Hexadecimal string to byte array in C
CStringC Problem Overview
Is there any standard C function that converts from hexadecimal string to byte array?
I do not want to write my own function.
C Solutions
Solution 1 - C
As far as I know, there's no standard function to do so, but it's simple to achieve in the following manner:
#include <stdio.h>
int main(int argc, char **argv) {
const char hexstring[] = "DEadbeef10203040b00b1e50", *pos = hexstring;
unsigned char val[12];
/* WARNING: no sanitization or error-checking whatsoever */
for (size_t count = 0; count < sizeof val/sizeof *val; count++) {
sscanf(pos, "%2hhx", &val[count]);
pos += 2;
}
printf("0x");
for(size_t count = 0; count < sizeof val/sizeof *val; count++)
printf("%02x", val[count]);
printf("\n");
return 0;
}
Edit
As Al pointed out, in case of an odd number of hex digits in the string, you have to make sure you prefix it with a starting 0. For example, the string "f00f5" will be evaluated as {0xf0, 0x0f, 0x05} erroneously by the above example, instead of the proper {0x0f, 0x00, 0xf5}.
Amended the example a little bit to address the comment from @MassimoCallegari
Solution 2 - C
I found this question by Googling for the same thing. I don't like the idea of calling sscanf() or strtol() since it feels like overkill. I wrote a quick function which does not validate that the text is indeed the hexadecimal presentation of a byte stream, but will handle odd number of hex digits:
uint8_t tallymarker_hextobin(const char * str, uint8_t * bytes, size_t blen)
{
uint8_t pos;
uint8_t idx0;
uint8_t idx1;
// mapping of ASCII characters to hex values
const uint8_t hashmap[] =
{
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, // ........
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, // ........
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, // ........
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, // ........
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, // !"#$%&'
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, // ()*+,-./
0x00, 0x01, 0x02, 0x03, 0x04, 0x05, 0x06, 0x07, // 01234567
0x08, 0x09, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, // 89:;<=>?
0x00, 0x0a, 0x0b, 0x0c, 0x0d, 0x0e, 0x0f, 0x00, // @ABCDEFG
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, // HIJKLMNO
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, // PQRSTUVW
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, // XYZ[\]^_
0x00, 0x0a, 0x0b, 0x0c, 0x0d, 0x0e, 0x0f, 0x00, // `abcdefg
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, // hijklmno
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, // pqrstuvw
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, // xyz{|}~.
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, // ........
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, // ........
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, // ........
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, // ........
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, // ........
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, // ........
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, // ........
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, // ........
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, // ........
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, // ........
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, // ........
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, // ........
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, // ........
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, // ........
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, // ........
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00 // ........
};
bzero(bytes, blen);
for (pos = 0; ((pos < (blen*2)) && (pos < strlen(str))); pos += 2)
{
idx0 = (uint8_t)str[pos+0];
idx1 = (uint8_t)str[pos+1];
bytes[pos/2] = (uint8_t)(hashmap[idx0] << 4) | hashmap[idx1];
};
return(0);
}
Solution 3 - C
Apart from the excellent answers above I though I would write a C function that does not use any libraries and has some guards against bad strings.
uint8_t* datahex(char* string) {
if(string == NULL)
return NULL;
size_t slength = strlen(string);
if((slength % 2) != 0) // must be even
return NULL;
size_t dlength = slength / 2;
uint8_t* data = malloc(dlength);
memset(data, 0, dlength);
size_t index = 0;
while (index < slength) {
char c = string[index];
int value = 0;
if(c >= '0' && c <= '9')
value = (c - '0');
else if (c >= 'A' && c <= 'F')
value = (10 + (c - 'A'));
else if (c >= 'a' && c <= 'f')
value = (10 + (c - 'a'));
else {
free(data);
return NULL;
}
data[(index/2)] += value << (((index + 1) % 2) * 4);
index++;
}
return data;
}
Explanation:
a. index / 2 | Division between integers will round down the value, so 0/2 = 0, 1/2 = 0, 2/2 = 1, 3/2 = 1, 4/2 = 2, 5/2 = 2, etc. So, for every 2 string characters we add the value to 1 data byte.
b. (index + 1) % 2 | We want odd numbers to result to 1 and even to 0 since the first digit of a hex string is the most significant and needs to be multiplied by 16. so for index 0 => 0 + 1 % 2 = 1, index 1 => 1 + 1 % 2 = 0 etc.
c. << 4 | Shift by 4 is multiplying by 16. example: b00000001 << 4 = b00010000
Solution 4 - C
For short strings, strtol, strtoll, and strtoimax will work just fine (note that the third argument is the base to use in processing the string...set it to 16). If your input is longer than number-of-bits-in-the-longest-integer-type/4 then you'll need one of the more flexible methods suggested by other answers.
