Longest Substring Without Repeating Characters

Solution

Let $n$ be size of the string, $m$ the size of the charset/alphabet, and $k$ the size of the hashset, which is upper bounded by $n$ and $m$.

Brute Force

The brute force solution is not shown.

Complexity

time: $O(n^3)$ (a nested for loop for the sliding window and to check if unique takes $O(n)$ time)
space: $O(\min(m,n))$ ($O(k)$ space for the sliding window)

Sliding Window Using HashSet

The idea is to use a sliding window to locate a substring, and a hashset to see if the new char is already seen previously.

Sliding window logic: i is the left boundary and j the right boundary. Increase j by 1 if s[j] has not occurred in the current subarray. Increase i if s[j] has occurred.

Invariant: i<=j

Complexity

time: $O(n)$ (worst case: $O(2n)$, all characters are the same and each will be visited by both $i$ and $j$)
space: $O(\min(m,n))$

def lengthOfLongestSubstring(self, s: str) -> int:
    if not s: return 0
    n = len(s)
    hashset = set()
    i = j = 0
    res = 1
    while i < n and j < n:
        if s[j] not in hashset:
            hashset.add(s[j])
            j += 1
            res = max(res, j-i)
        else:
            hashset.remove(s[i])
            i += 1
    return res

Sliding Window Using HashMap

If having seen the new char in existing window, could update the left boundary to d[val]+1 ($1$ + index of last duplicate in the window).

Several optimizations:

1. start and idx rather than i and j to refer to left and right boundaries of window
2. enumerate idx and val
3. dct to store the mapping from val to idx

Complexity

time: $O(n)$
space: $O(\min(m,n))$

def lengthOfLongestSubstring(self, s: str) -> int:
    res = 0
    start = 0
    dct = {}
    for idx, val in enumerate(s):
        if s[idx] in dct:
            start = max(start, dct[val]+1)
        res = max(res, idx-start+1)
        dct[val] = idx
    return res