xref: /freebsd/usr.bin/whois/whois.c (revision 2008043f386721d58158e37e0d7e50df8095942d)
1 /*-
2  * SPDX-License-Identifier: BSD-3-Clause
3  *
4  * Copyright (c) 1980, 1993
5  *	The Regents of the University of California.  All rights reserved.
6  *
7  * Redistribution and use in source and binary forms, with or without
8  * modification, are permitted provided that the following conditions
9  * are met:
10  * 1. Redistributions of source code must retain the above copyright
11  *    notice, this list of conditions and the following disclaimer.
12  * 2. Redistributions in binary form must reproduce the above copyright
13  *    notice, this list of conditions and the following disclaimer in the
14  *    documentation and/or other materials provided with the distribution.
15  * 3. Neither the name of the University nor the names of its contributors
16  *    may be used to endorse or promote products derived from this software
17  *    without specific prior written permission.
18  *
19  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
20  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
21  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
22  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
23  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
24  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
25  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
26  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
27  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
28  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
29  * SUCH DAMAGE.
30  */
31 
32 #ifndef lint
33 static const char copyright[] =
34 "@(#) Copyright (c) 1980, 1993\n\
35 	The Regents of the University of California.  All rights reserved.\n";
36 #endif /* not lint */
37 
38 #if 0
39 #ifndef lint
40 static char sccsid[] = "@(#)whois.c	8.1 (Berkeley) 6/6/93";
41 #endif /* not lint */
42 #endif
43 
44 #include <sys/cdefs.h>
45 #include <sys/types.h>
46 #include <sys/socket.h>
47 #include <sys/poll.h>
48 #include <netinet/in.h>
49 #include <arpa/inet.h>
50 #include <ctype.h>
51 #include <err.h>
52 #include <netdb.h>
53 #include <stdarg.h>
54 #include <stdio.h>
55 #include <stdlib.h>
56 #include <string.h>
57 #include <sysexits.h>
58 #include <unistd.h>
59 #include <fcntl.h>
60 #include <errno.h>
61 
62 #define	ABUSEHOST	"whois.abuse.net"
63 #define	ANICHOST	"whois.arin.net"
64 #define	DENICHOST	"whois.denic.de"
65 #define	DKNICHOST	"whois.dk-hostmaster.dk"
66 #define	FNICHOST	"whois.afrinic.net"
67 #define	GNICHOST	"whois.nic.gov"
68 #define	IANAHOST	"whois.iana.org"
69 #define	INICHOST	"whois.internic.net"
70 #define	KNICHOST	"whois.krnic.net"
71 #define	LNICHOST	"whois.lacnic.net"
72 #define	MNICHOST	"whois.ra.net"
73 #define	PDBHOST		"whois.peeringdb.com"
74 #define	PNICHOST	"whois.apnic.net"
75 #define	QNICHOST_TAIL	".whois-servers.net"
76 #define	RNICHOST	"whois.ripe.net"
77 #define	VNICHOST	"whois.verisign-grs.com"
78 
79 #define	DEFAULT_PORT	"whois"
80 
81 #define WHOIS_RECURSE	0x01
82 #define WHOIS_QUICK	0x02
83 #define WHOIS_SPAM_ME	0x04
84 
85 #define CHOPSPAM	">>> Last update of WHOIS database:"
86 
87 #define ishost(h) (isalnum((unsigned char)h) || h == '.' || h == '-')
88 
89 #define SCAN(p, end, check)					\
90 	while ((p) < (end))					\
91 		if (check) ++(p);				\
92 		else break
93 
94 static struct {
95 	const char *suffix, *server;
96 } whoiswhere[] = {
97 	/* Various handles */
98 	{ "-ARIN", ANICHOST },
99 	{ "-NICAT", "at" QNICHOST_TAIL },
100 	{ "-NORID", "no" QNICHOST_TAIL },
101 	{ "-RIPE", RNICHOST },
102 	/* Nominet's whois server doesn't return referrals to JANET */
103 	{ ".ac.uk", "ac.uk" QNICHOST_TAIL },
104 	{ ".gov.uk", "ac.uk" QNICHOST_TAIL },
105 	{ "", IANAHOST }, /* default */
106 	{ NULL, NULL } /* safety belt */
107 };
108 
109 #define WHOIS_REFERRAL(s) { s, sizeof(s) - 1 }
110 static struct {
111 	const char *prefix;
112 	size_t len;
113 } whois_referral[] = {
114 	WHOIS_REFERRAL("whois:"), /* IANA */
115 	WHOIS_REFERRAL("Whois Server:"),
116 	WHOIS_REFERRAL("Registrar WHOIS Server:"), /* corporatedomains.com */
117 	WHOIS_REFERRAL("ReferralServer:  whois://"), /* ARIN */
118 	WHOIS_REFERRAL("ReferralServer:  rwhois://"), /* ARIN */
119 	WHOIS_REFERRAL("descr:          region. Please query"), /* AfriNIC */
120 	{ NULL, 0 }
121 };
122 
123 /*
124  * We have a list of patterns for RIRs that assert ignorance rather than
125  * providing referrals. If that happens, we guess that ARIN will be more
126  * helpful. But, before following a referral to an RIR, we check if we have
127  * asked that RIR already, and if so we make another guess.
