xref: /freebsd/usr.bin/whois/whois.c (revision 5b31cc94b10d4bb7109c6b27940a0fc76a44a331)
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 #endif
40 
41 #include <sys/cdefs.h>
42 #include <sys/types.h>
43 #include <sys/socket.h>
44 #include <sys/poll.h>
45 #include <netinet/in.h>
46 #include <arpa/inet.h>
47 #include <ctype.h>
48 #include <err.h>
49 #include <netdb.h>
50 #include <stdarg.h>
51 #include <stdio.h>
52 #include <stdlib.h>
53 #include <string.h>
54 #include <sysexits.h>
55 #include <unistd.h>
56 #include <fcntl.h>
57 #include <errno.h>
58 
59 #define	ABUSEHOST	"whois.abuse.net"
60 #define	ANICHOST	"whois.arin.net"
61 #define	DENICHOST	"whois.denic.de"
62 #define	DKNICHOST	"whois.dk-hostmaster.dk"
63 #define	FNICHOST	"whois.afrinic.net"
64 #define	GNICHOST	"whois.nic.gov"
65 #define	IANAHOST	"whois.iana.org"
66 #define	INICHOST	"whois.internic.net"
67 #define	KNICHOST	"whois.krnic.net"
68 #define	LNICHOST	"whois.lacnic.net"
69 #define	MNICHOST	"whois.ra.net"
70 #define	PDBHOST		"whois.peeringdb.com"
71 #define	PNICHOST	"whois.apnic.net"
72 #define	QNICHOST_TAIL	".whois-servers.net"
73 #define	RNICHOST	"whois.ripe.net"
74 #define	VNICHOST	"whois.verisign-grs.com"
75 
76 #define	DEFAULT_PORT	"whois"
77 
78 #define WHOIS_RECURSE	0x01
79 #define WHOIS_QUICK	0x02
80 #define WHOIS_SPAM_ME	0x04
81 
82 #define CHOPSPAM	">>> Last update of WHOIS database:"
83 
84 #define ishost(h) (isalnum((unsigned char)h) || h == '.' || h == '-')
85 
86 #define SCAN(p, end, check)					\
87 	while ((p) < (end))					\
88 		if (check) ++(p);				\
89 		else break
90 
91 static struct {
92 	const char *suffix, *server;
93 } whoiswhere[] = {
94 	/* Various handles */
95 	{ "-ARIN", ANICHOST },
96 	{ "-NICAT", "at" QNICHOST_TAIL },
97 	{ "-NORID", "no" QNICHOST_TAIL },
98 	{ "-RIPE", RNICHOST },
99 	/* Nominet's whois server doesn't return referrals to JANET */
100 	{ ".ac.uk", "ac.uk" QNICHOST_TAIL },
101 	{ ".gov.uk", "ac.uk" QNICHOST_TAIL },
102 	{ "", IANAHOST }, /* default */
103 	{ NULL, NULL } /* safety belt */
104 };
105 
106 #define WHOIS_REFERRAL(s) { s, sizeof(s) - 1 }
107 static struct {
108 	const char *prefix;
109 	size_t len;
110 } whois_referral[] = {
111 	WHOIS_REFERRAL("whois:"), /* IANA */
112 	WHOIS_REFERRAL("Whois Server:"),
113 	WHOIS_REFERRAL("Registrar WHOIS Server:"), /* corporatedomains.com */
114 	WHOIS_REFERRAL("ReferralServer:  whois://"), /* ARIN */
115 	WHOIS_REFERRAL("ReferralServer:  rwhois://"), /* ARIN */
116 	WHOIS_REFERRAL("descr:          region. Please query"), /* AfriNIC */
117 	{ NULL, 0 }
118 };
119 
120 /*
121  * We have a list of patterns for RIRs that assert ignorance rather than
122  * providing referrals. If that happens, we guess that ARIN will be more
123  * helpful. But, before following a referral to an RIR, we check if we have
124  * asked that RIR already, and if so we make another guess.
