The initial recruiter screen followed my technical screen. Their process felt strange and difficult to prepare for. The interviewer advised me not to worry about performance at all, stating there would be no O(n) analysis required.

The problem was as follows:

Given a string of store logs, where 'Y' represents customers in the store and 'N' represents no customers (e.g., "Y N N Y"), and a closing time (integer), determine the total penalty. A penalty of +1 is incurred for every hour the store is open with no customers. An additional penalty of +1 is incurred for every hour the store is closed when there could have been customers.

The approach was to divide the array into two halves based on the closing time and calculate penalties for each half. The second part of the problem built upon this, requiring the calculation of the ideal closing time. I computed the penalty for all possible closing times and identified the lowest one. I solved both parts, passed all their tests, and handled edge cases, but was ultimately rejected.

Honestly, I have no idea what else I could have done. It was a bad experience. The problem description was also very wordy, taking me at least five minutes just to read.

Find total penalty and ideal closing time.