zoukankan      html  css  js  c++  java
  • [LeetCode] Word Break II

    Well, it seems that many people meet the TLE problem. Well, I use a simple trick in my code to aoivd TLE. That is, each time before I try to break s, I call isBreakable (just wordBreak from Word Break) to see whether it is breakable. If it is not breakable, then we need not move on. Otherwise, we break it using backtracking.

    The idea of backtracking is also typical. Start from index 0 of s, each time we see a word that is in wordDict, add it to a temporary string sentence. When we reach the end of the s, add sentence to sentences. Then we need to recover sentence back to its original status before word was added. So I simply record a temporary string temp each time before I add word to sentence and use it to recover the status.

    The code is as follows. I hope it to be self-explanatory enough.

     1 class Solution {
     2 public:
     3     bool isWordBreak(string& s, unordered_set<string>& wordDict) {
     4         vector<bool> dp(s.length() + 1, false);
     5         dp[0] = true;
     6         int minlen = INT_MAX;
     7         int maxlen = INT_MIN;
     8         for (string word : wordDict) {
     9             minlen = min(minlen, (int)word.length());
    10             maxlen = max(maxlen, (int)word.length());
    11         }
    12         for (int i = 1; i <= s.length(); i++) {
    13             for (int j = i - minlen; j >= max(0, i - maxlen); j--) {
    14                 if (dp[j] && wordDict.find(s.substr(j, i - j)) != wordDict.end()) {
    15                     dp[i] = true;
    16                     break;
    17                 }
    18             }
    19         }
    20         return dp[s.length()];
    21     }
    22     void breakWords(string s, int idx, unordered_set<string>& wordDict, string& sol, vector<string>& res) {
    23         if (idx == s.length()) {
    24             sol.resize(sol.length() - 1);
    25             res.push_back(sol);
    26             return;
    27         }
    28         for (int i = idx; i < s.length(); i++) {
    29             string word = s.substr(idx, i - idx + 1);
    30             if (wordDict.find(word) != wordDict.end()) {
    31                 string tmp = sol;
    32                 sol += word + " ";
    33                 breakWords(s, i + 1, wordDict, sol, res);
    34                 sol = tmp;
    35             }
    36         }
    37     }
    38     vector<string> wordBreak(string s, unordered_set<string>& wordDict) {
    39         string sol;
    40         vector<string> res;
    41         if (!isWordBreak(s, wordDict)) return res;
    42         breakWords(s, 0, wordDict, sol, res);
    43         return res;
    44     }
    45 };
  • 相关阅读:
    python3 进程_multiprocessing模块
    python3 线程_threading模块
    python3 异常处理
    python3 面向对象补充
    python3 多态,绑定方法与非绑定方法
    python3 uper(),继承实现原理,封装
    python3 面向对象、类、继承、组合、派生、接口、子类重用父类方法
    python3 正则表达式re模块
    python3 json、logging、sys模块
    mysql查询练习
  • 原文地址:https://www.cnblogs.com/jcliBlogger/p/4565214.html
Copyright © 2011-2022 走看看