xref: /openbsd-src/lib/libc/net/res_comp.c (revision 33b792a3c1c87b47219fdf9a73548c4003214de3)
1 /*	$OpenBSD: res_comp.c,v 1.9 2002/02/17 19:42:23 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. All advertising materials mentioning features or use of this software
18  *    must display the following acknowledgement:
19  * 	This product includes software developed by the University of
20  * 	California, Berkeley and its contributors.
21  * 4. Neither the name of the University nor the names of its contributors
22  *    may be used to endorse or promote products derived from this software
23  *    without specific prior written permission.
24  *
25  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
26  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
27  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
28  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
29  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
30  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
31  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
32  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
33  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
34  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
35  * SUCH DAMAGE.
36  * -
37  * Portions Copyright (c) 1993 by Digital Equipment Corporation.
38  *
39  * Permission to use, copy, modify, and distribute this software for any
40  * purpose with or without fee is hereby granted, provided that the above
41  * copyright notice and this permission notice appear in all copies, and that
42  * the name of Digital Equipment Corporation not be used in advertising or
43  * publicity pertaining to distribution of the document or software without
44  * specific, written prior permission.
45  *
46  * THE SOFTWARE IS PROVIDED "AS IS" AND DIGITAL EQUIPMENT CORP. DISCLAIMS ALL
47  * WARRANTIES WITH REGARD TO THIS SOFTWARE, INCLUDING ALL IMPLIED WARRANTIES
48  * OF MERCHANTABILITY AND FITNESS.   IN NO EVENT SHALL DIGITAL EQUIPMENT
49  * CORPORATION BE LIABLE FOR ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL
50  * DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR
51  * PROFITS, WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS
52  * ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS
53  * SOFTWARE.
54  * -
55  * --Copyright--
56  */
57 
58 #if defined(LIBC_SCCS) && !defined(lint)
59 #if 0
60 static char sccsid[] = "@(#)res_comp.c	8.1 (Berkeley) 6/4/93";
61 static char rcsid[] = "$From: res_comp.c,v 8.11 1996/12/02 09:17:22 vixie Exp $";
62 #else
63 static char rcsid[] = "$OpenBSD: res_comp.c,v 1.9 2002/02/17 19:42:23 millert Exp $";
64 #endif
65 #endif /* LIBC_SCCS and not lint */
66 
67 #include <sys/types.h>
68 #include <sys/param.h>
69 #include <netinet/in.h>
70 #include <arpa/nameser.h>
71 
72 #include <stdio.h>
73 #include <resolv.h>
74 #include <ctype.h>
75 
76 #include <unistd.h>
77 #include <string.h>
78 
79 static int dn_find(u_char *, u_char *, u_char **, u_char **);
80 
81 /*
82  * Expand compressed domain name 'comp_dn' to full domain name.
83  * 'msg' is a pointer to the begining of the message,
84  * 'eomorig' points to the first location after the message,
85  * 'exp_dn' is a pointer to a buffer of size 'length' for the result.
86  * Return size of compressed name or -1 if there was an error.
87  */
88 int
89 dn_expand(msg, eomorig, comp_dn, exp_dn, length)
90 	const u_char *msg, *eomorig, *comp_dn;
91 	char *exp_dn;
92 	int length;
93 {
94 	register const u_char *cp;
95 	register char *dn;
96 	register int n, c;
97 	char *eom;
98 	int len = -1, checked = 0;
99 
100 	dn = exp_dn;
101 	cp = comp_dn;
102 	if (length > MAXHOSTNAMELEN-1)
103 		length = MAXHOSTNAMELEN-1;
104 	eom = exp_dn + length;
105 	/*
106 	 * fetch next label in domain name
107 	 */
108 	while ((n = *cp++)) {
109 		/*
110 		 * Check for indirection
111 		 */
112 		switch (n & INDIR_MASK) {
113 		case 0:
114 			if (dn != exp_dn) {
115 				if (dn >= eom)
116 					return (-1);
117 				*dn++ = '.';
118 			}
119 			if (dn+n >= eom)
120 				return (-1);
121 			checked += n + 1;
122 			while (--n >= 0) {
123 				if (((c = *cp++) == '.') || (c == '\\')) {
124 					if (dn + n + 2 >= eom)
125 						return (-1);
126 					*dn++ = '\\';
127 				}
128 				*dn++ = c;
129 				if (cp >= eomorig)	/* out of range */
130 					return (-1);
131 			}
132 			break;
133 
134 		case INDIR_MASK:
135 			if (len < 0)
136 				len = cp - comp_dn + 1;
137 			cp = msg + (((n & 0x3f) << 8) | (*cp & 0xff));
138 			if (cp < msg || cp >= eomorig)	/* out of range */
139 				return (-1);
140 			checked += 2;
141 			/*
142 			 * Check for loops in the compressed name;
143 			 * if we've looked at the whole message,
144 			 * there must be a loop.
