base64: implements rfc4648, the base64 encoding

Encode buffers into base64 according to rfc4648.
Decode base64-encoded buffers according to the same standard.

Signed-off-by: <pb-ccan@barker.dropbear.id.au>
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>

Makefile-ccan

@@ -35,6 +35,7 @@ MODS_WITH_SRC := antithread \
 	asprintf \
 	autodata \
 	avl \
+	base64 \
 	bdelta \
 	block_pool \
 	breakpoint \

ccan/base64/LICENSE

+../../licenses/BSD-MIT
\ No newline at end of file

ccan/base64/_info

+#include "config.h"
+
+/**
+ * base64 - base64 encoding and decoding (rfc4648).
+ *
+ * base64 encoding is used to encode data in a 7-bit clean manner.
+ *  Commonly used for escaping data before encapsulation or transfer
+ *
+ * Example:
+ *      #include <stdio.h>
+ *      #include <string.h>
+ *      #include <ccan/base64/base64.h>
+ *
+ *      int main(int argc, char *argv[])
+ *      {
+ *      	char *base64_encoded_string;
+ *      	int i;
+ *
+ *      	// print the base64-encoded form of the program arguments
+ *      	for(i=1;i<argc;i++) {
+ *			size_t unencoded_length = strlen(argv[i]);
+ *      		size_t encoded_length = base64_encoded_length(unencoded_length);
+ *      		base64_encoded_string = malloc(encoded_length);
+ *      		base64_encode(base64_encoded_string, encoded_length,
+ *				 argv[i], unencoded_length);
+ *      		printf("%s\n", base64_encoded_string);
+ *      		free(base64_encoded_string);
+ *      	}
+ *
+ *      	return 0;
+ *      }
+ *
+ * License: BSD-MIT
+ */
+int main(int argc, char *argv[])
+{
+	if (argc != 2)
+		return 1;
+
+	return 1;
+}

