Convert output from MD5 to printable ascii

[Serman D.]

Hi,

I have very limited C knowledge. I want to convert to output from a
MD5 hash algorithm to printable ascii similar to the output of the
md5sum in GNU coreutils. Any help on how to do the conversion is
appreciated.

[aff@afflinux md5_xyssl]$ gcc -o test test.c md5.c
[aff@afflinux md5_xyssl]$ ./test
"J?n??CBW? ?}"
[aff@afflinux md5_xyssl]$ echo "This is my dearest secret: 12345" |
md5sum
749226c29c17114562745d9769fdab45 -
[aff@afflinux md5_xyssl]$

$ md5sum --version
md5sum (coreutils) 5.2.1

/*
* test.c
*
* MD5 source from http://xyssl.org/code/source/md5/
*
*/

#include "md5.h"
#include <string.h>

int main (void) {

unsigned char in[64] = "This is my dearest secret: 12345";
unsigned char *pout;

// compute MD5 hash
md5_csum(in, 64, pout);

printf("\"%s\"\n", pout);
}

[santosh]

Serman D. wrote:

Hi,

I have very limited C knowledge. I want to convert to output from a
MD5 hash algorithm to printable ascii similar to the output of the
md5sum in GNU coreutils. Any help on how to do the conversion is
appreciated.

Use printf with the %x or %X format specifier. Access the hash as an
array of unsigned char. Also the md5sum library may itself provide
formatting routines.

[Daniel Rudy]

At about the time of 2/23/2007 12:58 AM, Serman D. stated the following:

Hi,

I have very limited C knowledge. I want to convert to output from a
MD5 hash algorithm to printable ascii similar to the output of the
md5sum in GNU coreutils. Any help on how to do the conversion is
appreciated.

[aff@afflinux md5_xyssl]$ gcc -o test test.c md5.c
[aff@afflinux md5_xyssl]$ ./test
"J?n??CBW? ?}"
[aff@afflinux md5_xyssl]$ echo "This is my dearest secret: 12345" |
md5sum
749226c29c17114562745d9769fdab45 -
[aff@afflinux md5_xyssl]$

$ md5sum --version
md5sum (coreutils) 5.2.1

/*
* test.c
*
* MD5 source from http://xyssl.org/code/source/md5/
*
*/

#include "md5.h"
#include <string.h>

int main (void) {

unsigned char in[64] = "This is my dearest secret: 12345";
unsigned char *pout;

// compute MD5 hash
md5_csum(in, 64, pout);

printf("\"%s\"\n", pout);
}

int i;

for (i = 0; i < 16; i++) printf("%0x", pout[i]);
printf("\n");

I use 16 because MD6 hashes are 128 bits or 16 bytes long. This is
assuming an Intel platform.

--
Daniel Rudy

Email address has been base64 encoded to reduce spam
Decode email address using b64decode or uudecode -m

Why geeks like computers: look chat date touch grep make unzip
strip view finger mount fcsk more fcsk yes spray umount sleep

[Serman D.]

On Feb 23, 10:49 am, Daniel Rudy <[e-mail]:

At about the time of 2/23/2007 12:58 AM, Serman D. stated the following:

int main (void) {

unsigned char in[64] = "This is my dearest secret: 12345";
unsigned char *pout;

// compute MD5 hash
md5_csum(in, 64, pout);

printf("\"%s\"\n", pout);
}

int i;
for (i = 0; i < 16; i++) printf("%0x", pout[i]);
printf("\n");

Thanks Daniel,

Follow-up question:

How do I copy the contents to a string (unsigned char pointer)
variable instead of printing to stdout?

I'm completely lost in the C jungle of string manipulation..

--
Serman D.

[anyone.anon]

On Feb 23, 11:44 am, "Serman D." <[e-mail]:

How do I copy the contents to a string (unsigned char pointer)
variable instead of printing to stdout?

sprintf()

[Daniel Rudy]

At about the time of 2/23/2007 3:44 AM, Serman D. stated the following:

On Feb 23, 10:49 am, Daniel Rudy <[e-mail]:

>At about the time of 2/23/2007 12:58 AM, Serman D. stated the following:

>>int main (void) {
unsigned char in[64] = "This is my dearest secret: 12345";
unsigned char *pout;
// compute MD5 hash
md5_csum(in, 64, pout);
printf("\"%s\"\n", pout);
}

int i;
for (i = 0; i < 16; i++) printf("%0x", pout[i]);
printf("\n");

Thanks Daniel,

Follow-up question:

How do I copy the contents to a string (unsigned char pointer)
variable instead of printing to stdout?