128  */
129 static const char *actually_arin[] = {
130 	"netname:        ERX-NETBLOCK\n", /* APNIC */
131 	"netname:        NON-RIPE-NCC-MANAGED-ADDRESS-BLOCK\n",
132 	NULL
133 };
134 
135 static struct {
136 	int loop;
137 	const char *host;
138 } try_rir[] = {
139 	{ 0, ANICHOST },
140 	{ 0, RNICHOST },
141 	{ 0, PNICHOST },
142 	{ 0, FNICHOST },
143 	{ 0, LNICHOST },
144 	{ 0, NULL }
145 };
146 
147 static void
148 reset_rir(void) {
149 	int i;
150 
151 	for (i = 0; try_rir[i].host != NULL; i++)
152 		try_rir[i].loop = 0;
153 }
154 
155 static const char *port = DEFAULT_PORT;
156 
157 static const char *choose_server(char *);
158 static struct addrinfo *gethostinfo(const char *, const char *, int);
159 static void s_asprintf(char **ret, const char *format, ...) __printflike(2, 3);
160 static void usage(void) __dead2;
161 static void whois(const char *, const char *, const char *, int);
162 
163 int
164 main(int argc, char *argv[])
165 {
166 	const char *country, *host;
167 	int ch, flags;
168 
169 #ifdef	SOCKS
170 	SOCKSinit(argv[0]);
171 #endif
172 
173 	country = host = NULL;
174 	flags = 0;
175 	while ((ch = getopt(argc, argv, "aAbc:fgh:iIklmp:PQrRS")) != -1) {
176 		switch (ch) {
177 		case 'a':
178 			host = ANICHOST;
179 			break;
180 		case 'A':
181 			host = PNICHOST;
182 			break;
183 		case 'b':
184 			host = ABUSEHOST;
185 			break;
186 		case 'c':
187 			country = optarg;
188 			break;
189 		case 'f':
190 			host = FNICHOST;
191 			break;
192 		case 'g':
193 			host = GNICHOST;
194 			break;
195 		case 'h':
196 			host = optarg;
197 			break;
198 		case 'i':
199 			host = INICHOST;
200 			break;
201 		case 'I':
202 			host = IANAHOST;
203 			break;
204 		case 'k':
205 			host = KNICHOST;
206 			break;
207 		case 'l':
208 			host = LNICHOST;
209 			break;
210 		case 'm':
211 			host = MNICHOST;
212 			break;
213 		case 'p':
214 			port = optarg;
215 			break;
216 		case 'P':
217 			host = PDBHOST;
218 			break;
219 		case 'Q':
220 			flags |= WHOIS_QUICK;
221 			break;
222 		case 'r':
223 			host = RNICHOST;
224 			break;
225 		case 'R':
226 			flags |= WHOIS_RECURSE;
227 			break;
228 		case 'S':
229 			flags |= WHOIS_SPAM_ME;
230 			break;
231 		case '?':
232 		default:
233 			usage();
234 			/* NOTREACHED */
235 		}
236 	}
237 	argc -= optind;
238 	argv += optind;
239 
240 	if (!argc || (country != NULL && host != NULL))
241 		usage();
242 
243 	/*
244 	 * If no host or country is specified, rely on referrals from IANA.