125  */
126 static const char *actually_arin[] = {
127 	"netname:        ERX-NETBLOCK\n", /* APNIC */
128 	"netname:        NON-RIPE-NCC-MANAGED-ADDRESS-BLOCK\n",
129 	NULL
130 };
131 
132 static struct {
133 	int loop;
134 	const char *host;
135 } try_rir[] = {
136 	{ 0, ANICHOST },
137 	{ 0, RNICHOST },
138 	{ 0, PNICHOST },
139 	{ 0, FNICHOST },
140 	{ 0, LNICHOST },
141 	{ 0, NULL }
142 };
143 
144 static void
145 reset_rir(void) {
146 	int i;
147 
148 	for (i = 0; try_rir[i].host != NULL; i++)
149 		try_rir[i].loop = 0;
150 }
151 
152 static const char *port = DEFAULT_PORT;
153 
154 static const char *choose_server(char *);
155 static struct addrinfo *gethostinfo(const char *, const char *, int);
156 static void s_asprintf(char **ret, const char *format, ...) __printflike(2, 3);
157 static void usage(void) __dead2;
158 static void whois(const char *, const char *, const char *, int);
159 
160 int
161 main(int argc, char *argv[])
162 {
163 	const char *country, *host;
164 	int ch, flags;
165 
166 #ifdef	SOCKS
167 	SOCKSinit(argv[0]);
168 #endif
169 
170 	country = host = NULL;
171 	flags = 0;
172 	while ((ch = getopt(argc, argv, "aAbc:fgh:iIklmp:PQrRS")) != -1) {
173 		switch (ch) {
174 		case 'a':
175 			host = ANICHOST;
176 			break;
177 		case 'A':
178 			host = PNICHOST;
179 			break;
180 		case 'b':
181 			host = ABUSEHOST;
182 			break;
183 		case 'c':
184 			country = optarg;
185 			break;
186 		case 'f':
187 			host = FNICHOST;
188 			break;
189 		case 'g':
190 			host = GNICHOST;
191 			break;
192 		case 'h':
193 			host = optarg;
194 			break;
195 		case 'i':
196 			host = INICHOST;
197 			break;
198 		case 'I':
199 			host = IANAHOST;
200 			break;
201 		case 'k':
202 			host = KNICHOST;
203 			break;
204 		case 'l':
205 			host = LNICHOST;
206 			break;
207 		case 'm':
208 			host = MNICHOST;
209 			break;
210 		case 'p':
211 			port = optarg;
212 			break;
213 		case 'P':
214 			host = PDBHOST;
215 			break;
216 		case 'Q':
217 			flags |= WHOIS_QUICK;
218 			break;
219 		case 'r':
220 			host = RNICHOST;
221 			break;
222 		case 'R':
223 			flags |= WHOIS_RECURSE;
224 			break;
225 		case 'S':
226 			flags |= WHOIS_SPAM_ME;
227 			break;
228 		case '?':
229 		default:
230 			usage();
231 			/* NOTREACHED */
232 		}
233 	}
234 	argc -= optind;
235 	argv += optind;
236 
237 	if (!argc || (country != NULL && host != NULL))
238 		usage();
239 
240 	/*
241 	 * If no host or country is specified, rely on referrals from IANA.
242 	 */
243 	if (host == NULL && country == NULL) {
244 		if ((host = getenv("WHOIS_SERVER")) == NULL &&
245 		    (host = getenv("RA_SERVER")) == NULL) {
246 			if (!(flags & WHOIS_QUICK))
247 				flags |= WHOIS_RECURSE;
248 		}
249 	}
250 	while (argc-- > 0) {
251 		if (country != NULL) {
252 			char *qnichost;
253 			s_asprintf(&qnichost, "%s%s", country, QNICHOST_TAIL);
254 			whois(*argv, qnichost, port, flags);
255 			free(qnichost);
256 		} else
257 			whois(*argv, host != NULL ? host :
258 			      choose_server(*argv), port, flags);
259 		reset_rir();
260 		argv++;
261 	}
262 	exit(0);
263 }
264 
265 static const char *
266 choose_server(char *domain)
267 {
268 	size_t len = strlen(domain);
269 	int i;
270 
271 	for (i = 0; whoiswhere[i].suffix != NULL; i++) {
272 		size_t suffix_len = strlen(whoiswhere[i].suffix);
273 		if (len > suffix_len &&
274 		    strcasecmp(domain + len - suffix_len,
275 			       whoiswhere[i].suffix) == 0)
276 			return (whoiswhere[i].server);
277 	}
278 	errx(EX_SOFTWARE, "no default whois server");
279 }
280 
281 static struct addrinfo *
282 gethostinfo(const char *host, const char *hport, int exit_on_noname)
283 {
284 	struct addrinfo hints, *res;
285 	int error;
286 
287 	memset(&hints, 0, sizeof(hints));
288 	hints.ai_flags = AI_CANONNAME;
289 	hints.ai_family = AF_UNSPEC;
290 	hints.ai_socktype = SOCK_STREAM;
291 	res = NULL;
292 	error = getaddrinfo(host, hport, &hints, &res);
293 	if (error && (exit_on_noname || error != EAI_NONAME))
294 		err(EX_NOHOST, "%s: %s", host, gai_strerror(error));
295 	return (res);
296 }
297 
298 /*
299  * Wrapper for asprintf(3) that exits on error.