ccan/base64/base64.c

+/* Licensed under BSD-MIT - see LICENSE file for details */
+#include "base64.h"
+
+#include <errno.h>
+#include <string.h>
+#include <assert.h>
+#include <stdint.h>
+
+/**
+ * sixbit_to_b64 - maps a 6-bit value to the base64 alphabet
+ * @param map A base 64 map (see base64_init_map)
+ * @param sixbit Six-bit value to map
+ * @return a base 64 character
+ */
+static char sixbit_to_b64(const base64_maps_t *maps, const uint8_t sixbit)
+{
+	assert(sixbit >= 0);
+	assert(sixbit <= 63);
+
+	return maps->encode_map[(unsigned char)sixbit];
+}
+
+/**
+ * sixbit_from_b64 - maps a base64-alphabet character to its 6-bit value
+ * @param maps A base 64 maps structure (see base64_init_maps)
+ * @param sixbit Six-bit value to map
+ * @return a six-bit value
+ */
+static int8_t sixbit_from_b64(const base64_maps_t *maps,
+			      const unsigned char b64letter)
+{
+	int8_t ret;
+
+	ret = maps->decode_map[(unsigned char)b64letter];
+	if (ret == (char)0xff) {
+		errno = EDOM;
+		return -1;
+	}
+
+	return ret;
+}
+
+bool base64_char_in_alphabet(const base64_maps_t *maps, const char b64char)
+{
+	return (maps->decode_map[(const unsigned char)b64char] != (char)0xff);
+}
+
+void base64_init_maps(base64_maps_t *dest, const char src[64])
+{
+	unsigned char i;
+
+	memcpy(dest->encode_map,src,64);
+	memset(dest->decode_map,0xff,256);
+	for (i=0; i<64; i++) {
+		dest->decode_map[(unsigned char)src[i]] = i;
+	}
+}
+
+size_t base64_encoded_length(size_t srclen)
+{
+	return ((srclen + 2) / 3) * 4;
+}
+
+void base64_encode_triplet_using_maps(const base64_maps_t *maps,
+				      char dest[4], const char src[3])
+{
+	char a = src[0];
+	char b = src[1];
+	char c = src[2];
+
+	dest[0] = sixbit_to_b64(maps, (a & 0xfc) >> 2);
+	dest[1] = sixbit_to_b64(maps, ((a & 0x3) << 4) | ((b & 0xf0) >> 4));
+	dest[2] = sixbit_to_b64(maps, ((c & 0xc0) >> 6) | ((b & 0xf) << 2));
+	dest[3] = sixbit_to_b64(maps, c & 0x3f);
+}
+
+void base64_encode_tail_using_maps(const base64_maps_t *maps, char dest[4],
+				   const char *src, const size_t srclen)
+{
+	char longsrc[3] = { 0 };
+
+	assert(srclen <= 3);
+
+	memcpy(longsrc, src, srclen);
+	base64_encode_triplet_using_maps(maps, dest, longsrc);
+	memset(dest+1+srclen, '=', 3-srclen);
+}
+
+ssize_t base64_encode_using_maps(const base64_maps_t *maps,
+				 char *dest, const size_t destlen,
+				 const char *src, const size_t srclen)
+{
+	size_t src_offset = 0;
+	size_t dest_offset = 0;
+
+	if (destlen < base64_encoded_length(srclen)) {
+		errno = EOVERFLOW;
+		return -1;
+	}
+
+	while (srclen - src_offset >= 3) {
+		base64_encode_triplet_using_maps(maps, &dest[dest_offset], &src[src_offset]);
+		src_offset += 3;
+		dest_offset += 4;
+	}
+
+	if (src_offset < srclen) {
+		base64_encode_tail_using_maps(maps, &dest[dest_offset], &src[src_offset], srclen-src_offset);
+		dest_offset += 4;
+	}
+
+	memset(&dest[dest_offset], '\0', destlen-dest_offset);
+
+	return dest_offset;
+}
+
+size_t base64_decoded_length(size_t srclen)
+{
+	return ((srclen+3)/4*3);
+}
+
+int base64_decode_quartet_using_maps(const base64_maps_t *maps, char dest[3],
+				     const char src[4])
+{
+	signed char a;
+	signed char b;
+	signed char c;
+	signed char d;
+
+	a = sixbit_from_b64(maps, src[0]);
+	b = sixbit_from_b64(maps, src[1]);
+	c = sixbit_from_b64(maps, src[2]);
+	d = sixbit_from_b64(maps, src[3]);
+
+	if ((a == -1) || (b == -1) || (c == -1) || (d == -1)) {
+		return -1;
+	}
+
+	dest[0] = (a << 2) | (b >> 4);
+	dest[1] = ((b & 0xf) << 4) | (c >> 2);
+	dest[2] = ((c & 0x3) << 6) | d;
