xref: /freebsd/contrib/ldns/compat/b64_pton.c (revision 2008043f386721d58158e37e0d7e50df8095942d)
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 #include <ctype.h>
44 #include <stdlib.h>
45 #include <string.h>
46 
47 static const char Base64[] =
48 	"ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/";
49 static const char Pad64 = '=';
50 
51 /* (From RFC1521 and draft-ietf-dnssec-secext-03.txt)
52    The following encoding technique is taken from RFC 1521 by Borenstein
53    and Freed.  It is reproduced here in a slightly edited form for
54    convenience.
55 
56    A 65-character subset of US-ASCII is used, enabling 6 bits to be
57    represented per printable character. (The extra 65th character, "=",
58    is used to signify a special processing function.)
59 
60    The encoding process represents 24-bit groups of input bits as output
61    strings of 4 encoded characters. Proceeding from left to right, a
62    24-bit input group is formed by concatenating 3 8-bit input groups.
63    These 24 bits are then treated as 4 concatenated 6-bit groups, each
64    of which is translated into a single digit in the base64 alphabet.
65 
66    Each 6-bit group is used as an index into an array of 64 printable
67    characters. The character referenced by the index is placed in the
68    output string.
69 
70                          Table 1: The Base64 Alphabet
71 
72       Value Encoding  Value Encoding  Value Encoding  Value Encoding
73           0 A            17 R            34 i            51 z
74           1 B            18 S            35 j            52 0
75           2 C            19 T            36 k            53 1
76           3 D            20 U            37 l            54 2
77           4 E            21 V            38 m            55 3
78           5 F            22 W            39 n            56 4
79           6 G            23 X            40 o            57 5
80           7 H            24 Y            41 p            58 6
81           8 I            25 Z            42 q            59 7
82           9 J            26 a            43 r            60 8
83          10 K            27 b            44 s            61 9
84          11 L            28 c            45 t            62 +
85          12 M            29 d            46 u            63 /
86          13 N            30 e            47 v
87          14 O            31 f            48 w         (pad) =
88          15 P            32 g            49 x
89          16 Q            33 h            50 y
90 
91    Special processing is performed if fewer than 24 bits are available
92    at the end of the data being encoded.  A full encoding quantum is
93    always completed at the end of a quantity.  When fewer than 24 input
94    bits are available in an input group, zero bits are added (on the
95    right) to form an integral number of 6-bit groups.  Padding at the
96    end of the data is performed using the '=' character.
97 
98    Since all base64 input is an integral number of octets, only the
99          -------------------------------------------------
100    following cases can arise:
101 
102        (1) the final quantum of encoding input is an integral
103            multiple of 24 bits; here, the final unit of encoded
104 	   output will be an integral multiple of 4 characters
105 	   with no "=" padding,
106        (2) the final quantum of encoding input is exactly 8 bits;
107            here, the final unit of encoded output will be two
108 	   characters followed by two "=" padding characters, or
109        (3) the final quantum of encoding input is exactly 16 bits;
110            here, the final unit of encoded output will be three
111 	   characters followed by one "=" padding character.
112    */
113 
114 /* skips all whitespace anywhere.
115    converts characters, four at a time, starting at (or after)
116    src from base - 64 numbers into three 8 bit bytes in the target area.
117    it returns the number of data bytes stored at the target, or -1 on error.
118  */
119 
120 int
121 ldns_b64_pton(char const *origsrc, uint8_t *target, size_t targsize)
122 {
123 	unsigned char const* src = (unsigned char*)origsrc;
124 	int tarindex, state, ch;
125 	char *pos;
126 
127 	state = 0;
128 	tarindex = 0;
129 
130 	if (strlen(origsrc) == 0) {
131 		return 0;
132 	}
133 
134 	while ((ch = *src++) != '\0') {
135 		if (isspace((unsigned char)ch))        /* Skip whitespace anywhere. */
136 			continue;
137 
138 		if (ch == Pad64)
139 			break;
140 
141 		pos = strchr(Base64, ch);
142 		if (pos == 0) {
143 			/* A non-base64 character. */
144 			return (-1);
145 		}
146 
147 		switch (state) {
148 		case 0:
149 			if (target) {
150 				if ((size_t)tarindex >= targsize)
151 					return (-1);
152 				target[tarindex] = (pos - Base64) << 2;
153 			}
154 			state = 1;
155 			break;
156 		case 1:
157 			if (target) {
158 				if ((size_t)tarindex + 1 >= targsize)
159 					return (-1);
160 				target[tarindex]   |=  (pos - Base64) >> 4;
161 				target[tarindex+1]  = ((pos - Base64) & 0x0f)
162 							<< 4 ;
163 			}
164 			tarindex++;
165 			state = 2;
166 			break;
167 		case 2:
168 			if (target) {
169 				if ((size_t)tarindex + 1 >= targsize)
170 					return (-1);
171 				target[tarindex]   |=  (pos - Base64) >> 2;
172 				target[tarindex+1]  = ((pos - Base64) & 0x03)
173 							<< 6;
174 			}
175 			tarindex++;
176 			state = 3;
177 			break;
178 		case 3:
179 			if (target) {
180 				if ((size_t)tarindex >= targsize)
181 					return (-1);
182 				target[tarindex] |= (pos - Base64);
183 			}
184 			tarindex++;
185 			state = 0;
186 			break;
187 		default:
188 			abort();
189 		}
190 	}
191 
192 	/*
193 	 * We are done decoding Base-64 chars.  Let's see if we ended
194 	 * on a byte boundary, and/or with erroneous trailing characters.
195 	 */
196 
197 	if (ch == Pad64) {		/* We got a pad char. */
198 		ch = *src++;		/* Skip it, get next. */
199 		switch (state) {
200 		case 0:		/* Invalid = in first position */
201 		case 1:		/* Invalid = in second position */
202 			return (-1);
203 
204 		case 2:		/* Valid, means one byte of info */
205 			/* Skip any number of spaces. */
206 			for ((void)NULL; ch != '\0'; ch = *src++)
207 				if (!isspace((unsigned char)ch))
208 					break;
209 			/* Make sure there is another trailing = sign. */
210 			if (ch != Pad64)
211 				return (-1);
212 			ch = *src++;		/* Skip the = */
213 			/* Fall through to "single trailing =" case. */
214 			/* FALLTHROUGH */
215 
216 		case 3:		/* Valid, means two bytes of info */
217 			/*
218 			 * We know this char is an =.  Is there anything but
219 			 * whitespace after it?
220 			 */
221 			for ((void)NULL; ch != '\0'; ch = *src++)
222 				if (!isspace((unsigned char)ch))
223 					return (-1);
224 
225 			/*
226 			 * Now make sure for cases 2 and 3 that the "extra"
227 			 * bits that slopped past the last full byte were
228 			 * zeros.  If we don't check them, they become a
229 			 * subliminal channel.
230 			 */
231 			if (target && target[tarindex] != 0)
232 				return (-1);
233 		}
234 	} else {
235 		/*
236 		 * We ended by seeing the end of the string.  Make sure we
237 		 * have no partial bytes lying around.
238 		 */
239 		if (state != 0)
240 			return (-1);
241 	}
242 
243 	return (tarindex);
244 }
245