300  */
301 static void
302 s_asprintf(char **ret, const char *format, ...)
303 {
304 	va_list ap;
305 
306 	va_start(ap, format);
307 	if (vasprintf(ret, format, ap) == -1) {
308 		va_end(ap);
309 		err(EX_OSERR, "vasprintf()");
310 	}
311 	va_end(ap);
312 }
313 
314 static int
315 connect_to_any_host(struct addrinfo *hostres)
316 {
317 	struct addrinfo *res;
318 	nfds_t i, j;
319 	size_t count;
320 	struct pollfd *fds;
321 	int timeout = 180, s = -1;
322 
323 	for (res = hostres, count = 0; res; res = res->ai_next)
324 		count++;
325 	fds = calloc(count, sizeof(*fds));
326 	if (fds == NULL)
327 		err(EX_OSERR, "calloc()");
328 
329 	/*
330 	 * Traverse the result list elements and make non-block
331 	 * connection attempts.
332 	 */
333 	count = i = 0;
334 	for (res = hostres; res != NULL; res = res->ai_next) {
335 		s = socket(res->ai_family, res->ai_socktype | SOCK_NONBLOCK,
336 		    res->ai_protocol);
337 		if (s < 0)
338 			continue;
339 		if (connect(s, res->ai_addr, res->ai_addrlen) < 0) {
340 			if (errno == EINPROGRESS) {
341 				/* Add the socket to poll list */
342 				fds[i].fd = s;
343 				fds[i].events = POLLERR | POLLHUP |
344 						POLLIN | POLLOUT;
345 				/*
346 				 * From here until a socket connects, the
347 				 * socket fd is owned by the fds[] poll array.
348 				 */
349 				s = -1;
350 				count++;
351 				i++;
352 			} else {
353 				close(s);
354 				s = -1;
355 
356 				/*
357 				 * Poll only if we have something to poll,
358 				 * otherwise just go ahead and try next
359 				 * address
360 				 */
361 				if (count == 0)
362 					continue;
363 			}
364 		} else
365 			goto done;
366 
367 		/*
368 		 * If we are at the last address, poll until a connection is
369 		 * established or we failed all connection attempts.
370 		 */
371 		if (res->ai_next == NULL)
372 			timeout = INFTIM;
373 
374 		/*
375 		 * Poll the watched descriptors for successful connections:
376 		 * if we still have more untried resolved addresses, poll only
377 		 * once; otherwise, poll until all descriptors have errors,
378 		 * which will be considered as ETIMEDOUT later.
379 		 */
380 		do {
381 			int n;
382 
383 			n = poll(fds, i, timeout);
384 			if (n == 0) {
385 				/*
386 				 * No event reported in time.  Try with a
387 				 * smaller timeout (but cap at 2-3ms)
388 				 * after a new host have been added.
389 				 */
390 				if (timeout >= 3)
391 					timeout >>= 1;
392 
393 				break;
394 			} else if (n < 0) {
395 				/*
396 				 * errno here can only be EINTR which we would
397 				 * want to clean up and bail out.
398 				 */
399 				s = -1;
400 				goto done;
401 			}
402 
403 			/*
404 			 * Check for the event(s) we have seen.