+
+	return 0;
+}
+
+
+int base64_decode_tail_using_maps(const base64_maps_t *maps, char dest[3],
+				  const char * src, const size_t srclen)
+{
+	char longsrc[4];
+	int quartet_result;
+	size_t insize = srclen;
+
+	while (insize != 0 &&
+	       src[insize-1] == '=') { /* throw away padding symbols */
+		insize--;
+	}
+	if (insize == 0) {
+		return 0;
+	}
+	if (insize == 1) {
+		/* the input is malformed.... */
+		errno = EINVAL;
+		return -1;
+	}
+	memcpy(longsrc, src, insize);
+	memset(longsrc+insize, 'A', 4-insize);
+	quartet_result = base64_decode_quartet_using_maps(maps, dest, longsrc);
+	if (quartet_result == -1) {
+		return -1;
+	}
+
+	return insize - 1;
+}
+
+ssize_t base64_decode_using_maps(const base64_maps_t *maps,
+				 char *dest, const size_t destlen,
+				 const char *src, const size_t srclen)
+{
+	ssize_t dest_offset = 0;
+	ssize_t i;
+	size_t more;
+
+	if (destlen < base64_decoded_length(srclen)) {
+		errno = EOVERFLOW;
+		return -1;
+	}
+
+	for(i=0; srclen - i > 4; i+=4) {
+		if (base64_decode_quartet_using_maps(maps, &dest[dest_offset], &src[i]) == -1) {
+			return -1;
+		}
+		dest_offset += 3;
+	}
+
+	more = base64_decode_tail_using_maps(maps, &dest[dest_offset], &src[i], srclen - i);
+	if (more == -1) {
+		return -1;
+	}
+	dest_offset += more;
+
+	memset(&dest[dest_offset], '\0', destlen-dest_offset);
+
+	return dest_offset;
+}
+
+
+
+
+/**
+ * base64_maps_rfc4648 - pregenerated maps struct for rfc4648
+ */
+static const base64_maps_t base64_maps_rfc4648 = {
+  "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/",
+
+  "\xff\xff\xff\xff\xff" /* 0 */					\
+  "\xff\xff\xff\xff\xff" /* 5 */					\
+  "\xff\xff\xff\xff\xff" /* 10 */					\
+  "\xff\xff\xff\xff\xff" /* 15 */					\
+  "\xff\xff\xff\xff\xff" /* 20 */					\
+  "\xff\xff\xff\xff\xff" /* 25 */					\
+  "\xff\xff\xff\xff\xff" /* 30 */					\
+  "\xff\xff\xff\xff\xff" /* 35 */					\
+  "\xff\xff\xff\x3e\xff" /* 40 */					\
+  "\xff\xff\x3f\x34\x35" /* 45 */					\
+  "\x36\x37\x38\x39\x3a" /* 50 */					\
+  "\x3b\x3c\x3d\xff\xff" /* 55 */					\
+  "\xff\xff\xff\xff\xff" /* 60 */					\
+  "\x00\x01\x02\x03\x04" /* 65 A */					\
+  "\x05\x06\x07\x08\x09" /* 70 */					\
+  "\x0a\x0b\x0c\x0d\x0e" /* 75 */					\
+  "\x0f\x10\x11\x12\x13" /* 80 */					\
+  "\x14\x15\x16\x17\x18" /* 85 */					\
+  "\x19\xff\xff\xff\xff" /* 90 */					\
+  "\xff\xff\x1a\x1b\x1c" /* 95 */					\
+  "\x1d\x1e\x1f\x20\x21" /* 100 */					\
+  "\x22\x23\x24\x25\x26" /* 105 */					\
+  "\x27\x28\x29\x2a\x2b" /* 110 */					\
+  "\x2c\x2d\x2e\x2f\x30" /* 115 */					\
+  "\x31\x32\x33\xff\xff" /* 120 */					\
+  "\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff" /* 125 */			\
+  "\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff"				\
+  "\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff"				\
+  "\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff" /* 155 */			\
+  "\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff"				\
+  "\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff"				\
+  "\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff" /* 185 */			\
+  "\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff"				\
+  "\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff"				\
+  "\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff" /* 215 */			\
+  "\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff"				\
+  "\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff"				\
+  "\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff\xff" /* 245 */
+};

