1 /*	$OpenBSD: base64.c,v 1.8 2015/01/16 16:48:51 deraadt Exp $	*/
2 
3 /*
4  * Copyright (c) 1996 by Internet Software Consortium.
5  *
6  * Permission to use, copy, modify, and distribute this software for any
7  * purpose with or without fee is hereby granted, provided that the above
8  * copyright notice and this permission notice appear in all copies.
9  *
10  * THE SOFTWARE IS PROVIDED "AS IS" AND INTERNET SOFTWARE CONSORTIUM DISCLAIMS
11  * ALL WARRANTIES WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES
12  * OF MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL INTERNET SOFTWARE
13  * CONSORTIUM BE LIABLE FOR ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL
14  * DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR
15  * PROFITS, WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS
16  * ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS
17  * SOFTWARE.
18  */
19 
20 /*
21  * Portions Copyright (c) 1995 by International Business Machines, Inc.
22  *
23  * International Business Machines, Inc. (hereinafter called IBM) grants
24  * permission under its copyrights to use, copy, modify, and distribute this
25  * Software with or without fee, provided that the above copyright notice and
26  * all paragraphs of this notice appear in all copies, and that the name of IBM
27  * not be used in connection with the marketing of any product incorporating
28  * the Software or modifications thereof, without specific, written prior
29  * permission.
30  *
31  * To the extent it has a right to do so, IBM grants an immunity from suit
32  * under its patents, if any, for the use, sale or manufacture of products to
33  * the extent that such products are used for performing Domain Name System
34  * dynamic updates in TCP/IP networks by means of the Software.  No immunity is
35  * granted for any product per se or for any other function of any product.
36  *
37  * THE SOFTWARE IS PROVIDED "AS IS", AND IBM DISCLAIMS ALL WARRANTIES,
38  * INCLUDING ALL IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A
39  * PARTICULAR PURPOSE.  IN NO EVENT SHALL IBM BE LIABLE FOR ANY SPECIAL,
40  * DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES WHATSOEVER ARISING
41  * OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE, EVEN
42  * IF IBM IS APPRISED OF THE POSSIBILITY OF SUCH DAMAGES.
43  */
44 
45 #include <sys/types.h>
46 #include <sys/socket.h>
47 #include <netinet/in.h>
48 #include <arpa/inet.h>
49 #include <arpa/nameser.h>
50 
51 #include <ctype.h>
52 #include <resolv.h>
53 #include <stdio.h>
54 
55 #include <stdlib.h>
56 #include <string.h>
57 
58 static const char Base64[] =
59 	"ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/";
60 static const char Pad64 = '=';
61 
62 /* (From RFC1521 and draft-ietf-dnssec-secext-03.txt)
63    The following encoding technique is taken from RFC 1521 by Borenstein
64    and Freed.  It is reproduced here in a slightly edited form for
65    convenience.
66 
67    A 65-character subset of US-ASCII is used, enabling 6 bits to be
68    represented per printable character. (The extra 65th character, "=",
69    is used to signify a special processing function.)
70 
71    The encoding process represents 24-bit groups of input bits as output
72    strings of 4 encoded characters. Proceeding from left to right, a
73    24-bit input group is formed by concatenating 3 8-bit input groups.
74    These 24 bits are then treated as 4 concatenated 6-bit groups, each
75    of which is translated into a single digit in the base64 alphabet.
76 
77    Each 6-bit group is used as an index into an array of 64 printable
78    characters. The character referenced by the index is placed in the
79    output string.
