b64_pton.c revision 246827
1/*
2 * Copyright (c) 1996, 1998 by Internet Software Consortium.
3 *
4 * Permission to use, copy, modify, and distribute this software for any
5 * purpose with or without fee is hereby granted, provided that the above
6 * copyright notice and this permission notice appear in all copies.
7 *
8 * THE SOFTWARE IS PROVIDED "AS IS" AND INTERNET SOFTWARE CONSORTIUM DISCLAIMS
9 * ALL WARRANTIES WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES
10 * OF MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL INTERNET SOFTWARE
11 * CONSORTIUM BE LIABLE FOR ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL
12 * DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR
13 * PROFITS, WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS
14 * ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS
15 * SOFTWARE.
16 */
17
18/*
19 * Portions Copyright (c) 1995 by International Business Machines, Inc.
20 *
21 * International Business Machines, Inc. (hereinafter called IBM) grants
22 * permission under its copyrights to use, copy, modify, and distribute this
23 * Software with or without fee, provided that the above copyright notice and
24 * all paragraphs of this notice appear in all copies, and that the name of IBM
25 * not be used in connection with the marketing of any product incorporating
26 * the Software or modifications thereof, without specific, written prior
27 * permission.
28 *
29 * To the extent it has a right to do so, IBM grants an immunity from suit
30 * under its patents, if any, for the use, sale or manufacture of products to
31 * the extent that such products are used for performing Domain Name System
32 * dynamic updates in TCP/IP networks by means of the Software.  No immunity is
33 * granted for any product per se or for any other function of any product.
34 *
35 * THE SOFTWARE IS PROVIDED "AS IS", AND IBM DISCLAIMS ALL WARRANTIES,
36 * INCLUDING ALL IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A
37 * PARTICULAR PURPOSE.  IN NO EVENT SHALL IBM BE LIABLE FOR ANY SPECIAL,
38 * DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES WHATSOEVER ARISING
39 * OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE, EVEN
40 * IF IBM IS APPRISED OF THE POSSIBILITY OF SUCH DAMAGES.
41 */
42#include <ldns/config.h>
43
44#include <sys/types.h>
45#include <sys/param.h>
46#ifdef HAVE_SYS_SOCKET_H
47#include <sys/socket.h>
48#endif
49
50#ifdef HAVE_NETINET_IN_H
51#include <netinet/in.h>
52#endif
53#ifdef HAVE_ARPA_INET_H
54#include <arpa/inet.h>
55#endif
56
57#include <ctype.h>
58#include <stdio.h>
59#include <stdlib.h>
60#include <string.h>
61
62#define Assert(Cond) if (!(Cond)) abort()
63
64static const char Base64[] =
65	"ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/";
66static const char Pad64 = '=';
67
68/* (From RFC1521 and draft-ietf-dnssec-secext-03.txt)
69   The following encoding technique is taken from RFC 1521 by Borenstein
70   and Freed.  It is reproduced here in a slightly edited form for
71   convenience.
72
73   A 65-character subset of US-ASCII is used, enabling 6 bits to be
74   represented per printable character. (The extra 65th character, "=",
75   is used to signify a special processing function.)
76
77   The encoding process represents 24-bit groups of input bits as output
78   strings of 4 encoded characters. Proceeding from left to right, a
79   24-bit input group is formed by concatenating 3 8-bit input groups.
80   These 24 bits are then treated as 4 concatenated 6-bit groups, each
81   of which is translated into a single digit in the base64 alphabet.
82
83   Each 6-bit group is used as an index into an array of 64 printable
84   characters. The character referenced by the index is placed in the
85   output string.
86
87                         Table 1: The Base64 Alphabet
88
89      Value Encoding  Value Encoding  Value Encoding  Value Encoding
90          0 A            17 R            34 i            51 z
91          1 B            18 S            35 j            52 0
92          2 C            19 T            36 k            53 1
93          3 D            20 U            37 l            54 2
94          4 E            21 V            38 m            55 3
95          5 F            22 W            39 n            56 4
96          6 G            23 X            40 o            57 5
97          7 H            24 Y            41 p            58 6
98          8 I            25 Z            42 q            59 7
99          9 J            26 a            43 r            60 8
100         10 K            27 b            44 s            61 9
101         11 L            28 c            45 t            62 +
102         12 M            29 d            46 u            63 /
103         13 N            30 e            47 v
104         14 O            31 f            48 w         (pad) =
105         15 P            32 g            49 x
106         16 Q            33 h            50 y
107
108   Special processing is performed if fewer than 24 bits are available
109   at the end of the data being encoded.  A full encoding quantum is
110   always completed at the end of a quantity.  When fewer than 24 input
111   bits are available in an input group, zero bits are added (on the
112   right) to form an integral number of 6-bit groups.  Padding at the
113   end of the data is performed using the '=' character.
