// Resit Exam about data-mining a chat log

//=========================================

//

// This 

//

object Resit {

import io.Source

import scala.util._

// (1) The function below takes file name as argument.  It should read

//     the corresponding file and return its content. The content

//     should be returned as a list of strings, a string for each line

//     in the file. Since the file is a csv-file, the first line

//     should be dropped. Lines are separated by "\n".

def get_csv(name: String) : List[String] = {

  val csv = Source.fromFile(name)("ISO-8859-1")

  csv.mkString.split("\n").toList.drop(1)

}

// test cases:

//

// get_csv("log.csv")

// get_csv("log.csv").length  // should be 680

// (2) The function below takes a single line from the csv-file (as generated by

//     get_csv) and creates a Rec(ord) data structure. The data from the csv-file 

//     should be copied as follows:

//

//     csv-file         Rec data structure

//     -----------------------------------

//      counter      => num

//      id           => msg_id 

//      time_date    => date

//      name         => author

//      country,     => country (should be None if no country is given)

//      parent_id    => reply_id (should be None if there is no parent)

//      msg          => msg

//                   => parent is set to None  (will be calculated later)

//                   => children is set to Nil (will be calculated later)

//

//     You should use the function line.split(",").toList to separate

//     the items in the csv-line. BE CAREFUL that the message texts in 

//     the last field contain commas and therefore the split will not

//     always result into a list of 7 elements. You need to concatenate

//     anything beyond the 7th field into a string for the field msg.

case class Rec(num: Int, 

               msg_id: String,

               date: String,

               msg: String,

               author: String,

               country: Option[String],

               reply_id : Option[String],

               parent: Option[Int] = None,

               children: List[Int] = Nil)  

def process_line(line: String) : Rec = {

  val strs = line.split(",").toList

  Rec(num = strs(0).toInt,

      msg_id = strs(1),

      date = strs(2),

      author = strs(3),

      country = if (strs(4) == "") None else Some(strs(4)),

      reply_id = if (strs(5) == "") None else Some(strs(5)),

      msg = (for (i <- 6 until strs.length) yield strs(i)).mkString(","))

}

// test cases:

//

// process_line("""0,5ebeb459ac278d01301f1497,2020-05-15T15:25:13.413000,participant34,United Kingdom,5ebea6424923321d63155796,this question please?""")

// 

//      ==>  Rec(0,

//               "5ebeb459ac278d01301f1497",

//               "2020-05-15T15:25:13.413000",

//               "this question please?",

//               "participant34",

//               Some("United Kingdom"),

//               Some("5ebea6424923321d63155796"),

//               None,

//               List())

//

// process_line("""1,hash,date,p43,,,foo, bar""")

//

//      ==>  Rec(1, "hash", "date", "foo, bar",

//               "p43", None, None, None, List())

//

// (Note that in the last test case the message needs to be "foo, bar")

// (3) Each record in the log contains a unique hash code

//     identifying each message. Some messages also contain a hash

//     code identifying the parent message (to which question they reply).

//     The function post_process fills in the information about

//     potential children and a parent message. 

//  

//     The auxiliary function get_children takes a record e and a

//     record list rs as arguments, and returns the list of all direct

//     children (which have the hash code of e as reply_id. The list

//     of children are returned as a list of nums.

//      

//     The auxiliary function get_parent returns the number of the

//     record corresponding to the reply_id (if there exists one,

//     otherwise returns None).

def get_children(e: Rec, rs: List[Rec]) : List[Int] = {

  (rs.filter(r => r != e &&

                  Some(e.msg_id) == r.reply_id)).map(_.num)     

}

def get_parent(e: Rec, rs: List[Rec]) : Option[Int] = {

  (rs.find(r => r != e &&

                Some(r.msg_id) == e.reply_id)).map(_.num)     

}

def post_process(rs: List[Rec]) : List[Rec] =

  rs.map(r => r.copy(parent = get_parent(r, rs),

		     children = get_children(r, rs)))

// test cases:

//

//val recs = get_csv("log.csv").map(process_line)

//

//post_process(recs)(4).children   // List(12)

//post_process(recs)(23).children  // List(16,26)

//

//post_process(recs)(8).parent     // None

//post_process(recs)(9).parent     // Some(7)

//post_process(recs)(16).parent    // Some(23)

//post_process(recs)(26).parent    // Some(23)

// (4) The next two functions calculate the countries where

//     message authors are coming from and how many authors

//     come from each country (returned as a Map from countries

//     to numbers). In case an author did not specify a country,

//     the empty string is returned.

def get_countries(rs: List[Rec]) : Set[String] =

  rs.map(_.country.getOrElse("")).toSet

def get_countries_numbers(rs: List[Rec]) :  Map[String, Int] = {

  val name_countries = rs.map(r => (r.author, r.country.getOrElse(""))).distinct

  name_countries.groupBy(_._2).view.mapValues(_.size).toMap

}

// test cases:

//

//val recs = get_csv("log.csv").map(process_line)

//

// get_countries(recs) => 

//

//    Set("", Poland, Lebanon, Trinidad and

//        Tobago, Japan, Spain, Nigeria, Peru, India, Lithuania, Hong

//        Kong, Italy, Ireland, Uganda, England, Bangladesh, China,

//        Romania, Slovakia, United Kingdom, Norway, Pakistan, Indonesia,

//        Malaysia, Turkey, Portugal, Hungary)

//

// get_countries(recs) => 

//

//    Map("" -> 4, Poland -> 2, Lebanon -> 1, Trinidad and Tobago -> 2,

//        Japan -> 1, Spain -> 1, Nigeria -> 1, Peru -> 1, India -> 6,

//        Lithuania -> 1, Hong Kong -> 3, Italy -> 1, Ireland -> 2, Uganda -> 1,

//        England -> 4, Bangladesh -> 2, China -> 4, Romania -> 3,

//        Slovakia -> 1, United Kingdom -> 34, Norway -> 1, Pakistan -> 2,

//        Indonesia -> 2, Malaysia -> 1, Turkey -> 1, Portugal -> 1,

//        Hungary -> 2)

// (5) The function thread_sizes orders the message threads according to 

//     how many answers were given for one message (that is how many children, 

//     grand-children and so on one message received).

//

//     The auxiliary function search enumerates all children, grand-children and

//     so on for a given record r. Search returns the children and so on as

//     a list of Recs. 

//

//     The function thread_sizes generates for every message record

//     a pair

def search(r: Rec, rs: List[Rec]) : List[Rec] = 

  r :: (r.children.map(c => search(rs(c), rs)).flatten)

def thread_size(r: Rec, rs: List[Rec]) : (Int, Int) =

  (r.num, search(r, rs).size)

def ordered_thread_sizes(rs: List[Rec]) : List[(Int, Int)] = 

  rs.map(thread_size(_, rs)).sortBy(_._2).reverse

// test cases: 

//

//val recs_p = post_process(get_csv("log.csv").map(process_line))

//

//search(recs_p(459), recs_p).map(_.num)

//    => List(459, 401, 404, 426, 428, 399, 377, 357, 325, 304)

//

//thread_size(recs_p(459), recs_p) 

//    => (459,10)

//

//ordered_thread_sizes(recs_p).take(4)

//    => List((402,18), (95,12), (488,11), (459,10))

}