base64.c

/*
 * Copyright (c) 2000, 2001, 2002 X-Way Rights BV
 *
 * This library is free software; you can redistribute it and/or
 * modify it under the terms of the GNU Lesser General Public
 * License as published by the Free Software Foundation; either
 * version 2.1 of the License, or (at your option) any later version.
 *
 * This library is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
 * Lesser General Public License for more details.
 *
 * You should have received a copy of the GNU Lesser General Public
 * License along with this library; if not, write to the Free Software
 * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
 *
 */

/*!\file base64.c
 * \brief Base64 encoding and decoding.
 * \author Bob Deblier <bob.deblier@telenet.be>
 */

#define BEECRYPT_DLL_EXPORT

#if HAVE_CONFIG_H
# include "config.h"
#endif

#include "beecrypt/base64.h"

#if HAVE_ENDIAN_H && HAVE_ASM_BYTEORDER_H
# include <endian.h>
#endif

#include "beecrypt/endianness.h"

#if HAVE_CTYPE_H
# include <ctype.h>
#endif

static const char* to_b64 = "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/";

/* encode 64 characters per line */
#define CHARS_PER_LINE	64

char* b64enc(const memchunk* chunk)
{
	int div = chunk->size / 3;
	int rem = chunk->size % 3;
	int chars = div*4 + rem + 1;
	int newlines = (chars + CHARS_PER_LINE - 1) / CHARS_PER_LINE;

	const byte* data = chunk->data;
	char* string = (char*) malloc(chars + newlines + 1);

	if (string)
	{
		register char* buf = string;

		chars = 0;

		while (div > 0)
		{
			buf[0] = to_b64[ (data[0] >> 2) & 0x3f];
			buf[1] = to_b64[((data[0] << 4) & 0x30) | ((data[1] >> 4) & 0xf)];
			buf[2] = to_b64[((data[1] << 2) & 0x3c) | ((data[2] >> 6) & 0x3)];
			buf[3] = to_b64[  data[2] & 0x3f];
			data += 3;
			buf += 4;
			div--;
			chars += 4;
			if (chars == CHARS_PER_LINE)
			{
				chars = 0;
				*(buf++) = '\n';
			}
		}

		switch (rem)
		{
		case 2:
			buf[0] = to_b64[ (data[0] >> 2) & 0x3f];
			buf[1] = to_b64[((data[0] << 4) & 0x30) + ((data[1] >> 4) & 0xf)];
			buf[2] = to_b64[ (data[1] << 2) & 0x3c];
			buf[3] = '=';
			buf += 4;
			chars += 4;
			break;
		case 1:
			buf[0] = to_b64[ (data[0] >> 2) & 0x3f];
			buf[1] = to_b64[ (data[0] << 4) & 0x30];
			buf[2] = '=';
			buf[3] = '=';
			buf += 4;
			chars += 4;
			break;
		}

	/* 	*(buf++) = '\n'; This would result in a buffer overrun */
		*buf = '\0';
	}

	return string;
}

memchunk* b64dec(const char* string)
{
	/* return a decoded memchunk, or a null pointer in case of failure */

	memchunk* rc = 0;

	if (string)
	{
		register int length = strlen(string);

		/* do a format verification first */
		if (length > 0)
		{
			register int count = 0, rem = 0;
			register const char* tmp = string;

			while (length > 0)
			{
				register int skip = strspn(tmp, to_b64);
				count += skip;
				length -= skip;
				tmp += skip;
				if (length > 0)
				{
					register int i, vrfy = strcspn(tmp, to_b64);

					for (i = 0; i < vrfy; i++)
					{
						if (isspace(tmp[i]))
							continue;

						if (tmp[i] == '=')
						{
							/* we should check if we're close to the end of the string */
							rem = count % 4;

							/* rem must be either 2 or 3, otherwise no '=' should be here */
							if (rem < 2)
								return 0;

