Longest Substring Without Repeating Characters (leetcode 3)

version 1：

本问题是寻找最长的子串，且子串无重复字符，且注意是子串不是子序列。简单粗暴的用直觉想了个时间复杂度为的方法，简单粗暴的办法，先work再说，但是也能发现，效率是很低的，稍后的版本再选择进行提升。简单粗暴的办法就是使用双重循环（屡试不爽哈哈）：外重循环遍历整个字符串，内重循环发现以当前元素最为起始字符的最长字符串

class Solution:
    def lengthOfLongestSubstring(self, s):
        """
        :type s: str
        :rtype: int
        """

        # pick out the special case
        if len(s)<1 or s == None:
        	return 0
        # intuition lead us to do it like this
        # use two loop
        # the outside loop traverses all the character
        # the inside loop to find the longest substring without repeating charaters
        rlength = 1
        length = len(s)
        for i in range(0,length):
        	# the outside loop
        	for j in range(i+1,length):
        		# the inside loop

        		# if s[j] not in s[i:j](s[i...j-1],end of s[j-1])
        		# we can add the length of the substring which start at s[i]
        		# else we should deal with s[i+1]
        		if s[j] not in s[i:j]:
        			if j-i+1 > rlength:
        				rlength = j-i+1			
        		else:
        			break
       
        return rlength

version 2:

第一个版本的方法虽然思想简单，但耗时太长，所以经过一番冥思苦想，采用滑动窗口的方法，将两层循环变成一层循环。使用两个定位的变量begin 和 end，end随着循环遍历不断的往后游走，对当前处理的字符，检查是否出现在begin 到 end之间，如果没有，end往后一位，意味着滑动窗口增大；若出现了，与历史最大滑动窗口比较，记录下最大的滑动窗口，并将start移动到重复字符的后面位置。本次算法的时间复杂度是。结果时间上降低了一个量级。代码与结果如下：

class Solution:
    def lengthOfLongestSubstring(self, s):
        """
        :type s: str
        :rtype: int
        """

        # pick out the special case
        if len(s)<1 or s == None:
        	return 0
        begin = 0
        end = 1
        # begin and end is the begin and end of the sliding window
        maxWin = 1
        # update maxWin = end - begin

        for i in range(1,len(s)):
        	# if the current element exist in the window s[beging:end]
        	# we should start the window in the later element behind the repeat element
        	# at the same time, update the maxWin

        	# whatever situation,we should add 1 to end
        	if s[i] in s[begin:end]:
        		if end - begin > maxWin:
        			maxWin = end - begin
        		begin += s[begin:end].index(s[i]) + 1
        		end += 1
        	else:
        		end += 1
        # don't forget update the maxWin in the end
        if end - begin > maxWin:
        			maxWin = end - begin
       
        return maxWin		

这暂时是笔者想到的比较好的方法，可能会有更精妙的解法，等笔者解锁新解法的时候再来更新，也欢迎交流。