xref: /dragonfly/contrib/ldns/compat/b64_pton.c (revision 8a0bcd56)
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 <ldns/ldns.h>
45 
46 #include <sys/types.h>
47 #include <sys/param.h>
48 #ifdef HAVE_SYS_SOCKET_H
49 #include <sys/socket.h>
50 #endif
51 
52 #ifdef HAVE_NETINET_IN_H
53 #include <netinet/in.h>
54 #endif
55 #ifdef HAVE_ARPA_INET_H
56 #include <arpa/inet.h>
57 #endif
58 
59 #include <ctype.h>
60 #include <stdio.h>
61 #include <stdlib.h>
62 #include <string.h>
63 
64 #define Assert(Cond) if (!(Cond)) abort()
65 
66 static const char Base64[] =
67 	"ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/";
68 static const char Pad64 = '=';
69 
70 /* (From RFC1521 and draft-ietf-dnssec-secext-03.txt)
71    The following encoding technique is taken from RFC 1521 by Borenstein
72    and Freed.  It is reproduced here in a slightly edited form for
73    convenience.
74 
75    A 65-character subset of US-ASCII is used, enabling 6 bits to be
76    represented per printable character. (The extra 65th character, "=",
77    is used to signify a special processing function.)
78 
79    The encoding process represents 24-bit groups of input bits as output
80    strings of 4 encoded characters. Proceeding from left to right, a
81    24-bit input group is formed by concatenating 3 8-bit input groups.
82    These 24 bits are then treated as 4 concatenated 6-bit groups, each
83    of which is translated into a single digit in the base64 alphabet.
84 
85    Each 6-bit group is used as an index into an array of 64 printable
86    characters. The character referenced by the index is placed in the
87    output string.
88 
89                          Table 1: The Base64 Alphabet
90 
91       Value Encoding  Value Encoding  Value Encoding  Value Encoding
92           0 A            17 R            34 i            51 z
93           1 B            18 S            35 j            52 0
94           2 C            19 T            36 k            53 1
95           3 D            20 U            37 l            54 2
96           4 E            21 V            38 m            55 3
97           5 F            22 W            39 n            56 4
98           6 G            23 X            40 o            57 5
99           7 H            24 Y            41 p            58 6
100           8 I            25 Z            42 q            59 7
101           9 J            26 a            43 r            60 8
102          10 K            27 b            44 s            61 9
103          11 L            28 c            45 t            62 +
104          12 M            29 d            46 u            63 /
105          13 N            30 e            47 v
106          14 O            31 f            48 w         (pad) =
107          15 P            32 g            49 x
108          16 Q            33 h            50 y
109 
110    Special processing is performed if fewer than 24 bits are available
111    at the end of the data being encoded.  A full encoding quantum is
112    always completed at the end of a quantity.  When fewer than 24 input
113    bits are available in an input group, zero bits are added (on the
114    right) to form an integral number of 6-bit groups.  Padding at the
115    end of the data is performed using the '=' character.
116 
117    Since all base64 input is an integral number of octets, only the
118          -------------------------------------------------
119    following cases can arise:
120 
121        (1) the final quantum of encoding input is an integral
122            multiple of 24 bits; here, the final unit of encoded
123 	   output will be an integral multiple of 4 characters
124 	   with no "=" padding,
125        (2) the final quantum of encoding input is exactly 8 bits;
126            here, the final unit of encoded output will be two
127 	   characters followed by two "=" padding characters, or
128        (3) the final quantum of encoding input is exactly 16 bits;
129            here, the final unit of encoded output will be three
130 	   characters followed by one "=" padding character.
131    */
132 
133 /* skips all whitespace anywhere.
134    converts characters, four at a time, starting at (or after)
135    src from base - 64 numbers into three 8 bit bytes in the target area.
136    it returns the number of data bytes stored at the target, or -1 on error.
137  */
138 
139 int
140 ldns_b64_pton(char const *src, uint8_t *target, size_t targsize)
141 {
142 	int tarindex, state, ch;
143 	char *pos;
144 
145 	state = 0;
146 	tarindex = 0;
147 
148 	if (strlen(src) == 0) {
149 		return 0;
150 	}
151 
152 	while ((ch = *src++) != '\0') {
153 		if (isspace((unsigned char)ch))        /* Skip whitespace anywhere. */
154 			continue;
155 
156 		if (ch == Pad64)
157 			break;
158 
159 		pos = strchr(Base64, ch);
160 		if (pos == 0) {
161 			/* A non-base64 character. */
162 			return (-1);
163 		}
164 
165 		switch (state) {
166 		case 0:
167 			if (target) {
168 				if ((size_t)tarindex >= targsize)
169 					return (-1);
170 				target[tarindex] = (pos - Base64) << 2;
171 			}
172 			state = 1;
173 			break;
174 		case 1:
175 			if (target) {
176 				if ((size_t)tarindex + 1 >= targsize)
177 					return (-1);
178 				target[tarindex]   |=  (pos - Base64) >> 4;
179 				target[tarindex+1]  = ((pos - Base64) & 0x0f)
180 							<< 4 ;
181 			}
182 			tarindex++;
183 			state = 2;
184 			break;
185 		case 2:
186 			if (target) {
187 				if ((size_t)tarindex + 1 >= targsize)
188 					return (-1);
189 				target[tarindex]   |=  (pos - Base64) >> 2;
190 				target[tarindex+1]  = ((pos - Base64) & 0x03)
191 							<< 6;
192 			}
193 			tarindex++;
194 			state = 3;
195 			break;
196 		case 3:
197 			if (target) {
198 				if ((size_t)tarindex >= targsize)
199 					return (-1);
200 				target[tarindex] |= (pos - Base64);
201 			}
202 			tarindex++;
203 			state = 0;
204 			break;
205 		default:
206 			abort();
207 		}
208 	}
209 
210 	/*
211 	 * We are done decoding Base-64 chars.  Let's see if we ended
212 	 * on a byte boundary, and/or with erroneous trailing characters.
213 	 */
214 
215 	if (ch == Pad64) {		/* We got a pad char. */
216 		ch = *src++;		/* Skip it, get next. */
217 		switch (state) {
218 		case 0:		/* Invalid = in first position */
219 		case 1:		/* Invalid = in second position */
220 			return (-1);
221 
222 		case 2:		/* Valid, means one byte of info */
223 			/* Skip any number of spaces. */
224 			for ((void)NULL; ch != '\0'; ch = *src++)
225 				if (!isspace((unsigned char)ch))
226 					break;
227 			/* Make sure there is another trailing = sign. */
228 			if (ch != Pad64)
229 				return (-1);
230 			ch = *src++;		/* Skip the = */
231 			/* Fall through to "single trailing =" case. */
232 			/* FALLTHROUGH */
233 
234 		case 3:		/* Valid, means two bytes of info */
235 			/*
236 			 * We know this char is an =.  Is there anything but
237 			 * whitespace after it?
238 			 */
239 			for ((void)NULL; ch != '\0'; ch = *src++)
240 				if (!isspace((unsigned char)ch))
241 					return (-1);
242 
243 			/*
244 			 * Now make sure for cases 2 and 3 that the "extra"
245 			 * bits that slopped past the last full byte were
246 			 * zeros.  If we don't check them, they become a
247 			 * subliminal channel.
248 			 */
249 			if (target && target[tarindex] != 0)
250 				return (-1);
251 		}
252 	} else {
253 		/*
254 		 * We ended by seeing the end of the string.  Make sure we
255 		 * have no partial bytes lying around.
256 		 */
257 		if (state != 0)
258 			return (-1);
259 	}
260 
261 	return (tarindex);
262 }
263