xref: /openbsd-src/lib/libc/net/res_comp.c (revision db3296cf5c1dd9058ceecc3a29fe4aaa0bd26000)
1 /*	$OpenBSD: res_comp.c,v 1.11 2003/06/02 20:18:36 millert Exp $	*/
2 
3 /*
4  * ++Copyright++ 1985, 1993
5  * -
6  * Copyright (c) 1985, 1993
7  *    The Regents of the University of California.  All rights reserved.
8  *
9  * Redistribution and use in source and binary forms, with or without
10  * modification, are permitted provided that the following conditions
11  * are met:
12  * 1. Redistributions of source code must retain the above copyright
13  *    notice, this list of conditions and the following disclaimer.
14  * 2. Redistributions in binary form must reproduce the above copyright
15  *    notice, this list of conditions and the following disclaimer in the
16  *    documentation and/or other materials provided with the distribution.
17  * 3. Neither the name of the University nor the names of its contributors
18  *    may be used to endorse or promote products derived from this software
19  *    without specific prior written permission.
20  *
21  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
22  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
23  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
24  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
25  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
26  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
27  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
28  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
29  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
30  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
31  * SUCH DAMAGE.
32  * -
33  * Portions Copyright (c) 1993 by Digital Equipment Corporation.
34  *
35  * Permission to use, copy, modify, and distribute this software for any
36  * purpose with or without fee is hereby granted, provided that the above
37  * copyright notice and this permission notice appear in all copies, and that
38  * the name of Digital Equipment Corporation not be used in advertising or
39  * publicity pertaining to distribution of the document or software without
40  * specific, written prior permission.
41  *
42  * THE SOFTWARE IS PROVIDED "AS IS" AND DIGITAL EQUIPMENT CORP. DISCLAIMS ALL
43  * WARRANTIES WITH REGARD TO THIS SOFTWARE, INCLUDING ALL IMPLIED WARRANTIES
44  * OF MERCHANTABILITY AND FITNESS.   IN NO EVENT SHALL DIGITAL EQUIPMENT
45  * CORPORATION BE LIABLE FOR ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL
46  * DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR
47  * PROFITS, WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS
48  * ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS
49  * SOFTWARE.
50  * -
51  * --Copyright--
52  */
53 
54 #if defined(LIBC_SCCS) && !defined(lint)
55 #if 0
56 static char sccsid[] = "@(#)res_comp.c	8.1 (Berkeley) 6/4/93";
57 static char rcsid[] = "$From: res_comp.c,v 8.11 1996/12/02 09:17:22 vixie Exp $";
58 #else
59 static char rcsid[] = "$OpenBSD: res_comp.c,v 1.11 2003/06/02 20:18:36 millert Exp $";
60 #endif
61 #endif /* LIBC_SCCS and not lint */
62 
63 #include <sys/types.h>
64 #include <sys/param.h>
65 #include <netinet/in.h>
66 #include <arpa/nameser.h>
67 
68 #include <stdio.h>
69 #include <resolv.h>
70 #include <ctype.h>
71 
72 #include <unistd.h>
73 #include <string.h>
74 
75 static int dn_find(u_char *, u_char *, u_char **, u_char **);
76 
77 /*
78  * Expand compressed domain name 'comp_dn' to full domain name.
79  * 'msg' is a pointer to the begining of the message,
80  * 'eomorig' points to the first location after the message,
81  * 'exp_dn' is a pointer to a buffer of size 'length' for the result.
82  * Return size of compressed name or -1 if there was an error.