							/* end-of-message recognized */
							break;
						}
						else
						{
							/* Transmission error; RFC tells us to ignore this, but:
							 *  - the rest of the message is going to even more corrupt since we're sliding bits out of place
							 * If a message is corrupt, it should be dropped. Period.
							 */

							return 0;
						}
					}

					length -= vrfy;
					tmp += vrfy;
				}
			}

			rc = memchunkAlloc((count / 4) * 3 + (rem ? (rem - 1) : 0));

			if (rc)
			{
				if (count > 0)
				{
					register int i, qw = 0, tw = 0;
					register byte* data = rc->data;

					length = strlen(tmp = string);

					for (i = 0; i < length; i++)
					{
						register char ch = string[i];
						register byte bits = 0;

						if (isspace(ch))
							continue;

						if ((ch >= 'A') && (ch <= 'Z'))
						{
							bits = (byte) (ch - 'A');
						}
						else if ((ch >= 'a') && (ch <= 'z'))
						{
							bits = (byte) (ch - 'a' + 26);
						}
						else if ((ch >= '0') && (ch <= '9'))
						{
							bits = (byte) (ch - '0' + 52);
						}
						else if (ch == '+')
						{
							bits = 62;
						}
						else if (ch == '/')
						{
							bits = 63;
						}
						else if (ch == '=')
							break;

						switch (qw++)
						{
						case 0:
							data[tw+0] = (bits << 2) & 0xfc;
							break;
						case 1:
							data[tw+0] |= (bits >> 4) & 0x03;
							data[tw+1] = (bits << 4) & 0xf0;
							break;
						case 2:
							data[tw+1] |= (bits >> 2) & 0x0f;
							data[tw+2] = (bits << 6) & 0xc0;
							break;
						case 3:
							data[tw+2] |= bits & 0x3f;
							break;
						}

						if (qw == 4)
						{
							qw = 0;
							tw += 3;
						}
					}
				}
			}
		}
	}

	return rc;
}

int b64encode_chars_per_line = B64ENCODE_CHARS_PER_LINE;

const char * b64encode_eolstr = B64ENCODE_EOLSTR;

char* b64encode(const void* data, size_t ns)
{
	static char b64enc[] = "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/";
	const char *e;
	const unsigned char *s = data;
	unsigned char *t, *te;
	int nt;
	int lc;
	unsigned c;

	if (s == NULL) return NULL;

	if (ns == 0) ns = strlen((const char*) s);
	nt = ((ns + 2) / 3) * 4;

	/* Add additional bytes necessary for eol string(s). */
	if (b64encode_chars_per_line > 0 && b64encode_eolstr != NULL)
	{
		lc = (nt + b64encode_chars_per_line - 1) / b64encode_chars_per_line;
		if (((nt + b64encode_chars_per_line - 1) % b64encode_chars_per_line) != 0)
			++lc;
		nt += lc * strlen(b64encode_eolstr);
	}

	t = te = malloc(nt + 1);

	lc = 0;
	if (te)
	{
		while (ns > 0)
		{
			c = *s++;
			*te++ = b64enc[ (c >> 2) ], lc++;
			*te++ = b64enc[ ((c & 0x3) << 4) | (((ns-1) > 0 ? *s : 0) >> 4) ], lc++; 
			if (--ns == 0)
			{
				*te++ = '=';
				*te++ = '=';
				continue;
			}
			c = *s++;
			*te++ = b64enc[ ((c & 0xf) << 2) | (((ns-1) > 0 ? *s : 0) >> 6) ], lc++;
			if (--ns == 0)
			{
				*te++ = '=';
				continue;
			}
			*te++ = b64enc[ (int)(*s & 0x3f) ], lc++;

			/* Append eol string if desired. */
			if (b64encode_chars_per_line > 0 && b64encode_eolstr != NULL)
			{
				if (lc >= b64encode_chars_per_line)
				{
					for (e = b64encode_eolstr; *e != '\0'; e++)
						*te++ = *e;
					lc = 0;
				}
			}
			s++;
			--ns;
		}

