LeetCode 819: Most Common Word

Meta Data

#

Difficulty: easy First Attempt: 2022-09-14

• Submissions: Wrong Answer ×2 → Runtime Error → Wrong Answer → Accepted ×2 → Wrong Answer → Accepted

Approach

#

Normalize first: replace non-letters with spaces (deleting them would glue a,b into ab), lowercase, and split() (String Template 4). Put banned in a set, skip it while counting (Hash Table Template 1), and take the most frequent.

My Attempts

#

This took 8 submissions, each missing one normalization detail:

• Attempt 1 used paragraph[:-1] to drop trailing punctuation, but there may be none: "a" became an empty string.
• Attempt 2 deleted punctuation: "a'a" became "aa", while the answer is "a".
• After switching to spaces, one version had a misspelled variable (NameError) and two forgot to lowercase: "Bob" returned "Bob" instead of "bob".

Solution

#

class Solution:
    def mostCommonWord(self, paragraph: str, banned: List[str]) -> str:
        words = "".join(c.lower() if c.isalpha() else " " for c in paragraph).split()
        ban = set(banned)
        cnt = Counter(w for w in words if w not in ban)
        return cnt.most_common(1)[0][0]

Complexity

#

Time O(n + b), space O(n + b).

Mistakes to Avoid

#

• Replace punctuation with spaces rather than deleting it.
• Matching is case-insensitive; lowercase everything.
• Do not drop the last character: the paragraph need not end with punctuation.

Interview Explanation

#

I turn non-letters into spaces, lowercase, split, skip banned words via a set while counting, and return the most frequent word. Time O(n + b).