Thread: Parsing bittorrent metafile..

Hello,

I am trying to receive information from a bittorrent metafile. It uses the following format:

Code:

<string length encoded in base ten ASCII>:<string data>

Example:

Code:

5:Hello

This should return the string Hello.

Begin of a bittorrent metafile:

Code:

d8:announce39:http://www.point-blank.cc:6969/announce13:announce-list

After a few attempts, using strchr to find ':', I got stuck. Are there any other ways of extracting the right values from those files?

OS: Windows XP

Code:

#include <stdio.h>

int main()
{
   static const char filename[] = "file.txt";
   FILE *file = fopen(filename, "r");
   if ( file != NULL )
   {
      char line[BUFSIZ];
      while ( fgets(line, sizeof line, file) != NULL )
      {
         char text[20];
         if ( sscanf(line, "%*d:%19s", text) == 1 )
         {
            printf("text = \"%s\"\n", text);
         }
      }
   }
   else
   {
      perror(filename);
   }
   return 0;
}

/* my output
text = "Hello"
*/

Thank you for the responses, I can work further now I guess, if not, I will post my questions here.

The 'd' stands for dictionaries, you can find more info here

One does wonder......

Code:

int getintlen(int n, int base)
{
    int len = 0;
    while(n)
    {
        n /= base;
        ++len;
    }
    return len;
}

jafet's code counts the number of digits in an integral number, such as an int, not a string. It sounds like you're trying to check how many digits the number is after you've converted it to a string. You can do this, too, but not with jafet's code.

So you have a string that represents a number and you want to figure out how many digits are in it? Can you use strlen() or are there other things in the string?

Originally Posted by apsync

Thank you for the responses, I can work further now I guess, if not, I will post my questions here.

The 'd' stands for dictionaries, you can find more info here

I'd still go with sscanf, but that's me. For example a byte string and an integer could be done this way:

Code:

int bt_bytestring(const char **text)
{
   int size, n;
   if ( sscanf(*text, "%d:%n", &size, &n) == 1 )
   {
      *text += n;
      printf("\"%.*s\"", size, *text);
      *text += size;
      return 1;
   }
   return 0;
}

int bt_integer(const char **text)
{
   int value, n;
   if ( sscanf(*text, "i%de%n", &value, &n) == 1 )
   {
      *text += n;
      printf("%d", value);
      return 1;
   }
   return 0;
}

These could be called by some overall function that looks at the current location in a string.

Code:

int process(const char **text)
{
   return bt_bytestring (text) || 
          bt_integer    (text) || 
          bt_list       (text) || 
          bt_dictionary (text);
}

The list and the dictionary are a bit more troublesome, but they are based on the previous.

Code:

int bt_list(const char **text)
{
   if ( **text == 'l' )
   {
      fputs("[ ", stdout);
      ++*text;
      while ( process(text) )
      {
         if ( **text == 'e' )
         {
            fputs(" ]", stdout);
            ++*text;
            return 1;
         }
         fputs(", ", stdout);
      }
   }
   return 0;
}

int bt_dict(const char **text)
{
   if ( !process(text) )
   {
      return 0;
   }
   fputs(" => ", stdout);
   if ( !process(text) )
   {
      return 0;
   }
   return 1;
}

int bt_dictionary(const char **text)
{
   if ( **text != 'd' )
   {
      return 0;
   }
   fputs("{ ", stdout);
   ++*text;
   while ( bt_dict(text) && **text != 'e' )
   {
      fputs(", ", stdout);
   }
   fputs(" }", stdout);
   ++*text;
   return 1;
}

But a little test might look like this.

Code:

int main(void)
{
   static const char *text[] =
   {
      "4:spam",
      "i3e",
      "l4:spam4:eggse",
      "d3:cow3:moo4:spam4:eggse",
      "d4:spaml1:a1:bee",
   };
   size_t i;
   for ( i = 0; i < sizeof text / sizeof *text; ++i )
   {
      printf("\"%s\" : ", text[i]);
      if ( !process(&text[i]) )
      {
         fputs("bad format", stdout);
      }
      putchar('\n');
   }
   return 0;
}

/* my output
"4:spam" : "spam"
"i3e" : 3
"l4:spam4:eggse" : [ "spam", "eggs" ]
"d3:cow3:moo4:spam4:eggse" : { "cow" => "moo", "spam" => "eggs" }
"d4:spaml1:a1:bee" : { "spam" => [ "a", "b" ] }
*/