busybox/libbb/uuencode.c

224 lines
5.8 KiB
C
Raw Normal View History

/* vi: set sw=4 ts=4: */
/*
* Copyright 2003, Glenn McGrath
* Copyright 2006, Rob Landley <rob@landley.net>
* Copyright 2010, Denys Vlasenko
*
* Licensed under GPLv2 or later, see file LICENSE in this source tree.
*/
#include "libbb.h"
/* Conversion table. for base 64 */
const char bb_uuenc_tbl_base64[65 + 1] ALIGN1 = {
'A', 'B', 'C', 'D', 'E', 'F', 'G', 'H',
'I', 'J', 'K', 'L', 'M', 'N', 'O', 'P',
'Q', 'R', 'S', 'T', 'U', 'V', 'W', 'X',
'Y', 'Z', 'a', 'b', 'c', 'd', 'e', 'f',
'g', 'h', 'i', 'j', 'k', 'l', 'm', 'n',
'o', 'p', 'q', 'r', 's', 't', 'u', 'v',
'w', 'x', 'y', 'z', '0', '1', '2', '3',
'4', '5', '6', '7', '8', '9', '+', '/',
'=' /* termination character */,
'\0' /* needed for uudecode.c only */
};
const char bb_uuenc_tbl_std[65] ALIGN1 = {
'`', '!', '"', '#', '$', '%', '&', '\'',
'(', ')', '*', '+', ',', '-', '.', '/',
'0', '1', '2', '3', '4', '5', '6', '7',
'8', '9', ':', ';', '<', '=', '>', '?',
'@', 'A', 'B', 'C', 'D', 'E', 'F', 'G',
'H', 'I', 'J', 'K', 'L', 'M', 'N', 'O',
'P', 'Q', 'R', 'S', 'T', 'U', 'V', 'W',
'X', 'Y', 'Z', '[', '\\', ']', '^', '_',
'`' /* termination character */
};
/*
* Encode bytes at S of length LENGTH to uuencode or base64 format and place it
* to STORE. STORE will be 0-terminated, and must point to a writable
* buffer of at least 1+BASE64_LENGTH(length) bytes.
* where BASE64_LENGTH(len) = (4 * ((LENGTH + 2) / 3))
*/
void FAST_FUNC bb_uuencode(char *p, const void *src, int length, const char *tbl)
{
const unsigned char *s = src;
/* Transform the 3x8 bits to 4x6 bits */
while (length > 0) {
unsigned s1, s2;
/* Are s[1], s[2] valid or should be assumed 0? */
s1 = s2 = 0;
length -= 3; /* can be >=0, -1, -2 */
if (length >= -1) {
s1 = s[1];
if (length >= 0)
s2 = s[2];
}
*p++ = tbl[s[0] >> 2];
*p++ = tbl[((s[0] & 3) << 4) + (s1 >> 4)];
*p++ = tbl[((s1 & 0xf) << 2) + (s2 >> 6)];
*p++ = tbl[s2 & 0x3f];
s += 3;
}
/* Zero-terminate */
*p = '\0';
/* If length is -2 or -1, pad last char or two */
while (length) {
*--p = tbl[64];
length++;
}
}
/*
* Decode base64 encoded string. Stops on '\0'.
*
* Returns: pointer to the undecoded part of source.
* If points to '\0', then the source was fully decoded.
* (*pp_dst): advanced past the last written byte.
*/
const char* FAST_FUNC decode_base64(char **pp_dst, const char *src)
{
char *dst = *pp_dst;
const char *src_tail;
while (1) {
unsigned char six_bit[4];
int count = 0;
/* Fetch up to four 6-bit values */
src_tail = src;
while (count < 4) {
char *table_ptr;
int ch;
/* Get next _valid_ character.
* bb_uuenc_tbl_base64[] contains this string:
* 0 1 2 3 4 5 6
* 01234567890123456789012345678901234567890123456789012345678901234
* "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/="
*/
do {
ch = *src;
if (ch == '\0') {
if (count == 0) {
/* Example:
* If we decode "QUJD <NUL>", we want
* to return ptr to NUL, not to ' ',
* because we did fully decode
* the string (to "ABC").
*/
src_tail = src;
}
goto ret;
}
src++;
table_ptr = strchr(bb_uuenc_tbl_base64, ch);
//TODO: add BASE64_FLAG_foo to die on bad char?
} while (!table_ptr);
/* Convert encoded character to decimal */
ch = table_ptr - bb_uuenc_tbl_base64;
/* ch is 64 if char was '=', otherwise 0..63 */
if (ch == 64)
break;
six_bit[count] = ch;
count++;
}
/* Transform 6-bit values to 8-bit ones.