405 			 */
406 			for (j = 0; j < i; j++) {
407 				if (fds[j].fd == -1 || fds[j].events == 0 ||
408 				    fds[j].revents == 0)
409 					continue;
410 				if (fds[j].revents & ~(POLLIN | POLLOUT)) {
411 					close(fds[j].fd);
412 					fds[j].fd = -1;
413 					fds[j].events = 0;
414 					count--;
415 					continue;
416 				} else if (fds[j].revents & (POLLIN | POLLOUT)) {
417 					/* Connect succeeded. */
418 					s = fds[j].fd;
419 					fds[j].fd = -1;
420 
421 					goto done;
422 				}
423 
424 			}
425 		} while (timeout == INFTIM && count != 0);
426 	}
427 
428 	/* All attempts were failed */
429 	s = -1;
430 	if (count == 0)
431 		errno = ETIMEDOUT;
432 
433 done:
434 	/* Close all watched fds except the succeeded one */
435 	for (j = 0; j < i; j++)
436 		if (fds[j].fd != -1)
437 			close(fds[j].fd);
438 	free(fds);
439 	return (s);
440 }
441 
442 static void
443 whois(const char *query, const char *hostname, const char *hostport, int flags)
444 {
445 	FILE *fp;
446 	struct addrinfo *hostres;
447 	char *buf, *host, *nhost, *nport, *p;
448 	int comment, s, f;
449 	size_t len, i;
450 
451 	hostres = gethostinfo(hostname, hostport, 1);
452 	s = connect_to_any_host(hostres);
453 	if (s == -1)
454 		err(EX_OSERR, "connect()");
455 
456 	/* Restore default blocking behavior.  */
457 	if ((f = fcntl(s, F_GETFL)) == -1)
458 		err(EX_OSERR, "fcntl()");
459 	f &= ~O_NONBLOCK;
460 	if (fcntl(s, F_SETFL, f) == -1)
461 		err(EX_OSERR, "fcntl()");
462 
463 	fp = fdopen(s, "r+");
464 	if (fp == NULL)
465 		err(EX_OSERR, "fdopen()");
466 
467 	if (!(flags & WHOIS_SPAM_ME) &&
468 	    (strcasecmp(hostname, DENICHOST) == 0 ||
469 	     strcasecmp(hostname, "de" QNICHOST_TAIL) == 0)) {
470 		const char *q;
471 		int idn = 0;
472 		for (q = query; *q != '\0'; q++)
473 			if (!isascii(*q))
474 				idn = 1;
475 		fprintf(fp, "-T dn%s %s\r\n", idn ? "" : ",ace", query);
476 	} else if (!(flags & WHOIS_SPAM_ME) &&
477 		   (strcasecmp(hostname, DKNICHOST) == 0 ||
478 		    strcasecmp(hostname, "dk" QNICHOST_TAIL) == 0))
479 		fprintf(fp, "--show-handles %s\r\n", query);
480 	else if ((flags & WHOIS_SPAM_ME) ||
481 		 strchr(query, ' ') != NULL)
482 		fprintf(fp, "%s\r\n", query);
483 	else if (strcasecmp(hostname, ANICHOST) == 0) {
484 		if (strncasecmp(query, "AS", 2) == 0 &&
485 		    strspn(query+2, "0123456789") == strlen(query+2))
486 			fprintf(fp, "+ a %s\r\n", query+2);
487 		else
488 			fprintf(fp, "+ %s\r\n", query);
489 	} else if (strcasecmp(hostres->ai_canonname, VNICHOST) == 0)
490 		fprintf(fp, "domain %s\r\n", query);
491 	else
492 		fprintf(fp, "%s\r\n", query);
493 	fflush(fp);
494 
495 	comment = 0;
496 	if (!(flags & WHOIS_SPAM_ME) &&
497 	    (strcasecmp(hostname, ANICHOST) == 0 ||
498 	     strcasecmp(hostname, RNICHOST) == 0)) {
499 		comment = 2;
500 	}
501 
502 	nhost = NULL;
503 	while ((buf = fgetln(fp, &len)) != NULL) {
504 		/* Nominet */
505 		if (!(flags & WHOIS_SPAM_ME) &&
506 		    len == 5 && strncmp(buf, "-- \r\n", 5) == 0)
507 			break;
