[leetcode]187. Repeated DNA Sequences

题目：

All DNA is composed of a series of nucleotides abbreviated as A, C, G, and T, for example: “ACGAATTCCG”. When studying DNA, it is sometimes useful to identify repeated sequences within the DNA.

Write a function to find all the 10-letter-long sequences (substrings) that occur more than once in a DNA molecule.

For example,

Given s = “AAAAACCCCCAAAAACCCCCCAAAAAGGGTTT”,

Return:
[“AAAAACCCCC”, “CCCCCAAAAA”].

题解：

这个题题意不难理解，但是刚开始我以为有更多的解，后面才发现第一段连续的c有5个，第二段有6个，所以只存在这两个解。
思路是，将s分成10个一段，然后用map进行存储，将子串作为key，计算出重复次数就可以。

class Solution {
public:
    vector<string> findRepeatedDnaSequences(string s) {
    int i = 0;
    int j = 0;
    map<string, int> map;
    vector<string> vc;
    if(s.length()<=10){
        return vc;
    }
    for (; i <= s.length()-10; i++){
        string subs;
        subs = s.substr(i, 10);
        //cout << subs << endl;
        if (map[subs] == NULL){
            map[subs] = 1;
        }
        else{
            map[subs]++;
            if (map[subs]==2)
                vc.push_back(subs);
        }
    }
    return vc;
}
};

虽然ac，但是却用了208ms，比较耗时，然后就查了一下网上的题解，有一个比较有意思的思路是这样的;
1.将四个字符进行编码，A—00;B—01;C—10;D—11;
2.10长度的字符串可以用20位二进制数表示，int类型有4个字节，32位，可以用来表示子串
3.截取子串，可以让字符串左移两位，最高位补新字符的编码；
4.用hashtable 进行存储。
在线性时间就可以解决！666啊
学习到了，以后也要多想想如何省时间。
代码有空贴上

 #include <string>
 #include <vector>
 #include <unordered_set>
 #include <cstring>

 bool hashMap[1024*1024];

 class Solution {
 public:
     std::vector<std::string> findRepeatedDnaSequences(std::string s);
 };

 std::vector<std::string> Solution::findRepeatedDnaSequences(std::string s) {
     std::vector<std::string> rel;
     if (s.length() <= 10) {
         return rel;
     }

     // map char to code
     unsigned char convert[26];
     convert[0] = 0; // 'A' - 'A'  00
     convert[2] = 1; // 'C' - 'A'  01
     convert[6] = 2; // 'G' - 'A'  10
     convert[19] = 3; // 'T' - 'A' 11

     // initial process
     // as ten length string
     memset(hashMap, false, sizeof(hashMap));

     int hashValue = 0;

     for (int pos = 0; pos < 10; ++pos) {
         hashValue <<= 2;
         hashValue |= convert[s[pos] - 'A'];
     }

     hashMap[hashValue] = true;

     std::unordered_set<int> strHashValue;

     // 
     for (int pos = 10; pos < s.length(); ++pos) {
         hashValue <<= 2;
         hashValue |= convert[s[pos] - 'A'];
         hashValue &= ~(0x300000);

         if (hashMap[hashValue]) {
             if (strHashValue.find(hashValue) == strHashValue.end()) {
                rel.push_back(s.substr(pos - 9, 10));
                 strHashValue.insert(hashValue);
             }
         } else {
             hashMap[hashValue] = true;
         }
     }

    return rel; 
 }