245 	 */
246 	if (host == NULL && country == NULL) {
247 		if ((host = getenv("WHOIS_SERVER")) == NULL &&
248 		    (host = getenv("RA_SERVER")) == NULL) {
249 			if (!(flags & WHOIS_QUICK))
250 				flags |= WHOIS_RECURSE;
251 		}
252 	}
253 	while (argc-- > 0) {
254 		if (country != NULL) {
255 			char *qnichost;
256 			s_asprintf(&qnichost, "%s%s", country, QNICHOST_TAIL);
257 			whois(*argv, qnichost, port, flags);
258 			free(qnichost);
259 		} else
260 			whois(*argv, host != NULL ? host :
261 			      choose_server(*argv), port, flags);
262 		reset_rir();
263 		argv++;
264 	}
265 	exit(0);
266 }
267 
268 static const char *
269 choose_server(char *domain)
270 {
271 	size_t len = strlen(domain);
272 	int i;
273 
274 	for (i = 0; whoiswhere[i].suffix != NULL; i++) {
275 		size_t suffix_len = strlen(whoiswhere[i].suffix);
276 		if (len > suffix_len &&
277 		    strcasecmp(domain + len - suffix_len,
278 			       whoiswhere[i].suffix) == 0)
279 			return (whoiswhere[i].server);
280 	}
281 	errx(EX_SOFTWARE, "no default whois server");
282 }
283 
284 static struct addrinfo *
285 gethostinfo(const char *host, const char *hport, int exit_on_noname)
286 {
287 	struct addrinfo hints, *res;
288 	int error;
289 
290 	memset(&hints, 0, sizeof(hints));
291 	hints.ai_flags = AI_CANONNAME;
292 	hints.ai_family = AF_UNSPEC;
293 	hints.ai_socktype = SOCK_STREAM;
294 	res = NULL;
295 	error = getaddrinfo(host, hport, &hints, &res);
296 	if (error && (exit_on_noname || error != EAI_NONAME))
297 		err(EX_NOHOST, "%s: %s", host, gai_strerror(error));
298 	return (res);
299 }
300 
301 /*
302  * Wrapper for asprintf(3) that exits on error.
303  */
304 static void
305 s_asprintf(char **ret, const char *format, ...)
306 {
307 	va_list ap;
308 
309 	va_start(ap, format);
310 	if (vasprintf(ret, format, ap) == -1) {
311 		va_end(ap);
312 		err(EX_OSERR, "vasprintf()");
313 	}
314 	va_end(ap);
315 }
316 
317 static int
318 connect_to_any_host(struct addrinfo *hostres)
319 {
320 	struct addrinfo *res;
321 	nfds_t i, j;
322 	size_t count;
323 	struct pollfd *fds;
324 	int timeout = 180, s = -1;
325 
326 	for (res = hostres, count = 0; res; res = res->ai_next)
327 		count++;
328 	fds = calloc(count, sizeof(*fds));
329 	if (fds == NULL)
330 		err(EX_OSERR, "calloc()");
331 
332 	/*
333 	 * Traverse the result list elements and make non-block
334 	 * connection attempts.
335 	 */
336 	count = i = 0;
337 	for (res = hostres; res != NULL; res = res->ai_next) {
338 		s = socket(res->ai_family, res->ai_socktype | SOCK_NONBLOCK,
339 		    res->ai_protocol);
340 		if (s < 0)
341 			continue;
342 		if (connect(s, res->ai_addr, res->ai_addrlen) < 0) {
343 			if (errno == EINPROGRESS) {
344 				/* Add the socket to poll list */
345 				fds[i].fd = s;
346 				fds[i].events = POLLERR | POLLHUP |
347 						POLLIN | POLLOUT;
348 				/*
349 				 * From here until a socket connects, the
350 				 * socket fd is owned by the fds[] poll array.
351 				 */
352 				s = -1;
353 				count++;
354 				i++;
355 			} else {
356 				close(s);
357 				s = -1;
358 
359 				/*
360 				 * Poll only if we have something to poll,
361 				 * otherwise just go ahead and try next
362 				 * address
363 				 */
364 				if (count == 0)
365 					continue;
366 			}
367 		} else
368 			goto done;
369 
370 		/*
371 		 * If we are at the last address, poll until a connection is
372 		 * established or we failed all connection attempts.
373 		 */
374 		if (res->ai_next == NULL)
375 			timeout = INFTIM;
376 
377 		/*
378 		 * Poll the watched descriptors for successful connections:
379 		 * if we still have more untried resolved addresses, poll only
380 		 * once; otherwise, poll until all descriptors have errors,
381 		 * which will be considered as ETIMEDOUT later.
382 		 */
383 		do {
384 			int n;
385 
386 			n = poll(fds, i, timeout);
387 			if (n == 0) {
388 				/*
389 				 * No event reported in time.  Try with a
390 				 * smaller timeout (but cap at 2-3ms)
391 				 * after a new host have been added.