Solution 5 - C
By some modification form user411313's code, following works for me:
#include <stdio.h>
#include <stdint.h>
#include <string.h>
int main ()
{
char *hexstring = "deadbeef10203040b00b1e50";
int i;
unsigned int bytearray[12];
uint8_t str_len = strlen(hexstring);
for (i = 0; i < (str_len / 2); i++) {
sscanf(hexstring + 2*i, "%02x", &bytearray[i]);
printf("bytearray %d: %02x\n", i, bytearray[i]);
}
return 0;
}
Solution 6 - C
A fleshed out version of Michael Foukarakis post (since I don't have the "reputation" to add a comment to that post yet):
#include <stdio.h>
#include <string.h>
void print(unsigned char *byte_array, int byte_array_size)
{
int i = 0;
printf("0x");
for(; i < byte_array_size; i++)
{
printf("%02x", byte_array[i]);
}
printf("\n");
}
int convert(const char *hex_str, unsigned char *byte_array, int byte_array_max)
{
int hex_str_len = strlen(hex_str);
int i = 0, j = 0;
// The output array size is half the hex_str length (rounded up)
int byte_array_size = (hex_str_len+1)/2;
if (byte_array_size > byte_array_max)
{
// Too big for the output array
return -1;
}
if (hex_str_len % 2 == 1)
{
// hex_str is an odd length, so assume an implicit "0" prefix
if (sscanf(&(hex_str[0]), "%1hhx", &(byte_array[0])) != 1)
{
return -1;
}
i = j = 1;
}
for (; i < hex_str_len; i+=2, j++)
{
if (sscanf(&(hex_str[i]), "%2hhx", &(byte_array[j])) != 1)
{
return -1;
}
}
return byte_array_size;
}
void main()
{
char *examples[] = { "", "5", "D", "5D", "5Df", "deadbeef10203040b00b1e50", "02invalid55" };
unsigned char byte_array[128];
int i = 0;
for (; i < sizeof(examples)/sizeof(char *); i++)
{
int size = convert(examples[i], byte_array, 128);
if (size < 0)
{
printf("Failed to convert '%s'\n", examples[i]);
}
else if (size == 0)
{
printf("Nothing to convert for '%s'\n", examples[i]);
}
else
{
print(byte_array, size);
}
}
}
Solution 7 - C
Here is HexToBin and BinToHex relatively clean and readable. (Note originally there were returned enum error codes through an error logging system not a simple -1 or -2.)
typedef unsigned char ByteData;
ByteData HexChar (char c)
{
if ('0' <= c && c <= '9') return (ByteData)(c - '0');
if ('A' <= c && c <= 'F') return (ByteData)(c - 'A' + 10);
if ('a' <= c && c <= 'f') return (ByteData)(c - 'a' + 10);
return (ByteData)(-1);
}
ssize_t HexToBin (const char* s, ByteData * buff, ssize_t length)
{
ssize_t result = 0;
if (!s || !buff || length <= 0) return -2;
while (*s)
{
ByteData nib1 = HexChar(*s++);
if ((signed)nib1 < 0) return -3;
ByteData nib2 = HexChar(*s++);
if ((signed)nib2 < 0) return -4;
ByteData bin = (nib1 << 4) + nib2;
if (length-- <= 0) return -5;
*buff++ = bin;
++result;
}
return result;
}
void BinToHex (const ByteData * buff, ssize_t length, char * output, ssize_t outLength)
{
char binHex[] = "0123456789ABCDEF";
if (!output || outLength < 4) return (void)(-6);
*output = '\0';
if (!buff || length <= 0 || outLength <= 2 * length)
{
memcpy(output, "ERR", 4);
return (void)(-7);
}
for (; length > 0; --length, outLength -= 2)
{
ByteData byte = *buff++;
*output++ = binHex[(byte >> 4) & 0x0F];
*output++ = binHex[byte & 0x0F];
}
if (outLength-- <= 0) return (void)(-8);
*output++ = '\0';
}
Solution 8 - C
hextools.h
#ifndef HEX_TOOLS_H
#define HEX_TOOLS_H
char *bin2hex(unsigned char*, int);
unsigned char *hex2bin(const char*);
#endif // HEX_TOOLS_H
hextools.c
#include <stdlib.h>
char *bin2hex(unsigned char *p, int len)
{
char *hex = malloc(((2*len) + 1));
char *r = hex;