ccan/base64/base64.h

+/* Licensed under BSD-MIT - see LICENSE file for details */
+#ifndef CCAN_BASE64_H
+#define CCAN_BASE64_H
+
+#include <stddef.h>
+#include <stdbool.h>
+#include <sys/types.h>
+
+/**
+ * base64_maps_t - structure to hold maps for encode/decode
+ */
+typedef struct {
+	char encode_map[64];
+	signed char decode_map[256];
+} base64_maps_t;
+
+/**
+ * base64_encoded_length - Calculate encode buffer length
+ * @param srclen the size of the data to be encoded
+ * @note add 1 to this to get null-termination
+ * @return Buffer length required for encode
+ */
+size_t base64_encoded_length(size_t srclen);
+
+/**
+ * base64_decoded_length - Calculate decode buffer length
+ * @param srclen Length of the data to be decoded
+ * @note This does not return the size of the decoded data!  see base64_decode
+ * @return Minimum buffer length for safe decode
+ */
+size_t base64_decoded_length(size_t srclen);
+
+/**
+ * base64_init_maps - populate a base64_maps_t based on a supplied alphabet
+ * @param dest A base64 maps object
+ * @param src Alphabet to populate the maps from (e.g. base64_alphabet_rfc4648)
+ */
+void base64_init_maps(base64_maps_t *dest, const char src[64]);
+
+
+/**
+ * base64_encode_triplet_using_maps - encode 3 bytes into base64 using a specific alphabet
+ * @param maps Maps to use for encoding (see base64_init_maps)
+ * @param dest Buffer containing 3 bytes
+ * @param src Buffer containing 4 characters
+ */
+void base64_encode_triplet_using_maps(const base64_maps_t *maps,
+				      char dest[4], const char src[3]);
+
+/**
+ * base64_encode_tail_using_maps - encode the final bytes of a source using a specific alphabet
+ * @param maps Maps to use for encoding (see base64_init_maps)
+ * @param dest Buffer containing 4 bytes
+ * @param src Buffer containing srclen bytes
+ * @param srclen Number of bytes (<= 3) to encode in src
+ */
+void base64_encode_tail_using_maps(const base64_maps_t *maps, char dest[4],
+				   const char *src, size_t srclen);
+
+/**
+ * base64_encode_using_maps - encode a buffer into base64 using a specific alphabet
+ * @param maps Maps to use for encoding (see base64_init_maps)
+ * @param dest Buffer to encode into
+ * @param destlen Length of dest
+ * @param src Buffer to encode
+ * @param srclen Length of the data to encode
+ * @return Number of encoded bytes set in dest. -1 on error (and errno set)
+ * @note dest will be nul-padded to destlen (past any required padding)
+ * @note sets errno = EOVERFLOW if destlen is too small
+ */
+ssize_t base64_encode_using_maps(const base64_maps_t *maps,
+				 char *dest, size_t destlen,
+				 const char *src, size_t srclen);
+
+/*
+ * base64_char_in_alphabet - returns true if character can be part of an encoded string
+ * @param maps A base64 maps object (see base64_init_maps)
+ * @param b64char Character to check
+ */
+bool base64_char_in_alphabet(const base64_maps_t *maps, char b64char);
+
+/**
+ * base64_decode_using_maps - decode a base64-encoded string using a specific alphabet
+ * @param maps A base64 maps object (see base64_init_maps)
+ * @param dest Buffer to decode into
+ * @param destlen length of dest
+ * @param src the buffer to decode
+ * @param srclen the length of the data to decode
+ * @return Number of decoded bytes set in dest. -1 on error (and errno set)
+ * @note dest will be nul-padded to destlen
+ * @note sets errno = EOVERFLOW if destlen is too small
+ * @note sets errno = EDOM if src contains invalid characters
+ */
+ssize_t base64_decode_using_maps(const base64_maps_t *maps,
+				 char *dest, size_t destlen,
+				 const char *src, size_t srclen);
+
+/**
+ * base64_decode_quartet_using_maps - decode 4 bytes from base64 using a specific alphabet
+ * @param maps A base64 maps object (see base64_init_maps)
+ * @param dest Buffer containing 3 bytes
+ * @param src Buffer containing 4 bytes
+ * @return Number of decoded bytes set in dest. -1 on error (and errno set)
+ * @note sets errno = EDOM if src contains invalid characters
+ */
+int base64_decode_quartet_using_maps(const base64_maps_t *maps,
+				     char dest[3], const char src[4]);
+
+/**
+ * base64_decode_tail_using_maps - decode the final bytes of a base64 string using a specific alphabet
+ * @param maps A base64 maps object (see base64_init_maps)
+ * @param dest Buffer containing 3 bytes
+ * @param src Buffer containing 4 bytes - padded with '=' as required
