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