Here is the interview question prompt, presented for reference.

Assume that we are building software to determine how many planes are in the sky.

The data that we have on planes currently in the air comes in the form of a grid, modeled by a matrix array. See below:

[
  ['.', '.', '.', 'P'],
  ['.', '.', '.', 'P'],
  ['P', 'P', '.', 'P'],
  ['.', '.', '.', 'P']
]

Within that matrix, the dots (.s) represent available airspace. The Ps that are neighboring each other model one plane each. So in the above example, there are exactly 2 planes.

Given a multi-dimensional array of .s and Ps exclusively, can you find the number of planes that there are in the sky? Planes may only be placed horizontally or vertically. A diagonal plane like this would not count as one plane. Instead, there will be 3 separate planes.

[
  ['.', '.', 'P'],
  ['.', 'P', '.'],
  ['P', '.', '.']
]

Could you accomplish this in one pass?

sky = [
  ['.', '.', '.', 'P'],
  ['.', '.', '.', 'P'],
  ['P', 'P', '.', 'P'],
  ['.', '.', '.', 'P']
]

numOfPlanes(sky);
// 2

Constraints

• Total number of elements in the matrix <= 100000
• The elements will consist only of . and P characters
• The matrix can be empty
• Expected time complexity : O(n * m) where n and m are number of rows and columns respectively
• Expected space complexity : O(1)

You can see the full challenge with visuals at this link.