+ * @param srclen Number of bytes to decode in src
+ * @return Number of decoded bytes set in dest. -1 on error (and errno set)
+ * @note sets errno = EDOM if src contains invalid characters
+ * @note sets errno = EINVAL if src is an invalid base64 tail
+ */
+int base64_decode_tail_using_maps(const base64_maps_t *maps, char *dest,
+				  const char *src, size_t srclen);
+
+
+/* the rfc4648 functions: */
+
+static const base64_maps_t base64_maps_rfc4648;
+
+/**
+ * base64_encode - Encode a buffer into base64 according to rfc4648
+ * @param dest Buffer to encode into
+ * @param destlen Length of the destination buffer
+ * @param src Buffer to encode
+ * @param srclen Length of the data to encode
+ * @return Number of encoded bytes set in dest. -1 on error (and errno set)
+ * @note dest will be nul-padded to destlen (past any required padding)
+ * @note sets errno = EOVERFLOW if destlen is too small
+ *
+ * This function encodes src according to http://tools.ietf.org/html/rfc4648
+ *
+ * Example:
+ *	size_t encoded_length;
+ *	char dest[100];
+ *	const char *src = "This string gets encoded";
+ *	encoded_length = base64_encode(dest, sizeof(dest), src, strlen(src));
+ *	printf("Returned data of length %zd @%p\n", encoded_length, &dest);
+ */
+static inline
+ssize_t base64_encode(char *dest, size_t destlen,
+		      const char *src, size_t srclen)
+{
+	return base64_encode_using_maps(&base64_maps_rfc4648,
+					dest, destlen, src, srclen);
+}
+
+/**
+ * base64_encode_triplet - encode 3 bytes into base64 according to rfc4648
+ * @param dest Buffer containing 4 bytes
+ * @param src Buffer containing 3 bytes
+ */
+static inline
+void base64_encode_triplet(char dest[4], const char src[3])
+{
+	base64_encode_triplet_using_maps(&base64_maps_rfc4648, dest, src);
+}
+
+/**
+ * base64_encode_tail - encode the final bytes of a source according to rfc4648
+ * @param dest Buffer containing 4 bytes
+ * @param src Buffer containing srclen bytes
+ * @param srclen Number of bytes (<= 3) to encode in src
+ */
+static inline
+void base64_encode_tail(char dest[4], const char *src, size_t srclen)
+{
+	base64_encode_tail_using_maps(&base64_maps_rfc4648, dest, src, srclen);
+}
+
+
+/**
+ * base64_decode - decode An rfc4648 base64-encoded string
+ * @param dest Buffer to decode into
+ * @param destlen Length of the destination buffer
+ * @param src Buffer to decode
+ * @param srclen Length of the data to decode
+ * @return Number of decoded bytes set in dest. -1 on error (and errno set)
+ * @note dest will be nul-padded to destlen
+ * @note sets errno = EOVERFLOW if destlen is too small
+ * @note sets errno = EDOM if src contains invalid characters
+ *
+ * This function decodes the buffer according to
+ * http://tools.ietf.org/html/rfc4648
+ *
+ * Example:
+ *	size_t decoded_length;
+ *	char ret[100];
+ *	const char *src = "Zm9vYmFyYmF6";
+ *	decoded_length = base64_decode(ret, sizeof(ret), src, strlen(src));
+ *	printf("Returned data of length %zd @%p\n", decoded_length, &ret);
+ */
+static inline
+ssize_t base64_decode(char *dest, size_t destlen,
+		      const char *src, size_t srclen)
+{
+	return base64_decode_using_maps(&base64_maps_rfc4648,
+					dest, destlen, src, srclen);
+}
+
+/**
+ * base64_decode_quartet - decode the first 4 characters in src into dest
+ * @param dest Buffer containing 3 bytes
+ * @param src Buffer containing 4 characters
+ * @return Number of decoded bytes set in dest. -1 on error (and errno set)
+ * @note sets errno = EDOM if src contains invalid characters
+ */
+static inline
+int base64_decode_quartet(char dest[3], const char src[4])
+{
+	return base64_decode_quartet_using_maps(&base64_maps_rfc4648,
+						dest, src);
+}
+
+/**
+ * @brief decode the final bytes of a base64 string from src into dest
+ * @param dest Buffer containing 3 bytes
+ * @param src Buffer containing 4 bytes - padded with '=' as required
+ * @param srclen Number of bytes to decode in src
+ * @return Number of decoded bytes set in dest. -1 on error (and errno set)
+ * @note sets errno = EDOM if src contains invalid characters
+ * @note sets errno = EINVAL if src is an invalid base64 tail
+ */
+static inline
+ssize_t base64_decode_tail(char dest[3], const char *src, size_t srclen)
+{
+	return base64_decode_tail_using_maps(&base64_maps_rfc4648,
+					     dest, src, srclen);
+}
+
+/* end rfc4648 functions */
+
+
+
+#endif /* CCAN_BASE64_H */

