dupirefr/dupire.francois+pro@gmail.com [BAEL-2452] Caesar Cipher in Java (#8209)

* Added benchmarking on larger matrices * [BAEL-3452] Added cipher algorithm * [BAEL-3452] Added deciphering * [BAEL-3452] Added break algorithm * [BAEL-3452] Finalizing break algorithm * Revert "Added benchmarking on larger matrices" This reverts commit 4ea87c0aea44bd73be721691a3c427b03854b442.
2019-11-19 08:24:38 +01:00
parent 7d52a39289
commit 6b9f2c74ee
3 changed files with 173 additions and 0 deletions
@@ -0,0 +1,83 @@
+package com.baeldung.algorithms.caesarcipher;
+
+import org.apache.commons.math3.stat.inference.ChiSquareTest;
+
+import java.util.Arrays;
+import java.util.stream.IntStream;
+
+public class CaesarCipher {
+    private static final char LETTER_A = 'a';
+    private static final char LETTER_Z = 'z';
+    private static final int ALPHABET_SIZE = LETTER_Z - LETTER_A + 1;
+    private static final double[] ENGLISH_LETTERS_PROBABILITIES = {0.073, 0.009, 0.030, 0.044, 0.130, 0.028, 0.016, 0.035, 0.074, 0.002, 0.003, 0.035, 0.025, 0.078, 0.074, 0.027, 0.003, 0.077, 0.063, 0.093, 0.027, 0.013, 0.016, 0.005, 0.019, 0.001};
+
+    public String cipher(String message, int offset) {
+        StringBuilder result = new StringBuilder();
+
+        for (char character : message.toCharArray()) {
+            if (character != ' ') {
+                int originalAlphabetPosition = character - LETTER_A;
+                int newAlphabetPosition = (originalAlphabetPosition + offset) % ALPHABET_SIZE;
+                char newCharacter = (char) (LETTER_A + newAlphabetPosition);
+                result.append(newCharacter);
+            } else {
+                result.append(character);
+            }
+        }
+
+        return result.toString();
+    }
+
+    public String decipher(String message, int offset) {
+        return cipher(message, ALPHABET_SIZE - (offset % ALPHABET_SIZE));
+    }
+
+    public int breakCipher(String message) {
+        return probableOffset(chiSquares(message));
+    }
+
+    private double[] chiSquares(String message) {
+        double[] expectedLettersFrequencies = expectedLettersFrequencies(message.length());
+
+        double[] chiSquares = new double[ALPHABET_SIZE];
+
+        for (int offset = 0; offset < chiSquares.length; offset++) {
+            String decipheredMessage = decipher(message, offset);
+            long[] lettersFrequencies = observedLettersFrequencies(decipheredMessage);
+            double chiSquare = new ChiSquareTest().chiSquare(expectedLettersFrequencies, lettersFrequencies);
+            chiSquares[offset] = chiSquare;
+        }
+
+        return chiSquares;
+    }
+
+    private long[] observedLettersFrequencies(String message) {
+        return IntStream.rangeClosed(LETTER_A, LETTER_Z)
+          .mapToLong(letter -> countLetter((char) letter, message))
+          .toArray();
+    }
+
+    private long countLetter(char letter, String message) {
+        return message.chars()
+          .filter(character -> character == letter)
+          .count();
+    }
+
+    private double[] expectedLettersFrequencies(int messageLength) {
+        return Arrays.stream(ENGLISH_LETTERS_PROBABILITIES)
+          .map(probability -> probability * messageLength)
+          .toArray();
+    }
+
+    private int probableOffset(double[] chiSquares) {
+        int probableOffset = 0;
+        for (int offset = 0; offset < chiSquares.length; offset++) {
+            System.out.println(String.format("Chi-Square for offset %d: %.2f", offset, chiSquares[offset]));
+            if (chiSquares[offset] < chiSquares[probableOffset]) {
+                probableOffset = offset;
+            }
+        }
+
+        return probableOffset;
+    }
+}