Scala's syntax of of work with anonymous objects in map() & flatMap()

Question

As example, I have the following Scala class(and the same in C#, ofc):

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

class Flight{
  var FlDate:DateTime = new DateTime
  var Origin = ""
  var Destination = ""
}

flights = Flight[]

Original C# Code:

var rez = flights.SelectMany(p => new[] {
     new { Airport = p.Origin, IsOrigin = true },
     new { Airport = p.Destination, IsOrigin = false }
   })
   .GroupBy(x => x.Airport)
   .Select(g => new
   {
       Airport = g.Key,
       LeftCount = g.Count(x => x.IsOrigin),
       ArrivedCount = g.Count(x => !x.IsOrigin)
   });

Incorrect Scala code:

var rez = flights
  .flatMap(case (airport, isOrigin) => {
  new { airport = _.origin, isOrigin = true }, // issue - ?
  new { airport = _.destination, isOrigin = false } // issue - ?
})
.groupBy(_.airport)
.map {
  case (airport, leftCount, arrivedCount) => new {
    val airport = ??? // issue - g.Key?
    val leftCount = ??? // issue - g.Count(x => x.IsOrigin) ?
    val arrivedCount = ??? // issue - g.Count(x => !x.IsOrigin) ?
  }
}

So, as you see by the code, as result i need to have an array of objects with the following properties: ListBuffer[airport:String, leftCount :Int, arrivedCount:Int]

So...

Can someone help to fix the code?
Maybe, someone knows some good article about "syntax of work with anonymous objects in Scala" topic?

Also please always at least mention the errors you got. Saying "issue" is not very informative. — Jasper-M, Apr 04 '17 at 18:26

Jean Logeart · Accepted Answer · 2017-04-04T17:05:57.340

1

You can do:

val flights: Seq[Flight] = ...
// Let's represent the result in a Map[Airport, (Left, Arrived)]
val res: Map[String, (Int, Int)] = 
  flights
    .flatMap(flight => Seq((flight.origin, true), (flight.destination, false)))  // Seq[(String, Boolean)]
    .groupBy { case (airport, _) => airport }  // Map[String, List[(String, Boolean)]] 
    .mapValues { grouped =>
      val left = grouped.count { case (_, isOrigin) => isOrigin }
      val arrived = grouped.count { case (_, isOrigin) => !isOrigin }
      (left, arrived)
    }

edited Apr 04 '17 at 17:05

answered Apr 04 '17 at 14:58

Jean Logeart

52,687
11
83
118

1. can you explain flatMap string logic? as I still do not understand how to write the same + 2. There is incorrectly grouping. I think it's because of this code return ListBuffer[(String, (Int, Int))] -- (I had fixed mistakes in your code) -- but must return ListBuffer[String, Int, Int] – Andrew_STOP_RU_WAR_IN_UA Apr 04 '17 at 15:21
expected and actual result you can find here: http://prntscr.com/esebss – Andrew_STOP_RU_WAR_IN_UA Apr 04 '17 at 16:43
Try again with the latest edit. I just tried this exact code and it works :) – Jean Logeart Apr 04 '17 at 17:06
Thanks a lot. But, can you explain flatMap string logic? Or to give some link with explanations how to write the same? – Andrew_STOP_RU_WAR_IN_UA Apr 04 '17 at 17:41
@Andrew `flatMap` and `SelectMany` are the same thing. – Jasper-M Apr 04 '17 at 18:38

Scala's syntax of of work with anonymous objects in map() & flatMap()

1 Answers1