I'm completely lost in the C jungle of string manipulation..

--
Serman D.

The BIG thing you need to realize here is that strings in C are null
terminated (ASCII 0). The output of MD5 is BINARY, so for an unsigned
char, all values 0-255 are valid data...this includes the 0 or null char
for string termination. There's a set of functions that is used to
handle binary strings. To copy binary strings, take a look at memcpy or
memmove.

Something like:

unsigned char dst[16];
memcpy(dst, pout, 16);

Since the data is binary, there is no terminating null. You just have
to be careful when working with it and use the size correctly. The best
way to do this would be like this:

#include <stdio.h>
#include <string.h>
#include "md5.h"

#define MD5HASHSIZE 16

int main (void) {

unsigned char in[64] = "This is my dearest secret: 12345";
unsigned char *pout;
unsigned char dest[MD5HASHSIZE];
int i;

// compute MD5 hash
md5_csum(in, 64, pout);

printf("\"%s\"\n", pout);

for (i = 0; i < MD5HASHSIZE; i++) printf("%0x", pout[i]);
printf("\n");

memmove(dest, pout, MD5HASHSIZE);

return(0);
}

Note the #define directive. The pre-processor will replace all
instances of MD5HASHSIZE with 16. So if you change the value of
MD5HASHSIZE, it will change everywhere so you don't have to go through
your code and change each instance manually. Think of it as a constant.
It's not a const, but it's like a const.

I'm sure that someone will correct me on this....
--
Daniel Rudy

Email address has been base64 encoded to reduce spam
Decode email address using b64decode or uudecode -m

Why geeks like computers: look chat date touch grep make unzip
strip view finger mount fcsk more fcsk yes spray umount sleep

[Daniel Rudy]

At about the time of 2/23/2007 3:44 AM, Serman D. stated the following:

On Feb 23, 10:49 am, Daniel Rudy <[e-mail]:

>At about the time of 2/23/2007 12:58 AM, Serman D. stated the following:

>>int main (void) {
unsigned char in[64] = "This is my dearest secret: 12345";
unsigned char *pout;
// compute MD5 hash
md5_csum(in, 64, pout);
printf("\"%s\"\n", pout);
}

int i;
for (i = 0; i < 16; i++) printf("%0x", pout[i]);
printf("\n");

Thanks Daniel,

Follow-up question:

How do I copy the contents to a string (unsigned char pointer)
variable instead of printing to stdout?

I'm completely lost in the C jungle of string manipulation..

--
Serman D.

Disreguard my other post. I misunderstood what you were refering too.

int i;
char c[10];
char md5ascii[128];

memset(md5ascii, 0, sizeof(md5ascii);
for (i = 0; i < 16; i++)
{
snprintf(c, sizeof(c), "%0x", pout[i];
strncat(md5ascii, c, sizeof(md5ascii) - 1;
}

--
Daniel Rudy

Email address has been base64 encoded to reduce spam
Decode email address using b64decode or uudecode -m

Why geeks like computers: look chat date touch grep make unzip
strip view finger mount fcsk more fcsk yes spray umount sleep

[santosh]

Daniel Rudy wrote:

At about the time of 2/23/2007 3:44 AM, Serman D. stated the following:

<snip>

Follow-up question:

How do I copy the contents to a string (unsigned char pointer)
variable instead of printing to stdout?

[ ... ]

Disreguard my other post. I misunderstood what you were refering too.

int i;
char c[10];
char md5ascii[128];

memset(md5ascii, 0, sizeof(md5ascii);
for (i = 0; i < 16; i++)
{
snprintf(c, sizeof(c), "%0x", pout[i];
strncat(md5ascii, c, sizeof(md5ascii) - 1;
}

Hacked out on the spot eh? You've forgotten to close the argument list
for memset, snprintf and strncat. Required headers, definition of pout
and main are missing. Here's a compilable example based on yours:

