/* File:     char_freq.c.
   Contents: Source code for char_freq command.
   Author:   Jeffrey S. Leon
   Updated:  1/25/05

   The char_freq command counts the number of occurences of each letter
   (a-z or A-Z, with upper and lower case letters being treated as
   equivalent) in a text file.  Characters other than letters are ignored.
   The frequencies are printed in decreasing order.

   The syntax for invoking the command is
          char_freq  text_file  adjust_to
   where
          text_file  is the name of the text file.
          adjust_to  (optional)  is an integer that indicates that the
                     frequency of each letter is to be scaled so that the
                     frequencies sum to n, where n is the value of adjust_to.
                     In particular, if adjust_to == 1, the probability of each
                     letter is shown.  If adjust_to is omitted, no scaling is
                     performed.
*/

#include <stdio.h>
#include <ctype.h>
#include <stdlib.h>

typedef int bool;
const int false = 0, true = 1;

struct {
   int freq;
   int letter;
} count[26];

int compar( const void *f1, const void *f2)
{
   int *n1 = (int *)f1, *n2 = (int *)f2;
   return (*n1 > *n2) ? -1 : (*n1 < *n2) ? 1 : 0;
}

int main( int argc, char *argv[])
{
   int ch, total = 0, adjust_to = 0;
   FILE *input;
   char base = 'A';

   /* Check number of command-line arguments, and process arguments. */
   if ( argc != 2 && argc != 3 ) {
      printf( "Error: Wrong number of command-line arguments");
      exit(1);
   }

   input = fopen( argv[1], "r");
   if ( input == NULL ) {
      printf( "Error: File %s could not be opened.\n", argv[1]);
      exit(2);
   }

   if ( argc == 3 ) {
      char *p;
      adjust_to = strtol( argv[2], &p, 0);
      if ( p == argv[2] || adjust_to < 0 ) {
         printf( "Error: Invalid second argument %s.\n", argv[2]);
         exit(2);
      }
   }

   /* Count number of occurences of each letter. */
   for ( ch = 0 ; ch < 26; ++ch ) {
      count[ch].letter = ch;
      count[ch].freq = 0;
   }

   while ( (ch = getc(input)) != EOF )
      if ( isalpha(ch) ) {
         if ( islower(ch) )
            base = 'a';
         ++count[tolower(ch)-'a'].freq;
         ++total;
      }

   /* Sort letters by decreasing frequency. */
   qsort( count, 26, 2*sizeof(int), compar);

   /* Display total number of letters in file. */
   printf( "\n %d alphabetic characters in sample.\n\n", total);

   /* Display frequencies of individual letters, adjusted if necessary. */
   if (adjust_to == 0 ) {
      printf( " Letter Frequency\n");
      for ( ch = 0 ; ch < 26 ; ++ch )
         printf( "    %c  %8d\n", base+count[ch].letter, count[ch].freq);
    }
    else if (adjust_to == 1 ) {
       printf( " Letter   Prob\n");
       for ( ch = 0 ; ch < 26 ; ++ch )
          printf( "    %c  %8.4f\n", base+count[ch].letter, count[ch].freq / (double)total);
    }
    else if (adjust_to < 100 ) {
       printf( " Letter Frequency\n");
       printf( "       (per %d lets)\n", adjust_to);
       for ( ch = 0 ; ch < 26 ; ++ch )
          printf( "    %c  %8.3f\n", base+count[ch].letter, count[ch].freq / (double)total * (double)adjust_to);
    }
    else if (adjust_to < 1000 ) {
       printf( " Letter Frequency\n");
       printf( "        (per %d lets)\n", adjust_to);
       for ( ch = 0 ; ch < 26 ; ++ch )
          printf( "    %c  %8.2f\n", base+count[ch].letter, count[ch].freq / (double)total * (double)adjust_to);
    }
    else if (adjust_to < 10000 ) {
       printf( " Letter Frequency\n");
       printf( "        (per %d lets)\n", adjust_to);
       for ( ch = 0 ; ch < 26 ; ++ch )
          printf( "    %c  %8.1f\n", base+count[ch].letter, count[ch].freq / (double)total * (double)adjust_to);
    }
    else {
       printf( " Letter Frequency\n");
       printf( "        (per %d lets)\n", adjust_to);
       for ( ch = 0 ; ch < 26 ; ++ch )
         printf( "    %c  %8.0f\n", base+count[ch].letter, count[ch].freq / (double)total * (double)adjust_to);
    }

   /* All done. */
   fclose( input);
   return 0;
}