83  */
84 int
85 dn_expand(msg, eomorig, comp_dn, exp_dn, length)
86 	const u_char *msg, *eomorig, *comp_dn;
87 	char *exp_dn;
88 	int length;
89 {
90 	register const u_char *cp;
91 	register char *dn;
92 	register int n, c;
93 	char *eom;
94 	int len = -1, checked = 0;
95 
96 	dn = exp_dn;
97 	cp = comp_dn;
98 	if (length > MAXHOSTNAMELEN-1)
99 		length = MAXHOSTNAMELEN-1;
100 	eom = exp_dn + length;
101 	/*
102 	 * fetch next label in domain name
103 	 */
104 	while ((n = *cp++)) {
105 		/*
106 		 * Check for indirection
107 		 */
108 		switch (n & INDIR_MASK) {
109 		case 0:
110 			if (dn != exp_dn) {
111 				if (dn >= eom)
112 					return (-1);
113 				*dn++ = '.';
114 			}
115 			if (dn+n >= eom)
116 				return (-1);
117 			checked += n + 1;
118 			while (--n >= 0) {
119 				if (((c = *cp++) == '.') || (c == '\\')) {
120 					if (dn + n + 2 >= eom)
121 						return (-1);
122 					*dn++ = '\\';
123 				}
124 				*dn++ = c;
125 				if (cp >= eomorig)	/* out of range */
126 					return (-1);
127 			}
128 			break;
129 
130 		case INDIR_MASK:
131 			if (len < 0)
132 				len = cp - comp_dn + 1;
133 			cp = msg + (((n & 0x3f) << 8) | (*cp & 0xff));
134 			if (cp < msg || cp >= eomorig)	/* out of range */
135 				return (-1);
136 			checked += 2;
137 			/*
138 			 * Check for loops in the compressed name;
139 			 * if we've looked at the whole message,
140 			 * there must be a loop.
141 			 */
142 			if (checked >= eomorig - msg)
143 				return (-1);
144 			break;
145 
146 		default:
147 			return (-1);			/* flag error */
148 		}
149 	}
150 	*dn = '\0';
151 	if (len < 0)
152 		len = cp - comp_dn;
153 	return (len);
154 }
155 
156 /*
157  * Compress domain name 'exp_dn' into 'comp_dn'.
158  * Return the size of the compressed name or -1.
159  * 'length' is the size of the array pointed to by 'comp_dn'.
160  * 'dnptrs' is a list of pointers to previous compressed names. dnptrs[0]
161  * is a pointer to the beginning of the message. The list ends with NULL.
162  * 'lastdnptr' is a pointer to the end of the arrary pointed to
163  * by 'dnptrs'. Side effect is to update the list of pointers for
164  * labels inserted into the message as we compress the name.
165  * If 'dnptr' is NULL, we don't try to compress names. If 'lastdnptr'
166  * is NULL, we don't update the list.
167  */
168 int
169 dn_comp(exp_dn, comp_dn, length, dnptrs, lastdnptr)
170 	const char *exp_dn;
171 	u_char *comp_dn, **dnptrs, **lastdnptr;
172 	int length;
173 {
174 	register u_char *cp, *dn;
175 	register int c, l;
176 	u_char **cpp, **lpp, *sp, *eob;
177 	u_char *msg;
178 
179 	dn = (u_char *)exp_dn;
180 	cp = comp_dn;
181 	eob = cp + length;
182 	lpp = cpp = NULL;
183 	if (dnptrs != NULL) {
184 		if ((msg = *dnptrs++) != NULL) {
185 			for (cpp = dnptrs; *cpp != NULL; cpp++)
186 				;
187 			lpp = cpp;	/* end of list to search */
188 		}
189 	} else
190 		msg = NULL;
191 	for (c = *dn++; c != '\0'; ) {
192 		/* look to see if we can use pointers */
193 		if (msg != NULL) {
194 			if ((l = dn_find(dn-1, msg, dnptrs, lpp)) >= 0) {
195 				if (cp+1 >= eob)
196 					return (-1);
197 				*cp++ = (l >> 8) | INDIR_MASK;
198 				*cp++ = l % 256;
199 				return (cp - comp_dn);
200 			}
201 			/* not found, save it */
202 			if (lastdnptr != NULL && cpp < lastdnptr-1) {
203 				*cpp++ = cp;
204 				*cpp = NULL;
205 			}
206 		}
207 		sp = cp++;	/* save ptr to length byte */
208 		do {
209 			if (c == '.') {
210 				c = *dn++;
211 				break;
212 			}
213 			if (c == '\\') {
214 				if ((c = *dn++) == '\0')
215 					break;
216 			}
217 			if (cp >= eob) {
218 				if (msg != NULL)
219 					*lpp = NULL;
220 				return (-1);
221 			}
222 			*cp++ = c;
223 		} while ((c = *dn++) != '\0');
224 		/* catch trailing '.'s but not '..' */
225 		if ((l = cp - sp - 1) == 0 && c == '\0') {
226 			cp--;
227 			break;
228 		}
229 		if (l <= 0 || l > MAXLABEL) {
230 			if (msg != NULL)
231 				*lpp = NULL;
232 			return (-1);
233 		}
234 		*sp = l;
235 	}
236 	if (cp >= eob) {
237 		if (msg != NULL)
238 			*lpp = NULL;
239 		return (-1);
240 	}
241 	*cp++ = '\0';
242 	return (cp - comp_dn);
243 }
244 
245 /*
246  * Skip over a compressed domain name. Return the size or -1.