#include <stdio.h>
#include <string.h>

int i;
char c[10];
char md5ascii[128];
unsigned char pout[16];

int main(void) {
for(i = 0; i < sizeof(pout); i++) pout[i] = i + 1;

memset(md5ascii, 0, sizeof(md5ascii));
for (i = 0; i < 16; i++) {
snprintf(c, sizeof(c), "%0x", pout[i]);
strncat(md5ascii, c, sizeof(md5ascii) - 1);
}

for(i = 0; i < sizeof(md5ascii); i++) putchar(md5ascii[i]);

putchar('\n');
return 0;
}

[Daniel Rudy]

At about the time of 2/23/2007 10:05 AM, santosh stated the following:

Daniel Rudy wrote:

>At about the time of 2/23/2007 3:44 AM, Serman D. stated the following:

<snip>

>>Follow-up question:

How do I copy the contents to a string (unsigned char pointer)
variable instead of printing to stdout?

[ ... ]

>Disreguard my other post. I misunderstood what you were refering too.

int i;
char c[10];
char md5ascii[128];

memset(md5ascii, 0, sizeof(md5ascii);
for (i = 0; i < 16; i++)
{
snprintf(c, sizeof(c), "%0x", pout[i];
strncat(md5ascii, c, sizeof(md5ascii) - 1;
}

Hacked out on the spot eh?

Pretty much. :) That's what happens when you are up since 1 in the AM.

You've forgotten to close the argument list
for memset, snprintf and strncat. Required headers, definition of pout
and main are missing. Here's a compilable example based on yours:

#include <stdio.h>
#include <string.h>

int i;
char c[10];
char md5ascii[128];
unsigned char pout[16];

int main(void) {
for(i = 0; i < sizeof(pout); i++) pout[i] = i + 1;

memset(md5ascii, 0, sizeof(md5ascii));
for (i = 0; i < 16; i++) {
snprintf(c, sizeof(c), "%0x", pout[i]);
strncat(md5ascii, c, sizeof(md5ascii) - 1);
}

for(i = 0; i < sizeof(md5ascii); i++) putchar(md5ascii[i]);

putchar('\n');
return 0;
}

Opps.

Looks like my secret is out. I guess you can tell that I rely on
compiler warnings and errors *way* too much.

--
Daniel Rudy

Email address has been base64 encoded to reduce spam
Decode email address using b64decode or uudecode -m

Why geeks like computers: look chat date touch grep make unzip
strip view finger mount fcsk more fcsk yes spray umount sleep

[websnarf]

On Feb 23, 1:49 am, Daniel Rudy <[e-mail]:

At about the time of 2/23/2007 12:58 AM, Serman D. stated the
following:

I have very limited C knowledge. I want to convert to output from a
MD5 hash algorithm to printable ascii similar to the output of the
md5sum in GNU coreutils. Any help on how to do the conversion is
appreciated.

[aff@afflinux md5_xyssl]$ gcc -o test test.c md5.c
[aff@afflinux md5_xyssl]$ ./test
"J?n??CBW? ?}"
[aff@afflinux md5_xyssl]$ echo "This is my dearest secret: 12345" |
md5sum
749226c29c17114562745d9769fdab45 -
[aff@afflinux md5_xyssl]$

$ md5sum --version
md5sum (coreutils) 5.2.1

/*
* test.c
*
* MD5 source fromhttp://xyssl.org/code/source/md5/
*
*/

#include "md5.h"
#include <string.h>

int main (void) {

unsigned char in[64] = "This is my dearest secret: 12345";
unsigned char *pout;

// compute MD5 hash
md5_csum(in, 64, pout);

printf("\"%s\"\n", pout);
}

int i;

for (i = 0; i < 16; i++) printf("%0x", pout[i]);
printf("\n");

Let's try:

for (i=0; i < 16; i++) printf ("%02x", pout[i]);
printf ("\n");

The difference kind of matters.

However, I would, of course, just do this:

char out[33], *p;
for (p=out,i=0; i < 16; i++) {
*p++ = "0123456789ABCDEF"[(pout[i] >4) & 0xf];
*p++ = "0123456789ABCDEF"[pout[i] & 0xf];
}
*p = '\0';
puts (out);

to avoid linking in all of printf() unnecessarily.

--
Paul Hsieh
http://www.pobox.com/~qed/
http://bstring.sf.net/

[Keith Thompson]

we******@gmail.com writes:
[...]

