresh_name"



// The Option Type
//=================

// in Java, if something unusually happens, you return null 
// or raise an exception
//
//in Scala you use Options instead
//   - if the value is present, you use Some(value)
//   - if no value is present, you use None


List(7,2,3,4,5,6).find(_ < 4)
List(5,6,7,8,9).find(_ < 4)

// Int:      ..., 0, 1, 2,...
// Boolean:  true false
//
// List[Int]: Nil, List(_) 
//
// Option[Int]: None, Some(0), Some(1), ...
// Option[...]: None, Some(_)

def safe_div(x: Int, y: Int) : Option[Int] = 
  if (y == 0) None else Some(x / y)

List(1,2,3,4,5,6).indexOf(7)

def my_min(ls: List[Int]) : Option[Int] = ls.minOption

my_min(List(1,2,3,4))


// better error handling with Options (no exceptions)
//
//  Try(something).getOrElse(what_to_do_in_case_of_an_exception)
//

import scala.util._
import io.Source

val my_url = "https://nms.kcl.ac.uk/christian.urban2/"

Source.fromURL(my_url)("ISO-8859-1").mkString

Try(Source.fromURL(my_url)("ISO-8859-1").mkString).getOrElse("")

Try(Some(Source.fromURL(my_url)("ISO-8859-1").mkString)).getOrElse(None)


// the same for files
Try(Some(Source.fromFile("test.txt")("ISO-8859-1").mkString)).getOrElse(None)


// how to implement a function for reading 
// (lines) something from files...
//
def get_contents(name: String) : List[String] = 
  Source.fromFile(name)("ISO-8859-1").getLines.toList

get_contents("text.txt")
get_contents("test.txt")

// slightly better - return Nil
def get_contents(name: String) : List[String] = 
  Try(Source.fromFile(name)("ISO-8859-1").getLines.toList).getOrElse(List())

get_contents("text.txt")

// much better - you record in the type that things can go wrong 
def get_contents(name: String) : Option[List[String]] = 
  Try(Some(Source.fromFile(name)("ISO-8859-1").getLines.toList)).getOrElse(None)

get_contents("text.txt")
get_contents("test.txt")


// operations on options

val lst = List(None, Some(1), Some(2), None, Some(3))

lst.flatten

Some(1).get
None.get

Some(1).isDefined
None.isDefined

for (x <- lst) yield x.getOrElse(0)



val ps = List((3, 0), (4, 2), (6, 2), 
              (2, 0), (1, 0), (1, 1))

// division where possible

for ((x, y) <- ps) yield {
  if (y == 0) None else Some(x / y)
}



// getOrElse is for setting a default value

val lst = List(None, Some(1), Some(2), None, Some(3))




// a function that turns strings into numbers 
// (similar to .toInt)
Integer.parseInt("1234")


def get_me_an_int(s: String) : Option[Int] = 
 Try(Some(Integer.parseInt(s))).getOrElse(None)


// This may not look any better than working with null in Java, but to
// see the value, you have to put yourself in the shoes of the
// consumer of the get_me_an_int function, and imagine you didn't
// write that function.
//
// In Java, if you didn't write this function, you'd have to depend on
// the Javadoc of the get_me_an_int. If you didn't look at the Javadoc, 
// you might not know that get_me_an_int could return null, and your 
// code could potentially throw a NullPointerException.


// even Scala is not immune to problems like this:

List(5,6,7,8,9).indexOf(7)
List(5,6,7,8,9).indexOf(10)
List(5,6,7,8,9)(-1)


Try({
  val x = 3
  val y = 0
  Some(x / y)
}).getOrElse(None)


// minOption 
// maxOption 
// minByOption 
// maxByOption

// Higher-Order Functions
//========================

// functions can take functions as arguments
// and produce functions as result

def even(x: Int) : Boolean = x % 2 == 0
def odd(x: Int) : Boolean = x % 2 == 1

val lst = (1 to 10).toList

lst.filter(even)
lst.count(odd)
lst.find(even)
lst.exists(even)

lst.find(_ < 4)
lst.filter(_ < 4) 

def less4(x: Int) : Boolean = x < 4
lst.find(less4)
lst.find(_ < 4)

lst.filter(x => x % 2 == 0)
lst.filter(_ % 2 == 0)


lst.sortWith((x, y) => x < y)
lst.sortWith(_ > _)

// but this only works when the arguments are clear, but 
// not with multiple occurences
lst.find(n => odd(n) && n > 2)



val ps = List((3, 0), (3, 2), (4, 2), (2, 2), 
              (2, 0), (1, 1), (1, 0))

def lex(x: (Int, Int), y: (Int, Int)) : Boolean = 
  if (x._1 == y._1) x._2 < y._2 else x._1 < y._1

ps.sortWith(lex)

ps.sortBy(x => x._1)
ps.sortBy(_._2)

ps.maxBy(_._1)
ps.maxBy(_._2)


// maps (lower-case)
//===================

def double(x: Int): Int = x + x
def square(x: Int): Int = x * x

val lst = (1 to 10).toList

lst.map(square)
lst.map(x => (double(x), square(x)))

