LeetCode3_Longest Substring Without Repeating Characters

My Answer

1. 第四次提交（改进）

    

    /**
     * Before update algorithm: <br>
     * 1. 利用hashMap 存储被check过的char和它的index <br>
     * 2. remove hashMap里 所有在repeat char position之前的all chars<br>
     * 3. 利用hashMap.size()计算subString的长度 <br>
     * After update algorithm: <br>
     * 1. 利用hashMap 存储被check过的char和它的index<br>
     * 2. 利用two pointers: left, right作为substring的fromIndex, toIndex. subString长度为right-left <br>
     */
    public int lengthOfLongestSubstring(String s) {
        if (s == null || s.length() == 0) {
            return 0;
        }

        HashMap<Character, Integer> map = new HashMap<Character, Integer>();
        int left = 0, right = 0;
        int maxLgth = 0;

        while (right < s.length()) {
            char ch = s.charAt(right);
            Integer pos = map.get(ch);

            // ch already exists in the substring
            if (pos != null && pos >= left) {
                maxLgth = maxLgth < right - left ? right - left : maxLgth;
                left = pos + 1;
            }

            map.put(ch, right++);
        }

        // check the last one subString.length
        return maxLgth < right - left ? right - left : maxLgth;
    }

2. 第三次提交

<span style="font-weight: normal;">     public int lengthOfLongestSubstring(String s) {
        if (s == null || s.length() == 0) {
            return 0;
        }

        int start = 0;
        int maxLength = 0;
        char[] arr = s.toCharArray();
        HashMap<Character, Integer> charForIndex = new HashMap<Character, Integer>();

        for (int i = 0; i < arr.length; i++) {
            char ch = arr[i];
            Integer index = charForIndex.get(ch); // O(1)

            if (index != null) {
                maxLength = maxLength < charForIndex.size() ? charForIndex.size() : maxLength;
                for (int j = start; j < index + 1; j++) {
                    charForIndex.remove(arr[j]); // remove all chars before index
                }

                start = index + 1;
            }

            charForIndex.put(ch, i); // if key is repeat, update value
        }

        return maxLength = maxLength < charForIndex.size() ? charForIndex.size() : maxLength;
    }</span>

2. 第二次提交

<span style="font-weight: normal;">public int lengthOfLongestSubstring(String s) {
        if (s == null || s.length() == 0) {
            return 0;
        }

        int maxLength = 0;
        char[] arr = s.toCharArray();
        List<Character> subStr = new LinkedList<Character>();

        for (int i = 0; i < arr.length; i++) {
            char ch = arr[i];
            int index = subStr.indexOf(ch);

            if (index > -1) {
                int length = subStr.size();
                maxLength = maxLength < length ? length : maxLength;

                if (index == length - 1) {
                    subStr.clear();
                } else {
                    subStr = subStr.subList(index + 1, length);
                }
            }

            subStr.add(ch);
        }

        return maxLength < subStr.size() ? subStr.size() : maxLength;
    }</span>

3. 第一次提交

public int lengthOfLongestSubstring(String s) {
       if(s == null || s.length() == 0){
            return 0;
        }
        
        char[] arr = s.toCharArray();
        Set<Character> charSet = new HashSet<Character>();
        int maxSubLength = 0;
        for(int i=0 ; i<arr.length; i++){
            if(charSet.contains(arr[i])){
                if(maxSubLength < charSet.size()){
                    maxSubLength = charSet.size();
                }
                charSet.clear();
            }
            charSet.add(arr[i]);
        }
        
        if(maxSubLength < charSet.size()){
            maxSubLength = charSet.size();
        }
        
        return maxSubLength;
    }

Summary

1. 这个题目花了自己很长时间，主要是自己没完全理解题目就“开跑”。 题目tags “Hash Table”,"Two pointers" "String"  。虽然花了很长时间，看到提交运行时间还不错得到些许鼓励。要是正式面试自己这题早超时failure了。

2.  第一次提交基于Set,set是不能保证subString的顺序。一旦出现重复char就会将整个set clear。 例如"abcaxyz"最长子串为"bcaxyz"而不是"axyz"。基于set当遇到第二个a时,set会将'b','c'也clear掉。

3.  第二次提交基于String.indexOf(); String.subString(fromIndex,toIndex)。超时了 。这两个方法都是O(n),在加上外层for-loop.算法复杂度打到O(n^2)

4.  第三次提交基于hashMap. hashMap.get()/remove() running time都是O(1)。另外借助start pointer 去除重复char之前的all chars. 虽然hashMap不能保证subString里char的顺序  没关系，因为题目只需要返回substring的长度。

5. 第四次提交基于two pointers, hashMap. 相对于第三次提交代码更简洁，计算时间更短。

5. String的basic case记得考虑空字符串“”