[LeetCode] 30. Substring with Concatenation of All Words 串联全部单词的子串

2021年09月15日 阅读数:1
这篇文章主要向大家介绍[LeetCode] 30. Substring with Concatenation of All Words 串联全部单词的子串,主要内容包括基础应用、实用技巧、原理机制等方面,希望对大家有所帮助。

You are given a string, s, and a list of words, words, that are all of the same length. Find all starting indices of substring(s) in s that is a concatenation of each word in words exactly once and without any intervening characters.html

Example 1:java

Input:
  s = "barfoothefoobarman",
  words = ["foo","bar"]
Output: [0,9]
Explanation: Substrings starting at index 0 and 9 are "barfoor" and "foobar" respectively.
The output order does not matter, returning [9,0] is fine too.

Example 2:python

Input:
  s = "wordgoodstudentgoodword",
  words = ["word","student"]
Output: []

给一个字符串和一单词列表,找出串联列表中全部单词的子串的起始位置。app

解法:也就是要求出串联全部单词的子串,而后返回起始位置的坐标。post

Java:url

class Solution {
    public List<Integer> findSubstring(String s, String[] words) {
        final Map<String, Integer> counts = new HashMap<>();
        for (final String word : words) {
            counts.put(word, counts.getOrDefault(word, 0) + 1);
        }
        final List<Integer> indexes = new ArrayList<>();
        final int n = s.length(), num = words.length, len = words[0].length();
        for (int i = 0; i < n - num * len + 1; i++) {
            final Map<String, Integer> seen = new HashMap<>();
            int j = 0;
            while (j < num) {
                final String word = s.substring(i + j * len, i + (j + 1) * len);
                if (counts.containsKey(word)) {
                    seen.put(word, seen.getOrDefault(word, 0) + 1);
                    if (seen.get(word) > counts.getOrDefault(word, 0)) {
                        break;
                    }
                } else {
                    break;
                }
                j++;
            }
            if (j == num) {
                indexes.add(i);
            }
        }
        return indexes;
    }
}

Python:code

class Solution(object):
    def findSubstring(self, s, words):
        """
        :type s: str
        :type words: List[str]
        :rtype: List[int]
        """
        result, m, n, k = [], len(s), len(words), len(words[0])
        if m < n*k:
            return result

        lookup = collections.defaultdict(int)
        for i in words:
            lookup[i] += 1                # Space: O(n * k)

        for i in xrange(k):               # Time:  O(k)
            left, count = i, 0
            tmp = collections.defaultdict(int)
            for j in xrange(i, m-k+1, k): # Time:  O(m / k)
                s1 = s[j:j+k];            # Time:  O(k)
                if s1 in lookup:
                    tmp[s1] += 1
                    if tmp[s1] <= lookup[s1]:
                        count += 1
                    else:
                        while tmp[s1] > lookup[s1]:
                            s2 = s[left:left+k]
                            tmp[s2] -= 1
                            if tmp[s2] < lookup[s2]:
                                count -= 1
                            left += k
                    if count == n:
                        result.append(left)
                        tmp[s[left:left+k]] -= 1
                        count -= 1
                        left += k
                else:
                    tmp = collections.defaultdict(int)
                    count = 0
                    left = j+k
        return resultv

Python:htm

# Time:  O(m * n * k), where m is string length, n is dictionary size, k is word length
# Space: O(n * k)
class Solution2(object):
    def findSubstring(self, s, words):
        """
        :type s: str
        :type words: List[str]
        :rtype: List[int]
        """
        result, m, n, k = [], len(s), len(words), len(words[0])
        if m < n*k:
            return result

        lookup = collections.defaultdict(int)
        for i in words:
            lookup[i] += 1                            # Space: O(n * k)

        for i in xrange(m+1-k*n):                     # Time: O(m)
            cur, j = collections.defaultdict(int), 0
            while j < n:                              # Time: O(n)
                word = s[i+j*k:i+j*k+k]               # Time: O(k)
                if word not in lookup:
                    break
                cur[word] += 1
                if cur[word] > lookup[word]:
                    break
                j += 1
            if j == n:
                result.append(i)

        return result

C++:blog

class Solution {
public:
    vector<int> findSubstring(string s, vector<string>& words) {
        vector<int> res;
        if (s.empty() || words.empty()) return res;
        int n = words.size(), m = words[0].size();
        unordered_map<string, int> m1;
        for (auto &a : words) ++m1[a];
        for (int i = 0; i <= (int)s.size() - n * m; ++i) {
            unordered_map<string, int> m2;
            int j = 0; 
            for (j = 0; j < n; ++j) {
                string t = s.substr(i + j * m, m);
                if (m1.find(t) == m1.end()) break;
                ++m2[t];
                if (m2[t] > m1[t]) break;
            }
            if (j == n) res.push_back(i);
        }
        return res;
    }
};

C++:  字符串

class Solution {
public:
    vector<int> findSubstring(string s, vector<string>& words) {
        if (s.empty() || words.empty()) return {};
        vector<int> res;
        int n = s.size(), cnt = words.size(), len = words[0].size();
        unordered_map<string, int> m1;
        for (string w : words) ++m1[w];
        for (int i = 0; i < len; ++i) {
            int left = i, count = 0;
            unordered_map<string, int> m2;
            for (int j = i; j <= n - len; j += len) {
                string t = s.substr(j, len);
                if (m1.count(t)) {
                    ++m2[t];
                    if (m2[t] <= m1[t]) {
                        ++count;
                    } else {
                        while (m2[t] > m1[t]) {
                            string t1 = s.substr(left, len);
                            --m2[t1];
                            if (m2[t1] < m1[t1]) --count;
                            left += len;
                        }
                    }
                    if (count == cnt) {
                        res.push_back(left);
                        --m2[s.substr(left, len)];
                        --count;
                        left += len;
                    }
                } else {
                    m2.clear();
                    count = 0;
                    left = j + len;
                }
            }
        }
        return res;
    }
};

  

 

 

All LeetCode Questions List 题目汇总