/*----------------------------------------------------------------------*\
 | Ignore hits by robots and spiders; use key text in the agent string	|
 | to recognize these users.											|
\*----------------------------------------------------------------------*/

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

#include <sys/types.h>
#include <sys/stat.h>

static char *read_text (char *name) {
	long n;
	FILE *fp;
	char *dst,*end;
	long size;
	char *number;

	if (!name) return (NULL);

	dst = NULL;
	if (fp = fopen (name,"rb")) {
		if (fseek (fp,0L,SEEK_END) == 0) {
			size = 1L + ftell (fp);
			if (size > 1L) {
				if (dst = (char *) malloc (size)) {
					rewind (fp);
					n = fread (dst,1,size,fp);
					end = dst + n;
					*end = 0;
					}
				else
					fprintf (stderr,"Error: could not allocate space for file\n");
				}
			else
				fprintf (stderr,"Error: Input file is empty.\n");
			}
		fclose (fp);
		}
	else
		fprintf (stderr,"Error: could not open input file %s\n",name);
	return (dst);
	}

static char **spider_name = NULL;
static int spider_name_count = 0;
static char *spider_name_buffer = NULL;

static int compare (const void *e1, const void *e2) {
	const char *s1 = *(const char **) e1;
	const char *s2 = *(const char **) e2;
	return (strcmp (s1,s2));
	}

int read_spider_names (char *spider_name_file) {
	char *ptr, *eol;
	if (spider_name_buffer = read_text (spider_name_file)) {
		int n = 1;
		for (ptr = spider_name_buffer; *ptr; *ptr++)
			if (*ptr == '\n') n++;

		if (spider_name = (char **) malloc (n * sizeof (char *))) {
			char **p = spider_name;
			ptr = spider_name_buffer;
			while (*ptr) {
				for (eol=ptr; *eol; eol++) {
					if (*eol == '\r') *eol++ = 0;
					if (*eol == '\n') {
						*eol++ = 0;
						break;
						}
					}

				/*------------------------------------------------------*\
				 | Ignore blank lines and comments (any line beginning	|
				 | with #) in the list of spiders.						|
				\*------------------------------------------------------*/

				if (strlen(ptr) > 0 && *ptr != '#') *p++ = ptr;

				ptr = eol;
				}
			spider_name_count = p - spider_name;
			qsort (spider_name,spider_name_count,sizeof(char *),compare);
			}
		}
	return (spider_name_count);
	}

int is_spider (char *agent, char *host) {
	int i;
	if (agent && *agent)
		for (i=0; i < spider_name_count; i++)
			if (strstr (agent,spider_name[i])) return (1);
	if (host && *host)
		for (i=0; i < spider_name_count; i++)
			if (strstr (host,spider_name[i])) return (1);
	return (0);
	}

#ifdef TEST
int main (int argc, char *argv[]) {
	int i,n;
	n = read_spider_names (spider_name_file);
	if (n > 0) {
		printf ("%2d spider names from %s\n",spider_name_count,spider_name_file);
		if (argc > 1)
			for (i=1; i < argc; i++)
				if (is_spider (argv[i]))
					printf ("Yes: %s\n",argv[i]);
				else
					printf ("No:  %s\n",argv[i]);
		}
	else
		printf ("No spider names found in %s\n",spider_name_file);
	}
#endif

/*----------------------------------------------------------------------*\
\*----------------------------------------------------------------------*/