However, I would, of course, just do this:

char out[33], *p;
for (p=out,i=0; i < 16; i++) {
*p++ = "0123456789ABCDEF"[(pout[i] >4) & 0xf];
*p++ = "0123456789ABCDEF"[pout[i] & 0xf];
}
*p = '\0';
puts (out);

to avoid linking in all of printf() unnecessarily.

I presume you would first confirm that (a) the use of puts() doesn't
cause printf() to be linked in anyway on target implementations of
interest, and (b) the measured saving in final code size is actually
worth the additional effort (including testing the code to make sure
you got it right).

--
Keith Thompson (The_Other_Keith) ks***@mib.org <http://www.ghoti.net/~kst>
San Diego Supercomputer Center <* <http://users.sdsc.edu/~kst>
We must do something. This is something. Therefore, we must do this.

[Kenny McCormack]

In article <ln************@nuthaus.mib.org>,
Keith Thompson <ks***@mib.orgwrote:

>we******@gmail.com writes:
[...]

>However, I would, of course, just do this:

char out[33], *p;
for (p=out,i=0; i < 16; i++) {
*p++ = "0123456789ABCDEF"[(pout[i] >4) & 0xf];
*p++ = "0123456789ABCDEF"[pout[i] & 0xf];
}
*p = '\0';
puts (out);

to avoid linking in all of printf() unnecessarily.

I presume you would first confirm that (a) the use of puts() doesn't
cause printf() to be linked in anyway on target implementations of
interest, and (b) the measured saving in final code size is actually
worth the additional effort (including testing the code to make sure
you got it right).

Keithie, baby. Seriously. You need to get thee to www.senseofhumor.com
and buy one online. Whatever they're charging, it'll be worth it.
Trust me on this.

[websnarf]

On Feb 24, 1:42 pm, Keith Thompson <[e-mail]:

[e-mail] writes:

However, I would, of course, just do this:
char out[33], *p;
for (p=out,i=0; i < 16; i++) {
*p++ = "0123456789ABCDEF"[(pout[i] >4) & 0xf];
*p++ = "0123456789ABCDEF"[pout[i] & 0xf];
}
*p = '\0';
puts (out);

to avoid linking in all of printf() unnecessarily.

I presume you would first confirm that (a) the use of puts() doesn't
cause printf() to be linked in anyway on target implementations of
interest,

Well, I don't know what quality of implementations you are used to.
All the compilers I use basically do link these seperately. The
reason is that printf is a gargantuan link-in in comparison to puts.

[...] and (b) the measured saving in final code size is actually
worth the additional effort

printf needs to link in double -ASCII conversion. This cannot be
avoided unless the compiler really translates the format string at
compile time (which only gcc seems to bother trying) which itself is
risky since someone might want to override the implementation of
printf (which may not be standard endorsed, but it doesn't change the
fact that people want to do that.)

[...] (including testing the code to make sure you got it right).

Ironic that you suggest testing (is that mentioned somewhere in the
standard?) Daniel Rudy's code would have passed *some* testing while
still being wrong. Furthermore if you are unable to figure it out
logically, then *where* it is failing is on the inside printf's object
code, which in many compiler's debuggers you can't trace through. If
there's anything wrong in the code I showed, it will be pretty obvious
pretty quickly.

--
Paul Hsieh
http://www.pobox.com/~qed/
http://bstring.sf.net/

[santosh]

Keith Thompson wrote:

we******@gmail.com writes:
[...]

However, I would, of course, just do this:

char out[33], *p;
for (p=out,i=0; i < 16; i++) {
*p++ = "0123456789ABCDEF"[(pout[i] >4) & 0xf];
*p++ = "0123456789ABCDEF"[pout[i] & 0xf];
}
*p = '\0';
puts (out);

to avoid linking in all of printf() unnecessarily.

I presume you would first confirm that (a) the use of puts() doesn't
cause printf() to be linked in anyway on target implementations of
interest, and (b) the measured saving in final code size is actually
worth the additional effort (including testing the code to make sure
you got it right).

Interestingly, over here, (gcc 4.0.3 on an i686 Linux), gcc will emit
a call to puts for the corresponding printf invocation when the first
argument contains no format specifiers. This is done even when
optimisations are disabled.

I was able to turn of this "feature" by using a char pointer
initialised to string literal instead of the latter. In this case the
function mentioned in the source is called, whether it be printf or
puts.

Also linking in puts in the place of printf did nothing to reduce the
object code size. This probably means that printf and puts are in the
same object module and hence linked in together. I haven't actually
checked that.