80 
81                          Table 1: The Base64 Alphabet
82 
83       Value Encoding  Value Encoding  Value Encoding  Value Encoding
84           0 A            17 R            34 i            51 z
85           1 B            18 S            35 j            52 0
86           2 C            19 T            36 k            53 1
87           3 D            20 U            37 l            54 2
88           4 E            21 V            38 m            55 3
89           5 F            22 W            39 n            56 4
90           6 G            23 X            40 o            57 5
91           7 H            24 Y            41 p            58 6
92           8 I            25 Z            42 q            59 7
93           9 J            26 a            43 r            60 8
94          10 K            27 b            44 s            61 9
95          11 L            28 c            45 t            62 +
96          12 M            29 d            46 u            63 /
97          13 N            30 e            47 v
98          14 O            31 f            48 w         (pad) =
99          15 P            32 g            49 x
100          16 Q            33 h            50 y
101 
102    Special processing is performed if fewer than 24 bits are available
103    at the end of the data being encoded.  A full encoding quantum is
104    always completed at the end of a quantity.  When fewer than 24 input
105    bits are available in an input group, zero bits are added (on the
106    right) to form an integral number of 6-bit groups.  Padding at the
107    end of the data is performed using the '=' character.
108 
109    Since all base64 input is an integral number of octets, only the
110          -------------------------------------------------                       
111    following cases can arise:
112    
113        (1) the final quantum of encoding input is an integral
114            multiple of 24 bits; here, the final unit of encoded
115 	   output will be an integral multiple of 4 characters
116 	   with no "=" padding,
117        (2) the final quantum of encoding input is exactly 8 bits;
118            here, the final unit of encoded output will be two
119 	   characters followed by two "=" padding characters, or
120        (3) the final quantum of encoding input is exactly 16 bits;
121            here, the final unit of encoded output will be three
122 	   characters followed by one "=" padding character.
123    */
124 
125 int
126 b64_ntop(src, srclength, target, targsize)
127 	u_char const *src;
128 	size_t srclength;
129 	char *target;
130 	size_t targsize;
131 {
132 	size_t datalength = 0;
133 	u_char input[3];
134 	u_char output[4];
135 	int i;
136 
137 	while (2 < srclength) {
138 		input[0] = *src++;
139 		input[1] = *src++;
140 		input[2] = *src++;
141 		srclength -= 3;
142 
143 		output[0] = input[0] >> 2;
144 		output[1] = ((input[0] & 0x03) << 4) + (input[1] >> 4);
145 		output[2] = ((input[1] & 0x0f) << 2) + (input[2] >> 6);
146 		output[3] = input[2] & 0x3f;
147 
148 		if (datalength + 4 > targsize)
149 			return (-1);
150 		target[datalength++] = Base64[output[0]];
151 		target[datalength++] = Base64[output[1]];
152 		target[datalength++] = Base64[output[2]];
153 		target[datalength++] = Base64[output[3]];
154 	}
155     
156 	/* Now we worry about padding. */
157 	if (0 != srclength) {
158 		/* Get what's left. */
159 		input[0] = input[1] = input[2] = '\0';
160 		for (i = 0; i < srclength; i++)
161 			input[i] = *src++;
162 	
163 		output[0] = input[0] >> 2;
164 		output[1] = ((input[0] & 0x03) << 4) + (input[1] >> 4);
165 		output[2] = ((input[1] & 0x0f) << 2) + (input[2] >> 6);
166 
167 		if (datalength + 4 > targsize)
168 			return (-1);
169 		target[datalength++] = Base64[output[0]];
170 		target[datalength++] = Base64[output[1]];
171 		if (srclength == 1)
172 			target[datalength++] = Pad64;
173 		else
174 			target[datalength++] = Base64[output[2]];
175 		target[datalength++] = Pad64;
176 	}
177 	if (datalength >= targsize)
178 		return (-1);
179 	target[datalength] = '\0';	/* Returned value doesn't count \0. */
180 	return (datalength);
181 }
182 
183 /* skips all whitespace anywhere.
184    converts characters, four at a time, starting at (or after)
185    src from base - 64 numbers into three 8 bit bytes in the target area.
186    it returns the number of data bytes stored at the target, or -1 on error.
