From 2c26f4bbcc694f90568f71b4e4d8b551138f6069 Mon Sep 17 00:00:00 2001
From: Dimitri Lozeve <dimitri@lozeve.com>
Date: Fri, 5 Jun 2020 15:28:55 +0200
Subject: [PATCH] Find the most probable keysize using normalized edit distance

---
 ex06/main.c | 50 +++++++++++++++++++++++++++++++++++++++++++++++++-
 1 file changed, 49 insertions(+), 1 deletion(-)

diff --git a/ex06/main.c b/ex06/main.c
index 167c594..0fa5e45 100644
--- a/ex06/main.c
+++ b/ex06/main.c
@@ -1,8 +1,32 @@
 #include "utils.h"
 #include <ctype.h>
+#include <math.h>
 #include <stdio.h>
 #include <string.h>
 
+#define BUF_SIZE 4096
+
+double normalized_edit_distance(const unsigned char *buf,
+                                unsigned int keysize) {
+  char block1[keysize];
+  char block2[keysize];
+  char block3[keysize];
+  char block4[keysize];
+
+  for (size_t i = 0; i < keysize; ++i) {
+    block1[i] = buf[i];
+    block2[i] = buf[keysize + i];
+    block3[i] = buf[2 * keysize + i];
+    block4[i] = buf[3 * keysize + i];
+  }
+
+  double dist1 = hamming(block1, block2) / (double)keysize;
+  double dist2 = hamming(block2, block3) / (double)keysize;
+  double dist3 = hamming(block3, block4) / (double)keysize;
+
+  return (dist1 + dist2 + dist3) / 3;
+}
+
 int main(int argc, char *argv[]) {
   if (argc < 2) {
     printf("Usage: %s <filename>\n", argv[0]);
@@ -17,7 +41,31 @@ int main(int argc, char *argv[]) {
     return EXIT_FAILURE;
   }
 
-  printf("%u\n", hamming("this is a test", "wokka wokka!!!"));
+  char input[BUF_SIZE] = {'\0'};
+  size_t nread = fread(input, 1, BUF_SIZE, fp);
+  if (nread == 0) {
+    printf("Cannot read any character from file %s\n", filename);
+    return EXIT_FAILURE;
+  }
+
+  unsigned char buf[BUF_SIZE] = {0};
+  size_t len = base64_to_bytes(buf, input);
+  if (len == 0) {
+    printf("Could not decode base64\n");
+    return EXIT_FAILURE;
+  }
+
+  unsigned int keysize = 40;
+  double min_edit_dist = INFINITY;
+  for (unsigned int i = 40; i > 1; --i) {
+    double edit_dist = normalized_edit_distance(buf, i);
+    if (edit_dist < min_edit_dist) {
+      min_edit_dist = edit_dist;
+      keysize = i;
+    }
+  }
+
+  printf("keysize = %u, min edit distance = %f\n", keysize, min_edit_dist);
 
   if (fclose(fp)) {
     printf("Error closing file %s\n", filename);