while(len && p)
{
(*r) = ((*p) & 0xF0) >> 4;
(*r) = ((*r) <= 9 ? '0' + (*r) : 'A' - 10 + (*r));
r++;
(*r) = ((*p) & 0x0F);
(*r) = ((*r) <= 9 ? '0' + (*r) : 'A' - 10 + (*r));
r++;
p++;
len--;
}
*r = '\0';
return hex;
}
unsigned char *hex2bin(const char *str)
{
int len, h;
unsigned char *result, *err, *p, c;
err = malloc(1);
*err = 0;
if (!str)
return err;
if (!*str)
return err;
len = 0;
p = (unsigned char*) str;
while (*p++)
len++;
result = malloc((len/2)+1);
h = !(len%2) * 4;
p = result;
*p = 0;
c = *str;
while(c)
{
if(('0' <= c) && (c <= '9'))
*p += (c - '0') << h;
else if(('A' <= c) && (c <= 'F'))
*p += (c - 'A' + 10) << h;
else if(('a' <= c) && (c <= 'f'))
*p += (c - 'a' + 10) << h;
else
return err;
str++;
c = *str;
if (h)
h = 0;
else
{
h = 4;
p++;
*p = 0;
}
}
return result;
}
main.c
#include <stdio.h>
#include "hextools.h"
int main(void)
{
unsigned char s[] = { 0xa0, 0xf9, 0xc3, 0xde, 0x44 };
char *hex = bin2hex(s, sizeof s);
puts(hex);
unsigned char *bin;
bin = hex2bin(hex);
puts(bin2hex(bin, 5));
size_t k;
for(k=0; k<5; k++)
printf("%02X", bin[k]);
putchar('\n');
return 0;
}
Solution 9 - C
Following is the solution I wrote up for performance reasons:
void hex2bin(const char* in, size_t len, unsigned char* out) {
static const unsigned char TBL[] = {
0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 58, 59, 60, 61,
62, 63, 64, 10, 11, 12, 13, 14, 15, 71, 72, 73, 74, 75,
76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89,
90, 91, 92, 93, 94, 95, 96, 10, 11, 12, 13, 14, 15
};
static const unsigned char *LOOKUP = TBL - 48;
const char* end = in + len;
while(in < end) *(out++) = LOOKUP[*(in++)] << 4 | LOOKUP[*(in++)];
}
Example:
unsigned char seckey[32];
hex2bin("351aaaec0070d13d350afae2bc43b68c7e590268889869dde489f2f7988f3fee", 64, seckey);
/*
seckey = {
53, 26, 170, 236, 0, 112, 209, 61, 53, 10, 250, 226, 188, 67, 182, 140,
126, 89, 2, 104, 136, 152, 105, 221, 228, 137, 242, 247, 152, 143, 63, 238
};
*/
If you don't need to support lowercase:
static const unsigned char TBL[] = {
0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 58, 59,
60, 61, 62, 63, 64, 10, 11, 12, 13, 14, 15
};
Solution 10 - C
you can use this function which
- written with performance in mind (for embedded processors), no
scanf,strtolordynamic memory allocation - has guard against output buffer overflow and odd input str len
/// in: valid chars are 0-9 + A-F + a-f
/// out_len_max==0: convert until the end of input string, out_len_max>0 only convert this many numbers
/// returns actual out size
int hexStr2Arr(unsigned char* out, const char* in, size_t out_len_max = 0)
{
if (!out_len_max)
out_len_max = 2147483647; // INT_MAX
const int in_len = strnlen(in, out_len_max * 2);
if (in_len % 2 != 0)
return -1; // error, in str len should be even
// calc actual out len
const int out_len = out_len_max < (in_len / 2) ? out_len_max : (in_len / 2);
for (int i = 0; i < out_len; i++) {
char ch0 = in[2 * i];
char ch1 = in[2 * i + 1];
uint8_t nib0 = (ch0 & 0xF) + (ch0 >> 6) | ((ch0 >> 3) & 0x8);
uint8_t nib1 = (ch1 & 0xF) + (ch1 >> 6) | ((ch1 >> 3) & 0x8);
out[i] = (nib0 << 4) | nib1;
}
return out_len;
}
usage:
unsigned char result[128];
memset(result, 0, 128); // optional
printf("result len=%d\n", hexStr2Arr(result, "0a0B10")); // result = [0A 0B 10 00 00 ...]
memset(result, 0, 128); // optional
// only convert single number
printf("result len=%d\n", hexStr2Arr(result, "0a0B10", 1)); // result = [0A 00 00 00 00 ...]