392 				 */
393 				if (timeout >= 3)
394 					timeout >>= 1;
395 
396 				break;
397 			} else if (n < 0) {
398 				/*
399 				 * errno here can only be EINTR which we would
400 				 * want to clean up and bail out.
401 				 */
402 				s = -1;
403 				goto done;
404 			}
405 
406 			/*
407 			 * Check for the event(s) we have seen.
408 			 */
409 			for (j = 0; j < i; j++) {
410 				if (fds[j].fd == -1 || fds[j].events == 0 ||
411 				    fds[j].revents == 0)
412 					continue;
413 				if (fds[j].revents & ~(POLLIN | POLLOUT)) {
414 					close(fds[j].fd);
415 					fds[j].fd = -1;
416 					fds[j].events = 0;
417 					count--;
418 					continue;
419 				} else if (fds[j].revents & (POLLIN | POLLOUT)) {
420 					/* Connect succeeded. */
421 					s = fds[j].fd;
422 					fds[j].fd = -1;
423 
424 					goto done;
425 				}
426 
427 			}
428 		} while (timeout == INFTIM && count != 0);
429 	}
430 
431 	/* All attempts were failed */
432 	s = -1;
433 	if (count == 0)
434 		errno = ETIMEDOUT;
435 
436 done:
437 	/* Close all watched fds except the succeeded one */
438 	for (j = 0; j < i; j++)
439 		if (fds[j].fd != -1)
440 			close(fds[j].fd);
441 	free(fds);
442 	return (s);
443 }
444 
445 static void
446 whois(const char *query, const char *hostname, const char *hostport, int flags)
447 {
448 	FILE *fp;
449 	struct addrinfo *hostres;
450 	char *buf, *host, *nhost, *nport, *p;
451 	int comment, s, f;
452 	size_t len, i;
453 
454 	hostres = gethostinfo(hostname, hostport, 1);
455 	s = connect_to_any_host(hostres);
456 	if (s == -1)
457 		err(EX_OSERR, "connect()");
458 
459 	/* Restore default blocking behavior.  */
460 	if ((f = fcntl(s, F_GETFL)) == -1)
461 		err(EX_OSERR, "fcntl()");
462 	f &= ~O_NONBLOCK;
463 	if (fcntl(s, F_SETFL, f) == -1)
464 		err(EX_OSERR, "fcntl()");
465 
466 	fp = fdopen(s, "r+");
467 	if (fp == NULL)
468 		err(EX_OSERR, "fdopen()");
469 
470 	if (!(flags & WHOIS_SPAM_ME) &&
471 	    (strcasecmp(hostname, DENICHOST) == 0 ||
472 	     strcasecmp(hostname, "de" QNICHOST_TAIL) == 0)) {
473 		const char *q;
474 		int idn = 0;
475 		for (q = query; *q != '\0'; q++)
476 			if (!isascii(*q))
477 				idn = 1;
478 		fprintf(fp, "-T dn%s %s\r\n", idn ? "" : ",ace", query);
479 	} else if (!(flags & WHOIS_SPAM_ME) &&
480 		   (strcasecmp(hostname, DKNICHOST) == 0 ||
481 		    strcasecmp(hostname, "dk" QNICHOST_TAIL) == 0))
482 		fprintf(fp, "--show-handles %s\r\n", query);
483 	else if ((flags & WHOIS_SPAM_ME) ||
484 		 strchr(query, ' ') != NULL)
485 		fprintf(fp, "%s\r\n", query);
486 	else if (strcasecmp(hostname, ANICHOST) == 0) {
487 		if (strncasecmp(query, "AS", 2) == 0 &&
488 		    strspn(query+2, "0123456789") == strlen(query+2))
489 			fprintf(fp, "+ a %s\r\n", query+2);
490 		else
491 			fprintf(fp, "+ %s\r\n", query);
492 	} else if (strcasecmp(hostres->ai_canonname, VNICHOST) == 0)
493 		fprintf(fp, "domain %s\r\n", query);
494 	else
495 		fprintf(fp, "%s\r\n", query);
496 	fflush(fp);
497 
498 	comment = 0;
499 	if (!(flags & WHOIS_SPAM_ME) &&
500 	    (strcasecmp(hostname, ANICHOST) == 0 ||
501 	     strcasecmp(hostname, RNICHOST) == 0)) {
502 		comment = 2;
503 	}
504 
505 	nhost = NULL;