114
115   Since all base64 input is an integral number of octets, only the
116         -------------------------------------------------
117   following cases can arise:
118
119       (1) the final quantum of encoding input is an integral
120           multiple of 24 bits; here, the final unit of encoded
121	   output will be an integral multiple of 4 characters
122	   with no "=" padding,
123       (2) the final quantum of encoding input is exactly 8 bits;
124           here, the final unit of encoded output will be two
125	   characters followed by two "=" padding characters, or
126       (3) the final quantum of encoding input is exactly 16 bits;
127           here, the final unit of encoded output will be three
128	   characters followed by one "=" padding character.
129   */
130
131/* skips all whitespace anywhere.
132   converts characters, four at a time, starting at (or after)
133   src from base - 64 numbers into three 8 bit bytes in the target area.
134   it returns the number of data bytes stored at the target, or -1 on error.
135 */
136
137int
138ldns_b64_pton(char const *src, uint8_t *target, size_t targsize)
139{
140	int tarindex, state, ch;
141	char *pos;
142
143	state = 0;
144	tarindex = 0;
145
146	if (strlen(src) == 0) {
147		return 0;
148	}
149
150	while ((ch = *src++) != '\0') {
151		if (isspace((unsigned char)ch))        /* Skip whitespace anywhere. */
152			continue;
153
154		if (ch == Pad64)
155			break;
156
157		pos = strchr(Base64, ch);
158		if (pos == 0) {
159			/* A non-base64 character. */
160			return (-1);
161		}
162
163		switch (state) {
164		case 0:
165			if (target) {
166				if ((size_t)tarindex >= targsize)
167					return (-1);
168				target[tarindex] = (pos - Base64) << 2;
169			}
170			state = 1;
171			break;
172		case 1:
173			if (target) {
174				if ((size_t)tarindex + 1 >= targsize)
175					return (-1);
176				target[tarindex]   |=  (pos - Base64) >> 4;
177				target[tarindex+1]  = ((pos - Base64) & 0x0f)
178							<< 4 ;
179			}
180			tarindex++;
181			state = 2;
182			break;
183		case 2:
184			if (target) {
185				if ((size_t)tarindex + 1 >= targsize)
186					return (-1);
187				target[tarindex]   |=  (pos - Base64) >> 2;
188				target[tarindex+1]  = ((pos - Base64) & 0x03)
189							<< 6;
190			}
191			tarindex++;
192			state = 3;
193			break;
194		case 3:
195			if (target) {
196				if ((size_t)tarindex >= targsize)
197					return (-1);
198				target[tarindex] |= (pos - Base64);
199			}
200			tarindex++;
201			state = 0;
202			break;
203		default:
204			abort();
205		}
206	}
207
208	/*
209	 * We are done decoding Base-64 chars.  Let's see if we ended
210	 * on a byte boundary, and/or with erroneous trailing characters.
211	 */
212
213	if (ch == Pad64) {		/* We got a pad char. */
214		ch = *src++;		/* Skip it, get next. */
215		switch (state) {
216		case 0:		/* Invalid = in first position */
217		case 1:		/* Invalid = in second position */
218			return (-1);
219
220		case 2:		/* Valid, means one byte of info */
221			/* Skip any number of spaces. */
222			for ((void)NULL; ch != '\0'; ch = *src++)
223				if (!isspace((unsigned char)ch))
224					break;
225			/* Make sure there is another trailing = sign. */
226			if (ch != Pad64)
227				return (-1);
228			ch = *src++;		/* Skip the = */
229			/* Fall through to "single trailing =" case. */
230			/* FALLTHROUGH */
231
232		case 3:		/* Valid, means two bytes of info */
233			/*
234			 * We know this char is an =.  Is there anything but
235			 * whitespace after it?
236			 */
237			for ((void)NULL; ch != '\0'; ch = *src++)
238				if (!isspace((unsigned char)ch))
239					return (-1);
240
241			/*
242			 * Now make sure for cases 2 and 3 that the "extra"
243			 * bits that slopped past the last full byte were
244			 * zeros.  If we don't check them, they become a
245			 * subliminal channel.
246			 */
247			if (target && target[tarindex] != 0)
248				return (-1);
249		}
250	} else {
251		/*
252		 * We ended by seeing the end of the string.  Make sure we
253		 * have no partial bytes lying around.
254		 */
255		if (state != 0)
256			return (-1);
257	}
258
259	return (tarindex);
260}
261