508 		/* RIRs */
509 		if (comment == 1 && buf[0] == '#')
510 			break;
511 		else if (comment == 2) {
512 			if (strchr("#%\r\n", buf[0]) != NULL)
513 				continue;
514 			else
515 				comment = 1;
516 		}
517 
518 		printf("%.*s", (int)len, buf);
519 
520 		if ((flags & WHOIS_RECURSE) && nhost == NULL) {
521 			for (i = 0; whois_referral[i].prefix != NULL; i++) {
522 				p = buf;
523 				SCAN(p, buf+len, *p == ' ');
524 				if (strncasecmp(p, whois_referral[i].prefix,
525 					           whois_referral[i].len) != 0)
526 					continue;
527 				p += whois_referral[i].len;
528 				SCAN(p, buf+len, *p == ' ');
529 				host = p;
530 				SCAN(p, buf+len, ishost(*p));
531 				if (p > host) {
532 					char *pstr;
533 
534 					s_asprintf(&nhost, "%.*s",
535 						   (int)(p - host), host);
536 
537 					if (*p != ':') {
538 						s_asprintf(&nport, "%s", port);
539 						break;
540 					}
541 
542 					pstr = ++p;
543 					SCAN(p, buf+len, isdigit(*p));
544 					if (p > pstr && (p - pstr) < 6) {
545 						s_asprintf(&nport, "%.*s",
546 						    (int)(p - pstr), pstr);
547 						break;
548 					}
549 
550 					/* Invalid port; don't recurse */
551 					free(nhost);
552 					nhost = NULL;
553 				}
554 				break;
555 			}
556 			for (i = 0; actually_arin[i] != NULL; i++) {
557 				if (strncmp(buf, actually_arin[i], len) == 0) {
558 					s_asprintf(&nhost, "%s", ANICHOST);
559 					s_asprintf(&nport, "%s", port);
560 					break;
561 				}
562 			}
563 		}
564 		/* Verisign etc. */
565 		if (!(flags & WHOIS_SPAM_ME) &&
566 		    len >= sizeof(CHOPSPAM)-1 &&
567 		    (strncasecmp(buf, CHOPSPAM, sizeof(CHOPSPAM)-1) == 0 ||
568 		     strncasecmp(buf, CHOPSPAM+4, sizeof(CHOPSPAM)-5) == 0)) {
569 			printf("\n");
570 			break;
571 		}
572 	}
573 	fclose(fp);
574 	freeaddrinfo(hostres);
575 
576 	f = 0;
577 	for (i = 0; try_rir[i].host != NULL; i++) {
578 		/* Remember visits to RIRs */
579 		if (try_rir[i].loop == 0 &&
580 		    strcasecmp(try_rir[i].host, hostname) == 0)
581 			try_rir[i].loop = 1;
582 		/* Do we need to find an alternative RIR? */
583 		if (try_rir[i].loop != 0 && nhost != NULL &&
584 		    strcasecmp(try_rir[i].host, nhost) == 0) {
585 			free(nhost);
586 			nhost = NULL;
587 			free(nport);
588 			nport = NULL;
589 			f = 1;
590 		}
591 	}
592 	if (f) {
593 		/* Find a replacement RIR */
594 		for (i = 0; try_rir[i].host != NULL; i++) {
595 			if (try_rir[i].loop == 0) {
596 				s_asprintf(&nhost, "%s", try_rir[i].host);
597 				s_asprintf(&nport, "%s", port);
598 				break;
599 			}
600 		}
601 	}
602 	if (nhost != NULL) {
603 		/* Ignore self-referrals */
604 		if (strcasecmp(hostname, nhost) != 0) {
605 			printf("# %s\n\n", nhost);
606 			whois(query, nhost, nport, flags);
607 		}
608 		free(nhost);
609 		free(nport);
610 	}
611 }
612 
613 static void
614 usage(void)
615 {
616 	fprintf(stderr,
617 	    "usage: whois [-aAbfgiIklmPQrRS] [-c country-code | -h hostname] "
618 	    "[-p port] name ...\n");
619 	exit(EX_USAGE);
620 }
621