		/* Append eol string if desired. */
		if (b64encode_chars_per_line > 0 && b64encode_eolstr != NULL)
		{
			if (lc != 0)
			{
				for (e = b64encode_eolstr; *e != '\0'; e++)
					*te++ = *e;
			}
		}
		*te = '\0';
	}

	return (char*) t;
}

#define CRC24_INIT 0xb704ceL
#define CRC24_POLY 0x1864cfbL

char* b64crc (const unsigned char* data, size_t ns)
{
	const unsigned char *s = data;
	uint32_t crc = CRC24_INIT;

	while (ns-- > 0)
	{
		int i;
		crc ^= (*s++) << 16;
		for (i = 0; i < 8; i++)
		{
			crc <<= 1;
			if (crc & 0x1000000)
			crc ^= CRC24_POLY;
		}
	}
	crc &= 0xffffff;
	#if !WORDS_BIGENDIAN
	crc = swapu32(crc);
	#endif
	data = (byte *)&crc;
	data++;
	ns = 3;

	return b64encode(data, ns);
}

const char* b64decode_whitespace = B64DECODE_WHITESPACE;

int b64decode(const char* s, void** datap, size_t* lenp)
{
	unsigned char b64dec[256];
	const unsigned char *t;
	unsigned char *te;
	int ns, nt;
	unsigned a, b, c, d;

	if (s == NULL)	return 1;

	/* Setup character lookup tables. */
	memset(b64dec, 0x80, sizeof(b64dec));
	for (c = 'A'; c <= 'Z'; c++)
		b64dec[ c ] = 0 + (c - 'A');
	for (c = 'a'; c <= 'z'; c++)
		b64dec[ c ] = 26 + (c - 'a');
	for (c = '0'; c <= '9'; c++)
		b64dec[ c ] = 52 + (c - '0');
	b64dec[(unsigned)'+'] = 62;
	b64dec[(unsigned)'/'] = 63;
	b64dec[(unsigned)'='] = 0;

	/* Mark whitespace characters. */
	if (b64decode_whitespace)
	{
		const char *e;
		for (e = b64decode_whitespace; *e != '\0'; e++)
		{
			if (b64dec[ (unsigned)*e ] == 0x80)
				b64dec[ (unsigned)*e ] = 0x81;
		}
	}
	
	/* Validate input buffer */
	ns = 0;
	for (t = (unsigned char*) s; *t != '\0'; t++)
	{
		switch (b64dec[(unsigned) *t])
		{
		case 0x80:	/* invalid chararcter */
			return 3;
		case 0x81:	/* white space */
			break;
		default:
			ns++;
			break;
		}
	}
	
	if (((unsigned) ns) & 0x3)	return 2;

	nt = (ns / 4) * 3;
	t = te = malloc(nt + 1);

	while (ns > 0)
	{
		/* Get next 4 characters, ignoring whitespace. */
		while ((a = b64dec[ (unsigned)*s++ ]) == 0x81)
			;
		while ((b = b64dec[ (unsigned)*s++ ]) == 0x81)
			;
		while ((c = b64dec[ (unsigned)*s++ ]) == 0x81)
			;
		while ((d = b64dec[ (unsigned)*s++ ]) == 0x81)
			;

		ns -= 4;
		*te++ = (a << 2) | (b >> 4);
		if (s[-2] == '=') break;
		*te++ = (b << 4) | (c >> 2);
		if (s[-1] == '=') break;
		*te++ = (c << 6) | d;
	}

	if (ns != 0)
	{	/* XXX can't happen, just in case */
		if (t) free((void *)t);
		return 1;
	}
	if (lenp)
		*lenp = (te - t);

	if (datap)
		*datap = (void *)t;
	else
		if (t) free((void *)t);

	return 0;
}