// works also for strings
def tweet(c: Char) = c.toUpper

"Hello World".map(tweet)


// this can be iterated

lst.map(square).filter(_ > 4)

lst.map(square).find(_ > 4)
lst.map(square).find(_ > 4).map(double)

lst.map(square)
   .find(_ > 4)
   .map(double)


// Option Type and maps
//======================

// a function that turns strings into numbers (similar to .toInt)
Integer.parseInt("12u34")

// maps on Options

import scala.util._

def get_me_an_int(s: String) : Option[Int] = 
 Try(Some(Integer.parseInt(s))).getOrElse(None)

get_me_an_int("12345").map(_ % 2 == 0)
get_me_an_int("12u34").map(_ % 2 == 0)



val lst = List("12345", "foo", "5432", "bar", "x21", "456")
for (x <- lst) yield get_me_an_int(x)

// summing up all the numbers

lst.map(get_me_an_int).flatten.sum




// this is actually how for-comprehensions are
// defined in Scala

lst.map(n => square(n))
for (n <- lst) yield square(n)

// lets define our own higher-order functions
// type of functions is for example Int => Int


def my_map_int(lst: List[Int], f: Int => Int) : List[Int] = 
{
  if (lst == Nil) Nil
  else f(lst.head) :: my_map_int(lst.tail, f)
}

my_map_int(lst, square)

// same function using pattern matching: a kind
// of switch statement on steroids (see more later on)

def my_map_int(lst: List[Int], f: Int => Int) : List[Int] = 
  lst match {
    case Nil => Nil
    case x::xs => f(x)::my_map_int(xs, f)
  }



val biglst = (1 to 10000).toList
my_map_int(biglst, double)

(1 to 10000000).toList.map(double)

// other function types
//
// f1: (Int, Int) => Int
// f2: List[String] => Option[Int]
// ... 





// Map type (upper-case)
//=======================

// Note the difference between map and Map

val m = Map(1 -> "one", 2 -> "two", 10 -> "many")

List((1, "one"), (2, "two"), (10, "many")).toMap

m.get(1)
m.get(4)

m.getOrElse(1, "")
m.getOrElse(4, "")

val new_m = m + (10 -> "ten")

new_m.get(10)

val m2 = for ((k, v) <- m) yield (k, v.toUpperCase)



// groupBy function on Maps
val lst = List("one", "two", "three", "four", "five")
lst.groupBy(_.head)

lst.groupBy(_.length)

lst.groupBy(_.length).get(3)

val grps = lst.groupBy(_.length)
grps.keySet




// Pattern Matching
//==================

// A powerful tool which is supposed to come to Java in a few years
// time (https://www.youtube.com/watch?v=oGll155-vuQ)...Scala already
// has it for many years ;o)

// The general schema:
//
//    expression match {
//       case pattern1 => expression1
//       case pattern2 => expression2
//       ...
//       case patternN => expressionN
//    }


// recall
val lst = List(None, Some(1), Some(2), None, Some(3)).flatten

def my_flatten(xs: List[Option[Int]]): List[Int] = 
xs match {
  case Nil => Nil 
  case None::rest => my_flatten(rest)
  case Some(v)::rest => v :: my_flatten(rest)
}

my_flatten(List(None, Some(1), Some(2), None, Some(3)))


// another example with a default case
def get_me_a_string(n: Int): String = n match {
  case 0 | 1 | 2 => "small"
}

get_me_a_string(3)


// you can also have cases combined
def season(month: String) : String = month match {
  case "March" | "April" | "May" => "It's spring"
  case "June" | "July" | "August" => "It's summer"
  case "September" | "October" | "November" => "It's autumn"
  case "December" => "It's winter"
  case "January" | "February" => "It's unfortunately winter"
}
 
println(season("November"))

// What happens if no case matches?
println(season("foobar"))


// days of some months
def days(month: String) : Int = month match {
  case "March" | "April" | "May" => 31
  case "June" | "July" | "August" => 30
}


// Silly: fizz buzz
def fizz_buzz(n: Int) : String = (n % 3, n % 5) match {
  case (0, 0) => "fizz buzz"
  case (0, _) => "fizz"
  case (_, 0) => "buzz"
  case _ => n.toString  
}

for (n <- 0 to 20) 
 println(fizz_buzz(n))




// Recursion
//===========

// well-known example

def fib(n: Int) : Int = { 
  if (n == 0 || n == 1) 1
   else fib(n - 1) + fib(n - 2)
}


/* Say you have characters a, b, c.
   What are all the combinations of a certain length?