* count can be < 4 when we decode the tail:
* "eQ==" -> "y", not "y NUL NUL".
* Note that (count > 1) is always true,
* "x===" encoding is not valid:
* even a single zero byte encodes as "AA==".
* However, with current logic we come here with count == 1
* when we decode "==" tail.
*/
if (count > 1)
*dst++ = six_bit[0] << 2 | six_bit[1] >> 4;
if (count > 2)
*dst++ = six_bit[1] << 4 | six_bit[2] >> 2;
if (count > 3)
*dst++ = six_bit[2] << 6 | six_bit[3];
/* Note that if we decode "AA==" and ate first '=',
* we just decoded one char (count == 2) and now we'll
* do the loop once more to decode second '='.
*/
} /* while (1) */
ret:
*pp_dst = dst;
return src_tail;
}
/*
* Decode base64 encoded stream.
* Can stop on EOF, specified char, or on uuencode-style "====" line:
* flags argument controls it.
*/
void FAST_FUNC read_base64(FILE *src_stream, FILE *dst_stream, int flags)
{
/* Note that EOF _can_ be passed as exit_char too */
#define exit_char ((int)(signed char)flags)
#define uu_style_end (flags & BASE64_FLAG_UU_STOP)
/* uuencoded files have 61 byte lines. Use 64 byte buffer
* to process line at a time.
*/
enum { BUFFER_SIZE = 64 };
char in_buf[BUFFER_SIZE + 2];
char out_buf[BUFFER_SIZE / 4 * 3 + 2];
char *out_tail;
const char *in_tail;
int term_seen = 0;
int in_count = 0;
while (1) {
while (in_count < BUFFER_SIZE) {
int ch = fgetc(src_stream);
if (ch == exit_char) {
if (in_count == 0)
return;
term_seen = 1;
break;
}
if (ch == EOF) {
term_seen = 1;
break;
}
/* Prevent "====" line to be split: stop if we see '\n'.
* We can also skip other whitespace and skirt the problem
* of files with NULs by stopping on any control char or space:
*/
if (ch <= ' ')
break;
in_buf[in_count++] = ch;
}
in_buf[in_count] = '\0';
/* Did we encounter "====" line? */
if (uu_style_end && strcmp(in_buf, "====") == 0)
return;
out_tail = out_buf;
in_tail = decode_base64(&out_tail, in_buf);
fwrite(out_buf, (out_tail - out_buf), 1, dst_stream);
if (term_seen) {
/* Did we consume ALL characters? */
if (*in_tail == '\0')
return;
/* No */
libbb: reduce the overhead of single parameter bb_error_msg() calls Back in 2007, commit 0c97c9d43707 ("'simple' error message functions by Loic Grenie") introduced bb_simple_perror_msg() to allow for a lower overhead call to bb_perror_msg() when only a string was being printed with no parameters. This saves space for some CPU architectures because it avoids the overhead of a call to a variadic function. However there has never been a simple version of bb_error_msg(), and since 2007 many new calls to bb_perror_msg() have been added that only take a single parameter and so could have been using bb_simple_perror_message(). This changeset introduces 'simple' versions of bb_info_msg(), bb_error_msg(), bb_error_msg_and_die(), bb_herror_msg() and bb_herror_msg_and_die(), and replaces all calls that only take a single parameter, or use something like ("%s", arg), with calls to the corresponding 'simple' version. Since it is likely that single parameter calls to the variadic functions may be accidentally reintroduced in the future a new debugging config option WARN_SIMPLE_MSG has been introduced. This uses some macro magic which will cause any such calls to generate a warning, but this is turned off by default to avoid use of the unpleasant macros in normal circumstances. This is a large changeset due to the number of calls that have been replaced. The only files that contain changes other than simple substitution of function calls are libbb.h, libbb/herror_msg.c, libbb/verror_msg.c and libbb/xfuncs_printf.c. In miscutils/devfsd.c, networking/udhcp/common.h and util-linux/mdev.c additonal macros have been added for logging so that single parameter and multiple parameter logging variants exist. The amount of space saved varies considerably by architecture, and was found to be as follows (for 'defconfig' using GCC 7.4): Arm: -92 bytes MIPS: -52 bytes PPC: -1836 bytes x86_64: -938 bytes Note that for the MIPS architecture only an exception had to be made disabling the 'simple' calls for 'udhcp' (in networking/udhcp/common.h) because it made these files larger on MIPS. Signed-off-by: James Byrne <james.byrne@origamienergy.com> Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
2019-07-02 15:05:03 +05:30
bb_simple_error_msg_and_die("truncated base64 input");
}
/* It was partial decode */
in_count = strlen(in_tail);
memmove(in_buf, in_tail, in_count);
}
}