145 			 */
146 			if (checked >= eomorig - msg)
147 				return (-1);
148 			break;
149 
150 		default:
151 			return (-1);			/* flag error */
152 		}
153 	}
154 	*dn = '\0';
155 	if (len < 0)
156 		len = cp - comp_dn;
157 	return (len);
158 }
159 
160 /*
161  * Compress domain name 'exp_dn' into 'comp_dn'.
162  * Return the size of the compressed name or -1.
163  * 'length' is the size of the array pointed to by 'comp_dn'.
164  * 'dnptrs' is a list of pointers to previous compressed names. dnptrs[0]
165  * is a pointer to the beginning of the message. The list ends with NULL.
166  * 'lastdnptr' is a pointer to the end of the arrary pointed to
167  * by 'dnptrs'. Side effect is to update the list of pointers for
168  * labels inserted into the message as we compress the name.
169  * If 'dnptr' is NULL, we don't try to compress names. If 'lastdnptr'
170  * is NULL, we don't update the list.
171  */
172 int
173 dn_comp(exp_dn, comp_dn, length, dnptrs, lastdnptr)
174 	const char *exp_dn;
175 	u_char *comp_dn, **dnptrs, **lastdnptr;
176 	int length;
177 {
178 	register u_char *cp, *dn;
179 	register int c, l;
180 	u_char **cpp, **lpp, *sp, *eob;
181 	u_char *msg;
182 
183 	dn = (u_char *)exp_dn;
184 	cp = comp_dn;
185 	eob = cp + length;
186 	lpp = cpp = NULL;
187 	if (dnptrs != NULL) {
188 		if ((msg = *dnptrs++) != NULL) {
189 			for (cpp = dnptrs; *cpp != NULL; cpp++)
190 				;
191 			lpp = cpp;	/* end of list to search */
192 		}
193 	} else
194 		msg = NULL;
195 	for (c = *dn++; c != '\0'; ) {
196 		/* look to see if we can use pointers */
197 		if (msg != NULL) {
198 			if ((l = dn_find(dn-1, msg, dnptrs, lpp)) >= 0) {
199 				if (cp+1 >= eob)
200 					return (-1);
201 				*cp++ = (l >> 8) | INDIR_MASK;
202 				*cp++ = l % 256;
203 				return (cp - comp_dn);
204 			}
205 			/* not found, save it */
206 			if (lastdnptr != NULL && cpp < lastdnptr-1) {
207 				*cpp++ = cp;
208 				*cpp = NULL;
209 			}
210 		}
211 		sp = cp++;	/* save ptr to length byte */
212 		do {
213 			if (c == '.') {
214 				c = *dn++;
215 				break;
216 			}
217 			if (c == '\\') {
218 				if ((c = *dn++) == '\0')
219 					break;
220 			}
221 			if (cp >= eob) {
222 				if (msg != NULL)
223 					*lpp = NULL;
224 				return (-1);
225 			}
226 			*cp++ = c;
227 		} while ((c = *dn++) != '\0');
228 		/* catch trailing '.'s but not '..' */
229 		if ((l = cp - sp - 1) == 0 && c == '\0') {
230 			cp--;
231 			break;
232 		}
233 		if (l <= 0 || l > MAXLABEL) {
234 			if (msg != NULL)
235 				*lpp = NULL;
236 			return (-1);
237 		}
238 		*sp = l;
239 	}
240 	if (cp >= eob) {
241 		if (msg != NULL)
242 			*lpp = NULL;
243 		return (-1);
244 	}
245 	*cp++ = '\0';
246 	return (cp - comp_dn);
247 }
248 
249 /*
250  * Skip over a compressed domain name. Return the size or -1.