   All combinations of length 2:
  
     aa, ab, ac, ba, bb, bc, ca, cb, cc

   Combinations of length 3:
   
     aaa, baa, caa, and so on......
*/

def combs(cs: List[Char], n: Int) : List[String] = {
  if (n == 0) List("")
  else for (c <- cs; s <- combs(cs, n - 1)) yield s"$c$s"
}

combs(List('a', 'b', 'c'), 3)



def combs(cs: List[Char], l: Int) : List[String] = {
  if (l == 0) List("")
  else for (c <- cs; s <- combs(cs, l - 1)) yield s"$c$s"
}

combs("abc".toList, 2)


// When writing recursive functions you have to
// think about three points
// 
// - How to start with a recursive function
// - How to communicate between recursive calls
// - Exit conditions



// A Recursive Web Crawler / Email Harvester
//===========================================
//
// the idea is to look for links using the
// regular expression "https?://[^"]*" and for
// email addresses using another regex.

import io.Source
import scala.util._

// gets the first 10K of a web-page
def get_page(url: String) : String = {
  Try(Source.fromURL(url)("ISO-8859-1").take(10000).mkString).
    getOrElse { println(s"  Problem with: $url"); ""}
}

// regex for URLs and emails
val http_pattern = """"https?://[^"]*"""".r
val email_pattern = """([a-z0-9_\.-]+)@([\da-z\.-]+)\.([a-z\.]{2,6})""".r

//test case:
//email_pattern.findAllIn
//  ("foo bla christian@kcl.ac.uk 1234567").toList


// drops the first and last character from a string
def unquote(s: String) = s.drop(1).dropRight(1)

def get_all_URLs(page: String): Set[String] = 
  http_pattern.findAllIn(page).map(unquote).toSet

// naive version of crawl - searches until a given depth,
// visits pages potentially more than once
def crawl(url: String, n: Int) : Unit = {
  if (n == 0) ()
  else {
    println(s"  Visiting: $n $url")
    for (u <- get_all_URLs(get_page(url))) crawl(u, n - 1)
  }
}

// some starting URLs for the crawler
val startURL = """https://nms.kcl.ac.uk/christian.urban/"""

crawl(startURL, 2)


// a primitive email harvester
def emails(url: String, n: Int) : Set[String] = {
  if (n == 0) Set()
  else {
    println(s"  Visiting: $n $url")
    val page = get_page(url)
    val new_emails = email_pattern.findAllIn(page).toSet
    new_emails ++ (for (u <- get_all_URLs(page)) yield emails(u, n - 1)).flatten
  }
}

emails(startURL, 3)


// if we want to explore the internet "deeper", then we
// first have to parallelise the request of webpages:
//
// scala -cp scala-parallel-collections_2.13-0.2.0.jar 
// import scala.collection.parallel.CollectionConverters._





// Jumping Towers
//================


def moves(xs: List[Int], n: Int) : List[List[Int]] = 
 (xs, n) match {
   case (Nil, _) => Nil
   case (xs, 0) => Nil
   case (x::xs, n) => (x::xs) :: moves(xs, n - 1)
 }


moves(List(5,1,0), 1)
moves(List(5,1,0), 2)
moves(List(5,1,0), 5)

// checks whether a jump tour exists at all

def search(xs: List[Int]) : Boolean = xs match {
  case Nil => true
  case (x::xs) =>
    if (xs.length < x) true else moves(xs, x).exists(search(_))
}


search(List(5,3,2,5,1,1))
search(List(3,5,1,0,0,0,1))
search(List(3,5,1,0,0,0,0,1))
search(List(3,5,1,0,0,0,1,1))
search(List(3,5,1))
search(List(5,1,1))
search(Nil)
search(List(1))
search(List(5,1,1))
search(List(3,5,1,0,0,0,0,0,0,0,0,1))

// generate *all* jump tours
//    if we are only interested in the shortes one, we could
//    shortcircut the calculation and only return List(x) in
//    case where xs.length < x, because no tour can be shorter
//    than 1
// 

def jumps(xs: List[Int]) : List[List[Int]] = xs match {
  case Nil => Nil
  case (x::xs) => {
    val children = moves(xs, x)
    val results = children.map(cs => jumps(cs).map(x :: _)).flatten
    if (xs.length < x) List(x) :: results else results
  }
}

jumps(List(3,5,1,2,1,2,1))
jumps(List(3,5,1,2,3,4,1))
jumps(List(3,5,1,0,0,0,1))
jumps(List(3,5,1))
jumps(List(5,1,1))
jumps(Nil)
jumps(List(1))
jumps(List(5,1,2))
moves(List(1,2), 5)
jumps(List(1,5,1,2))
jumps(List(3,5,1,0,0,0,0,0,0,0,0,1))

jumps(List(5,3,2,5,1,1)).minBy(_.length)
jumps(List(1,3,5,8,9,2,6,7,6,8,9)).minBy(_.length)
jumps(List(1,3,6,1,0,9)).minBy(_.length)
jumps(List(2,3,1,1,2,4,2,0,1,1)).minBy(_.length)






/*
 *               1
 *             / |  \
 *           /   |   \
 *         /     |    \
 *        2      3     8
 *      /  \    / \   / \
 *     4    5  6   7 9  10
 * Preorder: 1,2,4,5,3,6,7,8,9,10
 * InOrder: 4,2,5,1,6,3,7,9,8,10
 * PostOrder: 4,5,2,6,7,3,9,10,8,1
 *
 
show inorder, preorder, postorder



*/