/*
 * freq_analysis.c
 * (c)Steve Parker, 31st Oct 2003
 * GPL v2 for now.
 * http://www.gnu.org/copyleft/gpl.html
 *
 * See also crypto.c
 * This is just a quick utility to generate a structure definition from a given piece of text.
*/

#include <stdio.h>
#define MAXCRYPTO 999999 
	// Longest string to be translated

struct freq {
  int f;
  char c;
  char t;
};

void bubblesort(struct freq a[], int n)
{
  int i,j;
  struct freq tmp;
 
  for (j=0; j<n-1; j++) {
  for (i = 0; i < n - 1; i++) {
    if (a[i].f < a[i+1].f) {
      tmp = a[i];
      a[i] = a[i+1];
      a[i+1] = tmp;
    }
  }
  }
}

void freq_analysis(char *p)
{
  int i,j;
      int k;
  struct freq f[255];
  char p2[MAXCRYPTO];

  for (i=0; i<255; i++) {
    f[i].f=0;
    f[i].c='\0';
    f[i].t='\0';
  }
  for (i=0; i<strlen(p); i++) {
    f[(int)p[i]].f++;
    f[(int)p[i]].c=p[i];
  }

  bubblesort(f, 256);

  printf("char english[]={ ");
  j=0;
  for (i=0; i<255; i++) { 
    if (f[i].f>0) { 
	while (f[i].c<0) f[i].c+=255;
      if (((int)f[i].c < 127) && ((int)f[i].c!=10)) { // don't bother with 8-bit ASCII for now
        //printf("%4d : %c (%03d)\n", f[i].f, f[i].c, f[i].c);
        printf("'%c', ", f[i].c);
      }
    }
    j++;
   }
   printf(" };\n");
}

int main(int argc, char *argv[])
{
 
  char *p;
  char *s;
  FILE *in;

  p=(char*)malloc(MAXCRYPTO);
  s=(char*)malloc(255);

  in=fopen(argv[1], "rb");
  while (fgets(s, 255, in)) {
    strcat(p, s);
  }
  fclose(in);

  freq_analysis(p);
  free(p);
  
  return 0;
}