ccan/base64/test/moretap.h

+#ifndef _BASE64_MORETAP_H
+#define _BASE64_MORETAP_H
+
+#include <ccan/str/str.h>
+
+/**
+ * is_str - OK if strings are equal
+ * @e1: expression for the variable string
+ * @e2: expression for the expected string
+ *
+ * If the strings are equal, the test passes.
+ *
+ * Example:
+ *     is_str(give_me_a_fred(),"fred");
+ */
+static void _is_str(char *got,const char *expected, const char *got_string, const char *expected_string, const char *func, const char *file, int line) {
+	if (streq(expected,got)) {
+		_gen_result(1, func, file, line,"%s eq %s",
+			    got_string,expected_string);
+	} else {
+		_gen_result(0, func, file, line,"%s eq %s",
+			    got_string,expected_string);
+		diag("Expected: %s",expected);
+		diag("     Got: %s",got);
+	}
+}
+# define is_str(got,expected) _is_str(got,expected,#got,#expected,__func__, __FILE__, __LINE__)
+
+
+/**
+ * is_int - OK if arguments are equal when cast to integers
+ * @e1: expression for the number
+ * @e2: expression for the expected number
+ *
+ * If the numbers are equal, the test passes.
+ *
+ * Example:
+ *     is_int(give_me_17(),17);
+ */
+# define is_int(e1,e2 ...)						\
+  (((int)e1)==((int)e2) ?						\
+   _gen_result(1, __func__, __FILE__, __LINE__,"%s == %s",#e1,#e2) :	\
+   (_gen_result(0, __func__, __FILE__, __LINE__,"%s == %s",#e1,#e2)) || (diag("Expected: %d",e2),diag("     Got: %d",e1),0)) /* diag is void; note commas. */
+
+
+
+/**
+ * is_mem - OK if arguments are identical up to length @e3
+ * @e1: expression for the buffer
+ * @e2: expression for the expected buffer
+ * @e2: length to compare in buffers
+ *
+ * If the buffers are equal up to @e2, the test passes.
+ *
+ * Example:
+ *     is_mem(give_me_foo(),"foo",3);
+ */
+static void _is_mem(const char *got, const char *expected, const size_t len,
+	      const char *got_string, const char *expected_string, const char *len_string,
+	      const char *func, const char *file, int line) {
+	size_t offset = 0;
+
+	for (offset=0; offset<len; offset++) {
+		if (got[offset] != expected[offset]) {
+			_gen_result(0, func, file, line,"%s eq %s",got_string,expected_string);
+			/* diag("Expected: %s",e2); */
+			/* diag("     Got: %s",e1); */
+			diag("Buffers differ at offset %zd (got=0x%02x expected=0x%02x)",
+			     offset,got[offset],expected[offset]);
+			return;
+		}
+	}
+
+	_gen_result(1, __func__, __FILE__, __LINE__,"%s eq %s",
+		    expected_string,got_string);
+}
+# define is_mem(got,expected,len) \
+	_is_mem(got,expected,len,#got,#expected,#len,__func__, __FILE__, __LINE__)
+
+/**
+ * is_size_t - OK if arguments are equal when cast to size_t
+ * @e1: expression for the number
+ * @e2: expression for the expected number
+ *
+ * If the numbers are equal, the test passes.
+ *
+ * Example:
+ *     is_size_t(give_me_17(),17);
+ */
+# define is_size_t(e1,e2 ...)						\
+  ((size_t)(e1)==((size_t)e2) ?						\
+   _gen_result(1, __func__, __FILE__, __LINE__,"%s == %s",#e1,#e2) :	\
+   (_gen_result(0, __func__, __FILE__, __LINE__,			\
+		"%s == %s",#e1,#e2)) || (diag("Expected: %zd",(size_t)e2),diag("     Got: %zd",(size_t)e1),0)) /* diag is void; note commas. */
+
+#endif