251  */
252 int
253 __dn_skipname(comp_dn, eom)
254 	const u_char *comp_dn, *eom;
255 {
256 	register const u_char *cp;
257 	register int n;
258 
259 	cp = comp_dn;
260 	while (cp < eom && (n = *cp++)) {
261 		/*
262 		 * check for indirection
263 		 */
264 		switch (n & INDIR_MASK) {
265 		case 0:			/* normal case, n == len */
266 			cp += n;
267 			continue;
268 		case INDIR_MASK:	/* indirection */
269 			cp++;
270 			break;
271 		default:		/* illegal type */
272 			return (-1);
273 		}
274 		break;
275 	}
276 	if (cp > eom)
277 		return (-1);
278 	return (cp - comp_dn);
279 }
280 
281 static int
282 mklower(ch)
283 	register int ch;
284 {
285 	if (isascii(ch) && isupper(ch))
286 		return (tolower(ch));
287 	return (ch);
288 }
289 
290 /*
291  * Search for expanded name from a list of previously compressed names.
292  * Return the offset from msg if found or -1.
293  * dnptrs is the pointer to the first name on the list,
294  * not the pointer to the start of the message.
295  */
296 static int
297 dn_find(exp_dn, msg, dnptrs, lastdnptr)
298 	u_char *exp_dn, *msg;
299 	u_char **dnptrs, **lastdnptr;
300 {
301 	register u_char *dn, *cp, **cpp;
302 	register int n;
303 	u_char *sp;
304 
305 	for (cpp = dnptrs; cpp < lastdnptr; cpp++) {
306 		dn = exp_dn;
307 		sp = cp = *cpp;
308 		while ((n = *cp++)) {
309 			/*
310 			 * check for indirection
311 			 */
312 			switch (n & INDIR_MASK) {
313 			case 0:		/* normal case, n == len */
314 				while (--n >= 0) {
315 					if (*dn == '.')
316 						goto next;
317 					if (*dn == '\\')
318 						dn++;
319 					if (mklower(*dn++) != mklower(*cp++))
320 						goto next;
321 				}
322 				if ((n = *dn++) == '\0' && *cp == '\0')
323 					return (sp - msg);
324 				if (n == '.')
325 					continue;
326 				goto next;
327 
328 			case INDIR_MASK:	/* indirection */
329 				cp = msg + (((n & 0x3f) << 8) | *cp);
330 				break;
331 
332 			default:	/* illegal type */
333 				return (-1);
334 			}
335 		}
336 		if (*dn == '\0')
337 			return (sp - msg);
338 	next:	;
339 	}
340 	return (-1);
341 }
342 
343 /*
344  * Verify that a domain name uses an acceptable character set.
345  */
346 
347 /*
348  * Note the conspicuous absence of ctype macros in these definitions.  On
349  * non-ASCII hosts, we can't depend on string literals or ctype macros to
350  * tell us anything about network-format data.  The rest of the BIND system
351  * is not careful about this, but for some reason, we're doing it right here.