187  */
188 
189 int
190 b64_pton(src, target, targsize)
191 	char const *src;
192 	u_char *target;
193 	size_t targsize;
194 {
195 	int tarindex, state, ch;
196 	u_char nextbyte;
197 	char *pos;
198 
199 	state = 0;
200 	tarindex = 0;
201 
202 	while ((ch = (unsigned char)*src++) != '\0') {
203 		if (isspace(ch))	/* Skip whitespace anywhere. */
204 			continue;
205 
206 		if (ch == Pad64)
207 			break;
208 
209 		pos = strchr(Base64, ch);
210 		if (pos == 0) 		/* A non-base64 character. */
211 			return (-1);
212 
213 		switch (state) {
214 		case 0:
215 			if (target) {
216 				if (tarindex >= targsize)
217 					return (-1);
218 				target[tarindex] = (pos - Base64) << 2;
219 			}
220 			state = 1;
221 			break;
222 		case 1:
223 			if (target) {
224 				if (tarindex >= targsize)
225 					return (-1);
226 				target[tarindex]   |=  (pos - Base64) >> 4;
227 				nextbyte = ((pos - Base64) & 0x0f) << 4;
228 				if (tarindex + 1 < targsize)
229 					target[tarindex+1] = nextbyte;
230 				else if (nextbyte)
231 					return (-1);
232 			}
233 			tarindex++;
234 			state = 2;
235 			break;
236 		case 2:
237 			if (target) {
238 				if (tarindex >= targsize)
239 					return (-1);
240 				target[tarindex]   |=  (pos - Base64) >> 2;
241 				nextbyte = ((pos - Base64) & 0x03) << 6;
242 				if (tarindex + 1 < targsize)
243 					target[tarindex+1] = nextbyte;
244 				else if (nextbyte)
245 					return (-1);
246 			}
247 			tarindex++;
248 			state = 3;
249 			break;
250 		case 3:
251 			if (target) {
252 				if (tarindex >= targsize)
253 					return (-1);
254 				target[tarindex] |= (pos - Base64);
255 			}
256 			tarindex++;
257 			state = 0;
258 			break;
259 		}
260 	}
261 
262 	/*
263 	 * We are done decoding Base-64 chars.  Let's see if we ended
264 	 * on a byte boundary, and/or with erroneous trailing characters.
265 	 */
266 
267 	if (ch == Pad64) {			/* We got a pad char. */
268 		ch = (unsigned char)*src++;	/* Skip it, get next. */
269 		switch (state) {
270 		case 0:		/* Invalid = in first position */
271 		case 1:		/* Invalid = in second position */
272 			return (-1);
273 
274 		case 2:		/* Valid, means one byte of info */
275 			/* Skip any number of spaces. */
276 			for (; ch != '\0'; ch = (unsigned char)*src++)
277 				if (!isspace(ch))
278 					break;
279 			/* Make sure there is another trailing = sign. */
280 			if (ch != Pad64)
281 				return (-1);
282 			ch = (unsigned char)*src++;		/* Skip the = */
283 			/* Fall through to "single trailing =" case. */
284 			/* FALLTHROUGH */
285 
286 		case 3:		/* Valid, means two bytes of info */
287 			/*
288 			 * We know this char is an =.  Is there anything but
289 			 * whitespace after it?
290 			 */
291 			for (; ch != '\0'; ch = (unsigned char)*src++)
292 				if (!isspace(ch))
293 					return (-1);
294 
295 			/*
296 			 * Now make sure for cases 2 and 3 that the "extra"
297 			 * bits that slopped past the last full byte were
298 			 * zeros.  If we don't check them, they become a
299 			 * subliminal channel.
300 			 */
301 			if (target && tarindex < targsize &&
302 			    target[tarindex] != 0)
303 				return (-1);
304 		}
305 	} else {
306 		/*
307 		 * We ended by seeing the end of the string.  Make sure we
308 		 * have no partial bytes lying around.
309 		 */
310 		if (state != 0)
311 			return (-1);
312 	}
313 
314 	return (tarindex);
315 }