1 /* base64.c -- Encode binary data using printable characters.
2 Copyright (C) 1999, 2000, 2001, 2004, 2005, 2006 Free Software
5 This program is free software; you can redistribute it and/or modify
6 it under the terms of the GNU General Public License as published by
7 the Free Software Foundation; either version 3, or (at your option)
10 This program is distributed in the hope that it will be useful,
11 but WITHOUT ANY WARRANTY; without even the implied warranty of
12 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
13 GNU General Public License for more details.
15 You should have received a copy of the GNU General Public License
16 along with this program; if not, write to the Free Software Foundation,
17 Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. */
19 /* Written by Simon Josefsson. Partially adapted from GNU MailUtils
20 * (mailbox/filter_trans.c, as of 2004-11-28). Improved by review
21 * from Paul Eggert, Bruno Haible, and Stepan Kasal.
23 * See also RFC 3548 <http://www.ietf.org/rfc/rfc3548.txt>.
25 * Be careful with error checking. Here is how you would typically
26 * use these functions:
28 * bool ok = base64_decode_alloc (in, inlen, &out, &outlen);
30 * FAIL: input was not valid base64
32 * FAIL: memory allocation error
33 * OK: data in OUT/OUTLEN
35 * size_t outlen = base64_encode_alloc (in, inlen, &out);
36 * if (out == NULL && outlen == 0 && inlen != 0)
37 * FAIL: input too long
39 * FAIL: memory allocation error
40 * OK: data in OUT/OUTLEN.
55 /* C89 compliant way to cast 'char' to 'unsigned char'. */
56 static inline unsigned char
62 /* Base64 encode IN array of size INLEN into OUT array of size OUTLEN.
63 If OUTLEN is less than BASE64_LENGTH(INLEN), write as many bytes as
64 possible. If OUTLEN is larger than BASE64_LENGTH(INLEN), also zero
65 terminate the output buffer. */
67 base64_encode (const char *restrict in
, size_t inlen
,
68 char *restrict out
, size_t outlen
)
70 static const char b64str
[64] =
71 "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/";
73 while (inlen
&& outlen
)
75 *out
++ = b64str
[(to_uchar (in
[0]) >> 2) & 0x3f];
78 *out
++ = b64str
[((to_uchar (in
[0]) << 4)
79 + (--inlen
? to_uchar (in
[1]) >> 4 : 0))
85 ? b64str
[((to_uchar (in
[1]) << 2)
86 + (--inlen
? to_uchar (in
[2]) >> 6 : 0))
91 *out
++ = inlen
? b64str
[to_uchar (in
[2]) & 0x3f] : '=';
104 /* Allocate a buffer and store zero terminated base64 encoded data
105 from array IN of size INLEN, returning BASE64_LENGTH(INLEN), i.e.,
106 the length of the encoded data, excluding the terminating zero. On
107 return, the OUT variable will hold a pointer to newly allocated
108 memory that must be deallocated by the caller. If output string
109 length would overflow, 0 is returned and OUT is set to NULL. If
110 memory allocation failed, OUT is set to NULL, and the return value
111 indicates length of the requested memory block, i.e.,
112 BASE64_LENGTH(inlen) + 1. */
114 base64_encode_alloc (const char *in
, size_t inlen
, char **out
)
116 size_t outlen
= 1 + BASE64_LENGTH (inlen
);
118 /* Check for overflow in outlen computation.
120 * If there is no overflow, outlen >= inlen.
122 * If the operation (inlen + 2) overflows then it yields at most +1, so
125 * If the multiplication overflows, we lose at least half of the
126 * correct value, so the result is < ((inlen + 2) / 3) * 2, which is
127 * less than (inlen + 2) * 0.66667, which is less than inlen as soon as
136 *out
= malloc (outlen
);
140 base64_encode (in
, inlen
, *out
, outlen
);
145 /* With this approach this file works independent of the charset used
146 (think EBCDIC). However, it does assume that the characters in the
147 Base64 alphabet (A-Za-z0-9+/) are encoded in 0..255. POSIX
148 1003.1-2001 require that char and unsigned char are 8-bit
149 quantities, though, taking care of that problem. But this may be a
150 potential problem on non-POSIX C99 platforms.
