isindex.c

this gives details of the network programming

/* CGI program for handling simple ISINDEX queries */
/* This program assumes that the query is passed to the
   CGI program in the environment variable QUERY_STRING 
   (as a result of a GET http operation)
   and that there are no spaces or other characters that
   the browser would encode (urlencoding).
   */

#include <stdio.h>
#include <stdlib.h>	/* for getenv */
#include <string.h>
#include <ctype.h>

#define MAXQUERY 1000

char *http_header = "content-type: text/html\n\n";
char *error_message = "<H2>Some kind of error occured while processing your query</H2>\n";

void fixstring( unsigned char *s);

int main( ) {
  char *query;
  FILE *cmd;
  char command[2001];
  char results[2000];


  /* get the query - we assume it is in the environment variable
     QUERY_STRING */

  query = getenv("QUERY_STRING");

  fixstring( query );

  if ((query==NULL)||(strlen(query)==0)) {
    /* No QUERY or it is null - generate error message and quit*/
    printf("%s%s\n",http_header,error_message);
    exit(1);
  } 

  if (strcspn(query,"&;!@#$()~\"") != strlen(query)) {
    printf("%s",http_header);
    printf("<H2>Nice try wise guy, but it's not gonna happen!</H2>\n");
    return(0);
  }

  /* create the grep command */
  snprintf(command,2000,"grep %s /usr/dict/words",query);
  /* start up the grep and pipe the result back here (to cmd) */
  cmd = popen(command,"r");

  if (cmd==NULL) {
    /* problem starting up the grep command */
    printf("%s%s\n",http_header,error_message);
    exit(1);
  }

  /* Grep is running - print out the results */
  printf("%s",http_header);
  printf("<H2>Words found that match \"%s\":</H2>\n",query);
  printf("<UL>\n");

  while (fgets(results,1000,cmd) !=NULL ) {
    printf("<LI>%s\n",results);
  }
  printf("</UL>\n");
  return(0);
}



/* hexval converts a ascii encoded hex digit to the
   corresponding binary value. If we get a numeric
   character ('0' - '9') we return the corresponding
   value 0-9.
   If we get an alphabetic character we return;

   'a' or 'A' -> 10
   'b' or 'B' -> 11
   ...
   'f' or 'F' -> 15

*/

char hexval (  char x ) {
  if (isdigit(x)) {
    /* converting '0' - '9' */
    return(x-'0');
  } else if ( (x>='a') && (x<='f')) {
    /* lowercase 'a' - 'f' */
    return(10+(x-'a'));
  } else if ( (x>='A') && (x<='F')) {
    /* uppercase 'A' - 'F' */
    return(10+(x-'A'));
  }
  /* we got something we didn't expect - return -1 */
  return(-1);
}
   

/* fixstring converts a urlencoded string to a normal string.
   This includes:
     replaces all '+' character by ' '
     replaces all hex equivalents by the single character encoded.
     
   The size of the string may change - but it will only shrink,
   so we change it in place (change the original string).
*/

void fixstring( unsigned char *s) {
  char *p = strdup(s);	/* make a copy of the string to work with */
  char *orig=p;		/* keep track of this copy */

  while (*p) {
    if (*p=='+') {
      /* substitute a blank for a plus */
      *s=' ';
      p++; s++;

      /* Look for a hex encoded character - don't trust the user 
         (make really sure this is hex encoded!) */
    } else if ((strlen(p)>=3) && (*p=='%') /*&& 
	       (isxdigit(p+1)) && (isxdigit(p+2))*/ ) {

      /* hex encoded - convert to a char */

      *s = hexval(p[1])<<4 | hexval(p[2]);
      s++; p+=3;
    } else {
      /* normal character - just copy it without changing */
      *s=*p;
      s++; p++;
    }
  }
  /* terminate the new string */
  *s=0;
  /* and get rid of the duplicate */
  free(orig);
}