247  */
248 int
249 __dn_skipname(comp_dn, eom)
250 	const u_char *comp_dn, *eom;
251 {
252 	register const u_char *cp;
253 	register int n;
254 
255 	cp = comp_dn;
256 	while (cp < eom && (n = *cp++)) {
257 		/*
258 		 * check for indirection
259 		 */
260 		switch (n & INDIR_MASK) {
261 		case 0:			/* normal case, n == len */
262 			cp += n;
263 			continue;
264 		case INDIR_MASK:	/* indirection */
265 			cp++;
266 			break;
267 		default:		/* illegal type */
268 			return (-1);
269 		}
270 		break;
271 	}
272 	if (cp > eom)
273 		return (-1);
274 	return (cp - comp_dn);
275 }
276 
277 static int
278 mklower(ch)
279 	register int ch;
280 {
281 	if (isascii(ch) && isupper(ch))
282 		return (tolower(ch));
283 	return (ch);
284 }
285 
286 /*
287  * Search for expanded name from a list of previously compressed names.
288  * Return the offset from msg if found or -1.
289  * dnptrs is the pointer to the first name on the list,
290  * not the pointer to the start of the message.
291  */
292 static int
293 dn_find(exp_dn, msg, dnptrs, lastdnptr)
294 	u_char *exp_dn, *msg;
295 	u_char **dnptrs, **lastdnptr;
296 {
297 	register u_char *dn, *cp, **cpp;
298 	register int n;
299 	u_char *sp;
300 
301 	for (cpp = dnptrs; cpp < lastdnptr; cpp++) {
302 		dn = exp_dn;
303 		sp = cp = *cpp;
304 		while ((n = *cp++)) {
305 			/*
306 			 * check for indirection
307 			 */
308 			switch (n & INDIR_MASK) {
309 			case 0:		/* normal case, n == len */
310 				while (--n >= 0) {
311 					if (*dn == '.')
312 						goto next;
313 					if (*dn == '\\')
314 						dn++;
315 					if (mklower(*dn++) != mklower(*cp++))
316 						goto next;
317 				}
318 				if ((n = *dn++) == '\0' && *cp == '\0')
319 					return (sp - msg);
320 				if (n == '.')
321 					continue;
322 				goto next;
323 
324 			case INDIR_MASK:	/* indirection */
325 				cp = msg + (((n & 0x3f) << 8) | *cp);
326 				break;
327 
328 			default:	/* illegal type */
329 				return (-1);
330 			}
331 		}
332 		if (*dn == '\0')
333 			return (sp - msg);
334 	next:	;
335 	}
336 	return (-1);
337 }
338 
339 /*
340  * Verify that a domain name uses an acceptable character set.
341  */
342 
343 /*
344  * Note the conspicuous absence of ctype macros in these definitions.  On
345  * non-ASCII hosts, we can't depend on string literals or ctype macros to
346  * tell us anything about network-format data.  The rest of the BIND system
347  * is not careful about this, but for some reason, we're doing it right here.
