301. Remove Invalid Parentheses
# O(n * 2 ^ n) time || O(2 ^ n) space
def remove_invalid_parentheses(self, s: str) -> List[str]:
def is_valid(string):
cnt = 0
for c in string:
if c == '(':
cnt += 1
elif c == ')':
cnt -= 1
if cnt < 0:
return False
return cnt == 0
res = []
dq, visited, found = collections.deque([s]), {str}, False
while dq:
curr_str = dq.popleft()
if curr_str not in visited:
visited.add(curr_str)
if is_valid(curr_str):
found = True
res.append(curr_str)
if not found:
for i, c in enumerate(curr_str):
if c in "()":
dq.append(curr_str[:i] + curr_str[i + 1:])
return res
The main idea is to use Breadth-First Search (BFS) to explore all possible strings formed by removing 0 or more parentheses. We start with the original string and, at each BFS level, we generate all possible strings by removing a single parenthesis. The key insight is that the first valid strings we encounter during our BFS traversal are the ones formed by removing the minimum number of invalid parentheses.
Detailed Steps:
- Initialization: Use a queue for BFS. Initially, the queue contains only the original string. Use a set (seen) to keep track of strings that have been visited to avoid re-processing. Initialize a found flag to False. This flag helps us stop BFS as soon as we find valid strings.
- BFS:
- While the queue is not empty:
- For each string in the current BFS level:
- Check if it’s valid.
- If it’s valid, mark found as True and add it to the result list.
- If no valid string is found at the current level, generate all possible strings by removing a single parenthesis. Add these to the queue if they haven’t been seen before.
- For each string in the current BFS level:
- While the queue is not empty:
- Validity Check: To determine if a string is valid, we use a counter. Increment the counter for every open parenthesis and decrement for every close parenthesis. The string is valid if the counter never goes negative and ends up being zero.
More details
newStr = currentStr[:j] + currentStr[j + 1:]
We’re creating a new string (newStr) by removing the character at index j from the current string (currentStr).
Let’s break it down:
currentStr[:j]
: This slices the string from the beginning up to, but not including, index j. It takes all characters before the j-th character.currentStr[j+1:]
: This slices the string from the position after j till the end. So, it takes all characters after the j-th character. By concatenating these two slices together, we essentially remove the character at index j.
Example:
Suppose currentStr = “abcde” and j = 2.
currentStr[:j]
would be “ab”currentStr[j+1:]
would be “de” When you concatenate them: “ab” + “de” = “abde”.
As you can see, the character at index 2 (which is “c”) has been removed from the original string.