Task

Mr.Right always tell the truth, Mr.Wrong always tell the lies.

Some people are having a conversation, trying to figure out how many in the group are Mr.Wrong. Listen to their conversation and determine how many Mr.Wrong there are.

[[Input]] A string array: conversation

They always talking about ... is Mr.Wrong., ... is Mr. Right, There are ... Mr.Wrong., There are ... Mr.Right.

Please note that everyone has at least one sentence and everyone is either Mr.Right or Mr.Wrong. Also, no one will ever have a paradoxical statement.

[[Output]] The number of Mr.Wrong in the group. If it is impossible to determine, return null.

Examples:

conversation=[
  "Joe:There are 1 Mr.Wrong.",
  "Bob:There are 2 Mr.Wrong.",
  "Tim:There are 3 Mr.Wrong."
]
howManyMrWrong(conversation) should return 2

conversation=[
  "Mario:Luigi is Mr.Wrong.",
  "Luigi:Mario is Mr.Wrong."
]
howManyMrWrong(conversation) should return 1

conversation=[
  "Musk:Zuckerberg is Mr.Right.",
  "Zuckerberg:There are 2 Mr.Wrong.",
  "Musk:Bezos is Mr.Right.",
  "Bezos:There are 1 Mr.Wrong."
]
howManyMrWrong(conversation) should return 3

conversation=[
  "John:Peter is Mr.Right.",
  "Peter:There are 2 Mr.Right",
  "Tom:John is Mr.Wrong."
]
howManyMrWrong(conversation) should return null

Two solutions are possible in the last example: either 1 Mr.Wrong (Tom) or 2 Mr. Wrong (John and Peter). In this case, the result is null.

Random Tests

$nn$ is the number of people.

• 50 tests with $n\le 150n\; \backslash le\; 150$ with one sentence of type There are ... Mr.Right/Wrong per person

• 50 tests with $n\le 150n\; \backslash le\; 150$ with one sentence of type ... is Mr.Right/Wrong per person

• 300 tests with $4\le n\le 104\; \backslash le\; n\; \backslash le\; 10$

• 100 tests with $300\le n\le 500300\; \backslash le\; n\; \backslash le\; 500$

Note for Beta:

The Random tests (including test generation) will likely be tweaked quite a lot before they are finished. I also need more good quality fixed tests, so if you run into specific kinds of scenarios in the random tests which aren't covered in the fixed tests, please suggest a fixed test in discourse.

With this in mind, please do not translate the kata until the tests have settled. After that I welcome translations :)

Series

• #1: Guess a number

• #2: Who is Mr.Wrong?

• #3: How many Mr.Wrong?