DEV Community

Prashant Mishra
Prashant Mishra

Posted on

Minimum window substring

Problem

Brute force approach:
TC: O(N^2), SC: O(256) which is constant
Note: This will lead to TLE

class Solution {
    public String minWindow(String s, String t) {


        int min = Integer.MAX_VALUE;
        int start =-1;
        int end = -1;
        //generating all possible substrings and returning the smallest substring having
        //all the character of string t
        for(int i =0;i<s.length();i++){
            int arr[] = new int[256]; // all the asci characters
            //initialize arr with count of character in t
            for(int  k =0;k<t.length();k++){
                arr[t.charAt(k)]++;
            }
            int count =0;
            for(int j =i;j<s.length();j++){
            // if character at j in s is also present in t, increase count
                if(arr[s.charAt(j)] >0){
                    count++;
                }
                arr[s.charAt(j)]--;//decrease the frequency of s.charAt(j) by 1 in arr
                if(count ==t.length()){// if count == length of t we have found one potential substring 
                    if(min> j-i+1){
                        min  = j-i +1;
                        start  = i;
                        end = j;
                    }
                    break;// break out as there is no need to propagate further in the current iteration
                }
            }
        }
        if(start ==-1 || end ==-1) return "";// if not found 
        return s.substring(start,end+1);// else
    }
}
Enter fullscreen mode Exit fullscreen mode

Optimal approach:
TC: O(n) , SC:O(256) wich is constant

// this intuition will require solving such problems as much as possible
class Solution {
    public String minWindow(String s, String t) {
        int left=0,right=0;
        int min = Integer.MAX_VALUE;
        int count =0;
        int arr[] = new int[256]; // has table for keeping count of characters
        // keeping track of start and end index
        int start =-1;
        int end = -1;
        for(int i=0;i<t.length();i++){
            arr[t.charAt(i)]++;
        }
        while(right<s.length()){
            char c = s.charAt(right);
            if(arr[c]>0){ // if the character is part of t, then increment count
                count++;
            }
            arr[c]--;// decrease the count of the character in the hash table once it is taken into consideration 
            while(count ==t.length()){ // if count == t.size() we have found a potential window
                if(min > right-left+1){
                    min = right-left+1;
                    //update the start and end accordingly 
                    start = left;
                    end = right;
                }
                arr[s.charAt(left)]++;//since we are removing this character from window and increasing the left pointer to point to next character after this, there will be one less
                //character that is not part of the t 
                if(arr[s.charAt(left)]>0){ // if after doing +1 for arr[s.charAt(left)] character, if it become positive then it will mean that this was part of t, and now we will have to look for this character in the next window 
                    count--; // since this was part of t, then count should decrement by 1
                }
                left++;
            }
            right++;
        }
        if(start ==-1 || end ==-1) return "";
        return s.substring(start,end+1);
    }
}
Enter fullscreen mode Exit fullscreen mode

Top comments (0)

Great read:

Is it Time to go Back to the Monolith?

History repeats itself. Everything old is new again and I’ve been around long enough to see ideas discarded, rediscovered and return triumphantly to overtake the fad. In recent years SQL has made a tremendous comeback from the dead. We love relational databases all over again. I think the Monolith will have its space odyssey moment again. Microservices and serverless are trends pushed by the cloud vendors, designed to sell us more cloud computing resources.

Microservices make very little sense financially for most use cases. Yes, they can ramp down. But when they scale up, they pay the costs in dividends. The increased observability costs alone line the pockets of the “big cloud” vendors.