152 IBM C V6 for AIX mishandles "#define B64(x) ...'x'...", so use "_"
153 as the formal parameter rather than "x". */
221 static const signed char b64
[0x100] = {
222 B64 (0), B64 (1), B64 (2), B64 (3),
223 B64 (4), B64 (5), B64 (6), B64 (7),
224 B64 (8), B64 (9), B64 (10), B64 (11),
225 B64 (12), B64 (13), B64 (14), B64 (15),
226 B64 (16), B64 (17), B64 (18), B64 (19),
227 B64 (20), B64 (21), B64 (22), B64 (23),
228 B64 (24), B64 (25), B64 (26), B64 (27),
229 B64 (28), B64 (29), B64 (30), B64 (31),
230 B64 (32), B64 (33), B64 (34), B64 (35),
231 B64 (36), B64 (37), B64 (38), B64 (39),
232 B64 (40), B64 (41), B64 (42), B64 (43),
233 B64 (44), B64 (45), B64 (46), B64 (47),
234 B64 (48), B64 (49), B64 (50), B64 (51),
235 B64 (52), B64 (53), B64 (54), B64 (55),
236 B64 (56), B64 (57), B64 (58), B64 (59),
237 B64 (60), B64 (61), B64 (62), B64 (63),
238 B64 (64), B64 (65), B64 (66), B64 (67),
239 B64 (68), B64 (69), B64 (70), B64 (71),
240 B64 (72), B64 (73), B64 (74), B64 (75),
241 B64 (76), B64 (77), B64 (78), B64 (79),
242 B64 (80), B64 (81), B64 (82), B64 (83),
243 B64 (84), B64 (85), B64 (86), B64 (87),
244 B64 (88), B64 (89), B64 (90), B64 (91),
245 B64 (92), B64 (93), B64 (94), B64 (95),
246 B64 (96), B64 (97), B64 (98), B64 (99),
247 B64 (100), B64 (101), B64 (102), B64 (103),
248 B64 (104), B64 (105), B64 (106), B64 (107),
249 B64 (108), B64 (109), B64 (110), B64 (111),
250 B64 (112), B64 (113), B64 (114), B64 (115),
251 B64 (116), B64 (117), B64 (118), B64 (119),
252 B64 (120), B64 (121), B64 (122), B64 (123),
253 B64 (124), B64 (125), B64 (126), B64 (127),
254 B64 (128), B64 (129), B64 (130), B64 (131),
255 B64 (132), B64 (133), B64 (134), B64 (135),
256 B64 (136), B64 (137), B64 (138), B64 (139),
257 B64 (140), B64 (141), B64 (142), B64 (143),
258 B64 (144), B64 (145), B64 (146), B64 (147),
259 B64 (148), B64 (149), B64 (150), B64 (151),
260 B64 (152), B64 (153), B64 (154), B64 (155),
261 B64 (156), B64 (157), B64 (158), B64 (159),
262 B64 (160), B64 (161), B64 (162), B64 (163),
263 B64 (164), B64 (165), B64 (166), B64 (167),
264 B64 (168), B64 (169), B64 (170), B64 (171),
265 B64 (172), B64 (173), B64 (174), B64 (175),
266 B64 (176), B64 (177), B64 (178), B64 (179),
267 B64 (180), B64 (181), B64 (182), B64 (183),
268 B64 (184), B64 (185), B64 (186), B64 (187),
269 B64 (188), B64 (189), B64 (190), B64 (191),
270 B64 (192), B64 (193), B64 (194), B64 (195),
271 B64 (196), B64 (197), B64 (198), B64 (199),
272 B64 (200), B64 (201), B64 (202), B64 (203),
273 B64 (204), B64 (205), B64 (206), B64 (207),
274 B64 (208), B64 (209), B64 (210), B64 (211),
275 B64 (212), B64 (213), B64 (214), B64 (215),
276 B64 (216), B64 (217), B64 (218), B64 (219),
277 B64 (220), B64 (221), B64 (222), B64 (223),
278 B64 (224), B64 (225), B64 (226), B64 (227),
279 B64 (228), B64 (229), B64 (230), B64 (231),
280 B64 (232), B64 (233), B64 (234), B64 (235),