352  */
353 #define PERIOD 0x2e
354 #define	hyphenchar(c) ((c) == 0x2d)
355 #define bslashchar(c) ((c) == 0x5c)
356 #define periodchar(c) ((c) == PERIOD)
357 #define asterchar(c) ((c) == 0x2a)
358 #define alphachar(c) (((c) >= 0x41 && (c) <= 0x5a) \
359 		   || ((c) >= 0x61 && (c) <= 0x7a))
360 #define digitchar(c) ((c) >= 0x30 && (c) <= 0x39)
361 
362 #define borderchar(c) (alphachar(c) || digitchar(c))
363 #define middlechar(c) (borderchar(c) || hyphenchar(c))
364 #define	domainchar(c) ((c) > 0x20 && (c) < 0x7f)
365 
366 int
367 res_hnok(dn)
368 	const char *dn;
369 {
370 	int pch = PERIOD, ch = *dn++;
371 
372 	while (ch != '\0') {
373 		int nch = *dn++;
374 
375 		if (periodchar(ch)) {
376 			;
377 		} else if (periodchar(pch)) {
378 			if (!borderchar(ch))
379 				return (0);
380 		} else if (periodchar(nch) || nch == '\0') {
381 			if (!borderchar(ch))
382 				return (0);
383 		} else {
384 			if (!middlechar(ch))
385 				return (0);
386 		}
387 		pch = ch, ch = nch;
388 	}
389 	return (1);
390 }
391 
392 /*
393  * hostname-like (A, MX, WKS) owners can have "*" as their first label
394  * but must otherwise be as a host name.
395  */
396 int
397 res_ownok(dn)
398 	const char *dn;
399 {
400 	if (asterchar(dn[0])) {
401 		if (periodchar(dn[1]))
402 			return (res_hnok(dn+2));
403 		if (dn[1] == '\0')
404 			return (1);
405 	}
406 	return (res_hnok(dn));
407 }
408 
409 /*
410  * SOA RNAMEs and RP RNAMEs can have any printable character in their first
411  * label, but the rest of the name has to look like a host name.
412  */
413 int
414 res_mailok(dn)
415 	const char *dn;
416 {
417 	int ch, escaped = 0;
418 
419 	/* "." is a valid missing representation */
420 	if (*dn == '\0')
421 		return(1);
422 
423 	/* otherwise <label>.<hostname> */
424 	while ((ch = *dn++) != '\0') {
425 		if (!domainchar(ch))
426 			return (0);
427 		if (!escaped && periodchar(ch))
428 			break;
429 		if (escaped)
430 			escaped = 0;
431 		else if (bslashchar(ch))
432 			escaped = 1;
433 	}
434 	if (periodchar(ch))
435 		return (res_hnok(dn));
436 	return(0);
437 }
438 
439 /*
440  * This function is quite liberal, since RFC 1034's character sets are only
441  * recommendations.
442  */
443 int
444 res_dnok(dn)
445 	const char *dn;
446 {
447 	int ch;
448 
449 	while ((ch = *dn++) != '\0')
450 		if (!domainchar(ch))
451 			return (0);
452 	return (1);
453 }
454 
455 /*
456  * Routines to insert/extract short/long's.
457  */
458 
459 u_int16_t
460 _getshort(msgp)
461 	register const u_char *msgp;
462 {
463 	register u_int16_t u;
464 
465 	GETSHORT(u, msgp);
466 	return (u);
467 }
468 
469 #ifdef NeXT
470 /*
471  * nExt machines have some funky library conventions, which we must maintain.
472  */
473 u_int16_t
474 res_getshort(msgp)
475 	register const u_char *msgp;
476 {
477 	return (_getshort(msgp));
478 }
479 #endif
480 
481 u_int32_t
482 _getlong(msgp)
483 	register const u_char *msgp;
484 {
485 	register u_int32_t u;
486 
487 	GETLONG(u, msgp);
488 	return (u);
489 }
490 
491 void
492 #if defined(__STDC__) || defined(__cplusplus)
493 __putshort(register u_int16_t s, register u_char *msgp)	/* must match proto */
494 #else
495 __putshort(s, msgp)
496 	register u_int16_t s;
497 	register u_char *msgp;
498 #endif
499 {
500 	PUTSHORT(s, msgp);
501 }
502 
503 void
504 __putlong(l, msgp)
505 	register u_int32_t l;
506 	register u_char *msgp;
507 {
508 	PUTLONG(l, msgp);
509 }
510