Qfyilyi

Caesar Cipher using Frequency Analysis** in Java: this is my code for the decode part:

public static String decode (String code){

    int key=0;

    final int ALPHABET_SIZE = 26;

    int freqs = new int[ALPHABET_SIZE];

    for (int l=0;l<freqs.length;l++){ 

        freqs[l]=0;

    }

    for (int k=0;k<code.length();k++){

        if (code.charAt(k)>='a' && code.charAt(k)<='z'){

            freqs[code.charAt(k)-'a']++;

        }

    }

    int biggest = 0;

    for (int t=0;t<freqs.length;t++){

        if (freqs[t]>biggest){

            biggest= t;

        }

    }

    if (biggest<4){

        key = (biggest + 26 - ('e'+'a'));



    }

    else{

        key = biggest + 'a' - 'e';

    }

    return (decode(code,key));

}

I cannot use maps, imports, lists, or add a key, I do know that the most freq letter is E but I do not know how to implement it in a different function. I would appreciate a more elegant solution, thank you. ** Frequency Analysis

I am not too sure what your definition of "elegant" is, but your solution looks fine in general.

A few comments:

1. You don't actually have to manually initialise your array to 0 at the start since Java will do that for you




2. Looking for the letter with the highest frequency can be done while you are counting the characters

For example, the for loop can become:

int highestFreq = 0;

int highestFreqIdx = -1;

for (int k=0; k < code.length(); k++) {

    if (code.charAt(k) >= 'a' && code.charAt(k) <= 'z') {

        int count = freqs[code.charAt(k)-'a']++;



        if (count > highestFreq) {

          highestFreq = count;

          highestFreqIdx = k;

        }

    }

}

3. Calculating the key can be simplified

If I am not wrong, your key is the number of positions that the letter with the highest frequency is away from 'e'. In this case, you can simply do this:

key = biggest - ('e' - 'a');

So your code becomes:

public static String decode (String code){

    int key = 0;

    final int ALPHABET_SIZE = 26;

    int freqs = new int[ALPHABET_SIZE];



    int highestFreq = 0;

    int highestFreqIdx = -1;

    for (int k=0; k < code.length(); k++) {

        if (code.charAt(k) >= 'a' && code.charAt(k) <= 'z') {

            int count = freqs[code.charAt(k)-'a']++;



            if (count > highestFreq) {

              highestFreq = count;

              highestFreqIdx = k;

            }

        }

    }



    key = highestFreqIdx - ('e' - 'a'); // Can also directly use 4 instead of 'e' - 'a'



    return (decode(code, key));

}

That said, your code will only work for messages that really have the letter 'e' as the most common letter...

P/S Such code review questions may be better off in this Stack Exchange site for code reviews instead