281 B64 (236), B64 (237), B64 (238), B64 (239),
282 B64 (240), B64 (241), B64 (242), B64 (243),
283 B64 (244), B64 (245), B64 (246), B64 (247),
284 B64 (248), B64 (249), B64 (250), B64 (251),
285 B64 (252), B64 (253), B64 (254), B64 (255)
289 # define uchar_in_range(c) true
291 # define uchar_in_range(c) ((c) <= 255)
294 /* Return true if CH is a character from the Base64 alphabet, and
295 false otherwise. Note that '=' is padding and not considered to be
296 part of the alphabet. */
300 return uchar_in_range (to_uchar (ch
)) && 0 <= b64
[to_uchar (ch
)];
303 /* Decode base64 encoded input array IN of length INLEN to output
304 array OUT that can hold *OUTLEN bytes. Return true if decoding was
305 successful, i.e. if the input was valid base64 data, false
306 otherwise. If *OUTLEN is too small, as many bytes as possible will
307 be written to OUT. On return, *OUTLEN holds the length of decoded
308 bytes in OUT. Note that as soon as any non-alphabet characters are
309 encountered, decoding is stopped and false is returned. This means
310 that, when applicable, you must remove any line terminators that is
311 part of the data stream before calling this function. */
313 base64_decode (const char *restrict in
, size_t inlen
,
314 char *restrict out
, size_t *outlen
)
316 size_t outleft
= *outlen
;
320 if (!isbase64 (in
[0]) || !isbase64 (in
[1]))
325 *out
++ = ((b64
[to_uchar (in
[0])] << 2)
326 | (b64
[to_uchar (in
[1])] >> 4));
344 if (!isbase64 (in
[2]))
349 *out
++ = (((b64
[to_uchar (in
[1])] << 4) & 0xf0)
350 | (b64
[to_uchar (in
[2])] >> 2));
364 if (!isbase64 (in
[3]))
369 *out
++ = (((b64
[to_uchar (in
[2])] << 6) & 0xc0)
370 | b64
[to_uchar (in
[3])]);
388 /* Allocate an output buffer in *OUT, and decode the base64 encoded
389 data stored in IN of size INLEN to the *OUT buffer. On return, the
390 size of the decoded data is stored in *OUTLEN. OUTLEN may be NULL,
391 if the caller is not interested in the decoded length. *OUT may be
392 NULL to indicate an out of memory error, in which case *OUTLEN
393 contains the size of the memory block needed. The function returns
394 true on successful decoding and memory allocation errors. (Use the
395 *OUT and *OUTLEN parameters to differentiate between successful
396 decoding and memory error.) The function returns false if the
397 input was invalid, in which case *OUT is NULL and *OUTLEN is
400 base64_decode_alloc (const char *in
, size_t inlen
, char **out
,
403 /* This may allocate a few bytes too much, depending on input,
404 but it's not worth the extra CPU time to compute the exact amount.
405 The exact amount is 3 * inlen / 4, minus 1 if the input ends
406 with "=" and minus another 1 if the input ends with "==".
407 Dividing before multiplying avoids the possibility of overflow. */
408 size_t needlen
= 3 * (inlen
/ 4) + 2;
410 *out
= malloc (needlen
);
414 if (!base64_decode (in
, inlen
, *out
, &needlen
))