[Ben Pfaff]

"santosh" <sa*********@gmail.comwrites:

Interestingly, over here, (gcc 4.0.3 on an i686 Linux), gcc will emit
a call to puts for the corresponding printf invocation when the first
argument contains no format specifiers. This is done even when
optimisations are disabled.

I hope it does that only when the format string ends in a
new-line character.
--
"Your correction is 100% correct and 0% helpful. Well done!"
--Richard Heathfield

[santosh]

Ben Pfaff wrote:

"santosh" <sa*********@gmail.comwrites:

Interestingly, over here, (gcc 4.0.3 on an i686 Linux), gcc will emit
a call to puts for the corresponding printf invocation when the first
argument contains no format specifiers. This is done even when
optimisations are disabled.

I hope it does that only when the format string ends in a
new-line character.

It seems to do as you say. Why that requirement?

[Joachim Schmitz]

"santosh" <sa*********@gmail.comschrieb im Newsbeitrag
news:11**********************@m58g2000cwm.googlegr oups.com...

>
Ben Pfaff wrote:

>"santosh" <sa*********@gmail.comwrites:

Interestingly, over here, (gcc 4.0.3 on an i686 Linux), gcc will emit
a call to puts for the corresponding printf invocation when the first
argument contains no format specifiers. This is done even when
optimisations are disabled.

I hope it does that only when the format string ends in a
new-line character.

It seems to do as you say. Why that requirement?

Because puts() appends a newline to the string it prints. It printf() hadn't
been told to add a newline, and would use puts() to do the printing, that
would add a newline, rendering the printf() behavoir non-conforming.

Bye, Jojo

[Flash Gordon]

santosh wrote, On 25/02/07 08:56:

Ben Pfaff wrote:

>"santosh" <sa*********@gmail.comwrites:

>>Interestingly, over here, (gcc 4.0.3 on an i686 Linux), gcc will emit
a call to puts for the corresponding printf invocation when the first
argument contains no format specifiers. This is done even when
optimisations are disabled.

I hope it does that only when the format string ends in a
new-line character.

It seems to do as you say. Why that requirement?

Because puts appends a newline character, so it can't be used when you
want to print a partial line not ending in a newline (possibly because
some other part of the program will print the rest of the line).
--
Flash Gordon

[santosh]

Flash Gordon wrote:

santosh wrote, On 25/02/07 08:56:

Ben Pfaff wrote:

"santosh" <sa*********@gmail.comwrites:

Interestingly, over here, (gcc 4.0.3 on an i686 Linux), gcc will emit
a call to puts for the corresponding printf invocation when the first
argument contains no format specifiers. This is done even when
optimisations are disabled.

I hope it does that only when the format string ends in a
new-line character.

It seems to do as you say. Why that requirement?

Because puts appends a newline character, so it can't be used when you
want to print a partial line not ending in a newline (possibly because
some other part of the program will print the rest of the line).

Thanks. I see it now.

[Randy Howard]

On Sat, 24 Feb 2007 22:13:21 -0600, we******@gmail.com wrote
(in article <11**********************@j27g2000cwj.googlegroups .com>):

On Feb 24, 1:42 pm, Keith Thompson <[e-mail]:

>[e-mail] writes:

>>However, I would, of course, just do this:
char out[33], *p;
for (p=out,i=0; i < 16; i++) {
*p++ = "0123456789ABCDEF"[(pout[i] >4) & 0xf];
*p++ = "0123456789ABCDEF"[pout[i] & 0xf];
}
*p = '\0';
puts (out);

>>to avoid linking in all of printf() unnecessarily.

I presume you would first confirm that (a) the use of puts() doesn't
cause printf() to be linked in anyway on target implementations of
interest,

Well, I don't know what quality of implementations you are used to.
All the compilers I use basically do link these seperately. The
reason is that printf is a gargantuan link-in in comparison to puts.

I just tried this with gcc 4.0.1 using gcc on a Hello, World program
with printf and puts on OS X. The output size is exactly the same.
So, I used a slightly trickier printf() call that shouldn't be switched
out to a straight puts(), and that output file went up by 4284 bytes.

Outside of embedded land, that is completely meaningless.

--
Randy Howard (2reply remove FOOBAR)
"The power of accurate observation is called cynicism by those
who have not got it." - George Bernard Shaw