Solution 11 - C
char *hexstring = "deadbeef10203040b00b1e50", *pos = hexstring;
unsigned char val[12];
while( *pos )
{
if( !((pos-hexstring)&1) )
sscanf(pos,"%02x",&val[(pos-hexstring)>>1]);
++pos;
}
sizeof(val)/sizeof(val[0]) is redundant!
Solution 12 - C
In main()
{
printf("enter string :\n");
fgets(buf, 200, stdin);
unsigned char str_len = strlen(buf);
k=0;
unsigned char bytearray[100];
for(j=0;j<str_len-1;j=j+2)
{ bytearray[k++]=converttohex(&buffer[j]);
printf(" %02X",bytearray[k-1]);
}
}
Use this
int converttohex(char * val)
{
unsigned char temp = toupper(*val);
unsigned char fin=0;
if(temp>64)
temp=10+(temp-65);
else
temp=temp-48;
fin=(temp<<4)&0xf0;
temp = toupper(*(val+1));
if(temp>64)
temp=10+(temp-65);
else
temp=temp-48;
fin=fin|(temp & 0x0f);
return fin;
}
Solution 13 - C
This is a modified function from a similar question, modified as per the suggestion of https://stackoverflow.com/a/18267932/700597.
This function will convert a hexadecimal string - NOT prepended with "0x" - with an even number of characters to the number of bytes specified. It will return -1 if it encounters an invalid character, or if the hex string has an odd length, and 0 on success.
//convert hexstring to len bytes of data
//returns 0 on success, -1 on error
//data is a buffer of at least len bytes
//hexstring is upper or lower case hexadecimal, NOT prepended with "0x"
int hex2data(unsigned char *data, const unsigned char *hexstring, unsigned int len)
{
unsigned const char *pos = hexstring;
char *endptr;
size_t count = 0;
if ((hexstring[0] == '\0') || (strlen(hexstring) % 2)) {
//hexstring contains no data
//or hexstring has an odd length
return -1;
}
for(count = 0; count < len; count++) {
char buf[5] = {'0', 'x', pos[0], pos[1], 0};
data[count] = strtol(buf, &endptr, 0);
pos += 2 * sizeof(char);
if (endptr[0] != '\0') {
//non-hexadecimal character encountered
return -1;
}
}
return 0;
}
Solution 14 - C
Two short routines to parse a byte or a word, using strchr().
// HexConverter.h
#ifndef HEXCONVERTER_H
#define HEXCONVERTER_H
unsigned int hexToByte (const char *hexString);
unsigned int hexToWord (const char *hexString);
#endif
// HexConverter.c
#include <string.h> // for strchr()
#include <ctype.h> // for toupper()
unsigned int hexToByte (const char *hexString)
{
unsigned int value;
const char *hexDigits = "0123456789ABCDEF";
value = 0;
if (hexString != NULL)
{
char *ptr;
ptr = strchr (hexDigits, toupper(hexString[0]));
if (ptr != NULL)
{
value = (ptr - hexDigits) << 4;
ptr = strchr (hexDigits, toupper(hexString[1]));
if (ptr != NULL)
{
value = value | (ptr - hexDigits);
}
}
}
return value;
}
unsigned int hexToWord (const char *hexString)
{
unsigned int value;
value = 0;
if (hexString != NULL)
{
value = (hexToByte (&hexString[0]) << 8) |
(hexToByte (&hexString[2]));
}
return value;
}
// HexConverterTest.c
#include <stdio.h>
#include "HexConverter.h"
int main (int argc, char **argv)
{
(void)argc; // not used
(void)argv; // not used
unsigned int value;
char *hexString;
hexString = "2a";
value = hexToByte (hexString);
printf ("%s == %x (%u)\n", hexString, value, value);
hexString = "1234";
value = hexToWord (hexString);
printf ("%s == %x (%u)\n", hexString, value, value);
hexString = "0102030405060708090a10ff";
printf ("Hex String: %s\n", hexString);
for (unsigned int idx = 0; idx < strlen(hexString); idx += 2)
{
value = hexToByte (&hexString[idx]);
printf ("%c%c == %x (%u)\n", hexString[idx], hexString[idx+1],
value, value);
}
return EXIT_SUCCESS;
}
Solution 15 - C
Try the following code:
static unsigned char ascii2byte(char *val)
{
unsigned char temp = *val;
if(temp > 0x60) temp -= 39; // convert chars a-f
temp -= 48; // convert chars 0-9
temp *= 16;
temp += *(val+1);
if(*(val+1) > 0x60) temp -= 39; // convert chars a-f
temp -= 48; // convert chars 0-9
return temp;
}
Solution 16 - C
Here's my version:
/* Convert a hex char digit to its integer value. */
int hexDigitToInt(char digit) {
digit = tolower(digit);
if ('0' <= digit && digit <= '9') //if it's decimal
return (int)(digit - '0');
else if ('a' <= digit && digit <= 'f') //if it's abcdef
return (int)(digit - ('a' - 10));
else
return -1; //value not in [0-9][a-f] range
}
/* Decode a hex string. */
char *decodeHexString(const char *hexStr) {
char* decoded = malloc(strlen(hexStr)/2+1);
char* hexStrPtr = (char *)hexStr;
char* decodedPtr = decoded;
while (*hexStrPtr != '\0') { /* Step through hexStr, two chars at a time. */
*decodedPtr = 16 * hexDigitToInt(*hexStrPtr) + hexDigitToInt(*(hexStrPtr+1));
hexStrPtr += 2;
decodedPtr++;
}
*decodedPtr = '\0'; /* final null char */
return decoded;
}
Solution 17 - C
Could it be simpler?!