348  */
349 #define PERIOD 0x2e
350 #define	hyphenchar(c) ((c) == 0x2d)
351 #define bslashchar(c) ((c) == 0x5c)
352 #define periodchar(c) ((c) == PERIOD)
353 #define asterchar(c) ((c) == 0x2a)
354 #define alphachar(c) (((c) >= 0x41 && (c) <= 0x5a) \
355 		   || ((c) >= 0x61 && (c) <= 0x7a))
356 #define digitchar(c) ((c) >= 0x30 && (c) <= 0x39)
357 
358 #define borderchar(c) (alphachar(c) || digitchar(c))
359 #define middlechar(c) (borderchar(c) || hyphenchar(c))
360 #define	domainchar(c) ((c) > 0x20 && (c) < 0x7f)
361 
362 int
363 res_hnok(dn)
364 	const char *dn;
365 {
366 	int pch = PERIOD, ch = *dn++;
367 
368 	while (ch != '\0') {
369 		int nch = *dn++;
370 
371 		if (periodchar(ch)) {
372 			;
373 		} else if (periodchar(pch)) {
374 			if (!borderchar(ch))
375 				return (0);
376 		} else if (periodchar(nch) || nch == '\0') {
377 			if (!borderchar(ch))
378 				return (0);
379 		} else {
380 			if (!middlechar(ch))
381 				return (0);
382 		}
383 		pch = ch, ch = nch;
384 	}
385 	return (1);
386 }
387 
388 /*
389  * hostname-like (A, MX, WKS) owners can have "*" as their first label
390  * but must otherwise be as a host name.
391  */
392 int
393 res_ownok(dn)
394 	const char *dn;
395 {
396 	if (asterchar(dn[0])) {
397 		if (periodchar(dn[1]))
398 			return (res_hnok(dn+2));
399 		if (dn[1] == '\0')
400 			return (1);
401 	}
402 	return (res_hnok(dn));
403 }
404 
405 /*
406  * SOA RNAMEs and RP RNAMEs can have any printable character in their first
407  * label, but the rest of the name has to look like a host name.
408  */
409 int
410 res_mailok(dn)
411 	const char *dn;
412 {
413 	int ch, escaped = 0;
414 
415 	/* "." is a valid missing representation */
416 	if (*dn == '\0')
417 		return(1);
418 
419 	/* otherwise <label>.<hostname> */
420 	while ((ch = *dn++) != '\0') {
421 		if (!domainchar(ch))
422 			return (0);
423 		if (!escaped && periodchar(ch))
424 			break;
425 		if (escaped)
426 			escaped = 0;
427 		else if (bslashchar(ch))
428 			escaped = 1;
429 	}
430 	if (periodchar(ch))
431 		return (res_hnok(dn));
432 	return(0);
433 }
434 
435 /*
436  * This function is quite liberal, since RFC 1034's character sets are only
437  * recommendations.
438  */
439 int
440 res_dnok(dn)
441 	const char *dn;
442 {
443 	int ch;
444 
445 	while ((ch = *dn++) != '\0')
446 		if (!domainchar(ch))
447 			return (0);
448 	return (1);
449 }
450 
451 /*
452  * Routines to insert/extract short/long's.
453  */
454 
455 u_int16_t
456 _getshort(msgp)
457 	register const u_char *msgp;
458 {
459 	register u_int16_t u;
460 
461 	GETSHORT(u, msgp);
462 	return (u);
463 }
464 
465 #ifdef NeXT
466 /*
467  * nExt machines have some funky library conventions, which we must maintain.
468  */
469 u_int16_t
470 res_getshort(msgp)
471 	register const u_char *msgp;
472 {
473 	return (_getshort(msgp));
474 }
475 #endif
476 
477 u_int32_t
478 _getlong(msgp)
479 	register const u_char *msgp;
480 {
481 	register u_int32_t u;
482 
483 	GETLONG(u, msgp);
484 	return (u);
485 }
486 
487 void
488 __putshort(register u_int16_t s, register u_char *msgp)
489 {
490 	PUTSHORT(s, msgp);
491 }
492 
493 void
494 __putlong(l, msgp)
495 	register u_int32_t l;
496 	register u_char *msgp;
497 {
498 	PUTLONG(l, msgp);
499 }
500