506 	while ((buf = fgetln(fp, &len)) != NULL) {
507 		/* Nominet */
508 		if (!(flags & WHOIS_SPAM_ME) &&
509 		    len == 5 && strncmp(buf, "-- \r\n", 5) == 0)
510 			break;
511 		/* RIRs */
512 		if (comment == 1 && buf[0] == '#')
513 			break;
514 		else if (comment == 2) {
515 			if (strchr("#%\r\n", buf[0]) != NULL)
516 				continue;
517 			else
518 				comment = 1;
519 		}
520 
521 		printf("%.*s", (int)len, buf);
522 
523 		if ((flags & WHOIS_RECURSE) && nhost == NULL) {
524 			for (i = 0; whois_referral[i].prefix != NULL; i++) {
525 				p = buf;
526 				SCAN(p, buf+len, *p == ' ');
527 				if (strncasecmp(p, whois_referral[i].prefix,
528 					           whois_referral[i].len) != 0)
529 					continue;
530 				p += whois_referral[i].len;
531 				SCAN(p, buf+len, *p == ' ');
532 				host = p;
533 				SCAN(p, buf+len, ishost(*p));
534 				if (p > host) {
535 					char *pstr;
536 
537 					s_asprintf(&nhost, "%.*s",
538 						   (int)(p - host), host);
539 
540 					if (*p != ':') {
541 						s_asprintf(&nport, "%s", port);
542 						break;
543 					}
544 
545 					pstr = ++p;
546 					SCAN(p, buf+len, isdigit(*p));
547 					if (p > pstr && (p - pstr) < 6) {
548 						s_asprintf(&nport, "%.*s",
549 						    (int)(p - pstr), pstr);
550 						break;
551 					}
552 
553 					/* Invalid port; don't recurse */
554 					free(nhost);
555 					nhost = NULL;
556 				}
557 				break;
558 			}
559 			for (i = 0; actually_arin[i] != NULL; i++) {
560 				if (strncmp(buf, actually_arin[i], len) == 0) {
561 					s_asprintf(&nhost, "%s", ANICHOST);
562 					s_asprintf(&nport, "%s", port);
563 					break;
564 				}
565 			}
566 		}
567 		/* Verisign etc. */
568 		if (!(flags & WHOIS_SPAM_ME) &&
569 		    len >= sizeof(CHOPSPAM)-1 &&
570 		    (strncasecmp(buf, CHOPSPAM, sizeof(CHOPSPAM)-1) == 0 ||
571 		     strncasecmp(buf, CHOPSPAM+4, sizeof(CHOPSPAM)-5) == 0)) {
572 			printf("\n");
573 			break;
574 		}
575 	}
576 	fclose(fp);
577 	freeaddrinfo(hostres);
578 
579 	f = 0;
580 	for (i = 0; try_rir[i].host != NULL; i++) {
581 		/* Remember visits to RIRs */
582 		if (try_rir[i].loop == 0 &&
583 		    strcasecmp(try_rir[i].host, hostname) == 0)
584 			try_rir[i].loop = 1;
585 		/* Do we need to find an alternative RIR? */
586 		if (try_rir[i].loop != 0 && nhost != NULL &&
587 		    strcasecmp(try_rir[i].host, nhost) == 0) {
588 			free(nhost);
589 			nhost = NULL;
590 			free(nport);
591 			nport = NULL;
592 			f = 1;
593 		}
594 	}
595 	if (f) {
596 		/* Find a replacement RIR */
597 		for (i = 0; try_rir[i].host != NULL; i++) {
598 			if (try_rir[i].loop == 0) {
599 				s_asprintf(&nhost, "%s", try_rir[i].host);
600 				s_asprintf(&nport, "%s", port);
601 				break;
602 			}
603 		}
604 	}
605 	if (nhost != NULL) {
606 		/* Ignore self-referrals */
607 		if (strcasecmp(hostname, nhost) != 0) {
608 			printf("# %s\n\n", nhost);
609 			whois(query, nhost, nport, flags);
610 		}
611 		free(nhost);
612 		free(nport);
613 	}
614 }
615 
616 static void
617 usage(void)
618 {
619 	fprintf(stderr,
620 	    "usage: whois [-aAbfgiIklmPQrRS] [-c country-code | -h hostname] "
621 	    "[-p port] name ...\n");
622 	exit(EX_USAGE);
623 }
624