uint8_t hex(char ch) {
uint8_t r = (ch > 57) ? (ch - 55) : (ch - 48);
return r & 0x0F;
}
int to_byte_array(const char *in, size_t in_size, uint8_t *out) {
int count = 0;
if (in_size % 2) {
while (*in && out) {
*out = hex(*in++);
if (!*in)
return count;
*out = (*out << 4) | hex(*in++);
*out++;
count++;
}
return count;
} else {
while (*in && out) {
*out++ = (hex(*in++) << 4) | hex(*in++);
count++;
}
return count;
}
}
int main() {
char hex_in[] = "deadbeef10203040b00b1e50";
uint8_t out[32];
int res = to_byte_array(hex_in, sizeof(hex_in) - 1, out);
for (size_t i = 0; i < res; i++)
printf("%02x ", out[i]);
printf("\n");
system("pause");
return 0;
}
Solution 18 - C
The best way I know:
int hex2bin_by_zibri(char *source_str, char *dest_buffer)
{
char *line = source_str;
char *data = line;
int offset;
int read_byte;
int data_len = 0;
while (sscanf(data, " %02x%n", &read_byte, &offset) == 1) {
dest_buffer[data_len++] = read_byte;
data += offset;
}
return data_len;
}
The function returns the number of converted bytes saved in dest_buffer. The input string can contain spaces and mixed case letters.
"01 02 03 04 ab Cd eF garbage AB"
translates to dest_buffer containing 01 02 03 04 ab cd ef
and also "01020304abCdeFgarbageAB"
translates as before.
Parsing stops at the first "error".
Solution 19 - C
Here is a solution to deal with files, which may be used more frequently...
int convert(char *infile, char *outfile) {
char *source = NULL;
FILE *fp = fopen(infile, "r");
long bufsize;
if (fp != NULL) {
/* Go to the end of the file. */
if (fseek(fp, 0L, SEEK_END) == 0) {
/* Get the size of the file. */
bufsize = ftell(fp);
if (bufsize == -1) { /* Error */ }
/* Allocate our buffer to that size. */
source = malloc(sizeof(char) * (bufsize + 1));
/* Go back to the start of the file. */
if (fseek(fp, 0L, SEEK_SET) != 0) { /* Error */ }
/* Read the entire file into memory. */
size_t newLen = fread(source, sizeof(char), bufsize, fp);
if ( ferror( fp ) != 0 ) {
fputs("Error reading file", stderr);
} else {
source[newLen++] = '\0'; /* Just to be safe. */
}
}
fclose(fp);
}
int sourceLen = bufsize - 1;
int destLen = sourceLen/2;
unsigned char* dest = malloc(destLen);
short i;
unsigned char highByte, lowByte;
for (i = 0; i < sourceLen; i += 2)
{
highByte = toupper(source[i]);
lowByte = toupper(source[i + 1]);
if (highByte > 0x39)
highByte -= 0x37;
else
highByte -= 0x30;
if (lowByte > 0x39)
lowByte -= 0x37;
else
lowByte -= 0x30;
dest[i / 2] = (highByte << 4) | lowByte;
}
FILE *fop = fopen(outfile, "w");
if (fop == NULL) return 1;
fwrite(dest, 1, destLen, fop);
fclose(fop);
free(source);
free(dest);
return 0;
}
Solution 20 - C
No. But it's relatively trivial to achieve using sscanf in a loop.