Print a file in binary form (1 and 0)

Question

In the context of a GZip decoder I need a unix tool or a C solution to print my compressed gzip file on my screen in a binary form. What I exactly need is hexdump able to print in binary instead of octal,decimal or hexadecimal.

I am pretty sure this tool exists but I can't find it :-(

EDIT I guess such editor is hard to find because the use case must not be frequent. In my case I have very small files and need for debug purpose to search for binary "patterns" such as 10010011 that may cross several bytes

if you search for multi-byte-patterns you probably want a solution that doesn't impose a line wrap to the stream? — moooeeeep, Jul 20 '12 at 13:08
See this: http://stackoverflow.com/questions/111928/is-there-a-printf-converter-to-print-in-binary-format — qwertz, Jul 20 '12 at 13:13

Nahuel Fouilleul · Accepted Answer · 2012-07-20T14:45:22.690

2

perl -ne 'print unpack(B8,$_),$/for split//' FILE

or to have 8 elements in a line

perl -ne 'print unpack(B8,$_),++$i%8?" ":"\n"for split//;END{print"\n"}'

without linebreaks

perl -ne 'print unpack(B8,$_)for split//' FILE

edited Jul 20 '12 at 14:45

answered Jul 20 '12 at 14:08

Nahuel Fouilleul

18,726
2
31
36

Perfect ! and without line breaks, how should I modify this ? – Manuel Selva Jul 20 '12 at 14:25

alinsoar · Answer 2 · 2012-08-04T14:48:30.863

1

Because Manuel told me to repost, I completed the commentary that I wrote last time, and here is the call:

od -t x1 FILE | sed 'h;x;s:^\(........\).*:\1:;x;s:^........::;s:0:0000:g;s:1:0001:g;s:2:0010:g;s:3:0011:g;s:4:0100:g;s:5:0101:g;s:6:0110:g;s:7:0111:g;s:8:1000:g;s:9:1001:g;s:a:1010:g;s:b:1011:g;s:c:1100:g;s:d:1101:g;s:e:1110:g;s:f:1111:g;H;x;s:\n::'

edited Aug 04 '12 at 14:48

answered Jul 20 '12 at 15:50

alinsoar

15,386
4
57
74

Thanks and do not blame me for my ignorance about sed ;-) – Manuel Selva Jul 20 '12 at 17:15

Some programmer dude · Answer 3 · 2012-07-20T13:26:09.007

0

From another answer I made (that I unfortunately can't find right now) I have this function to print an int as binary digits:

void print_bin(int n)
{
    for (int i = sizeof(n) * 8 - 1; i >= 0; i--)
        printf("%d", (int) ((n >> i) & 1));
}

You can read the file as integers and use this function to print the values as binary digits.

edited Jul 20 '12 at 13:26

answered Jul 20 '12 at 13:02

Some programmer dude

400,186
35
402
621

thanks for the tips. If such an heditor doesn't exist I'll write a little tool starting from this. – Manuel Selva Jul 20 '12 at 13:03
what does int(((n >> i) & 1)) mean ? it doesn't compile for me ? – Manuel Selva Jul 20 '12 at 13:23
@ManuelSelva Ah, I used this in a C++ program originally, edited to C-style casting now. – Some programmer dude Jul 20 '12 at 13:26
Thanks I get it ;-) but why I didn't want to do that is that your solution doesn't print the same result that the other C hand coded solution proposed in an other answer ;-) – Manuel Selva Jul 20 '12 at 14:10
@ManuelSelva What do you mean "doesn't print the same result"? The function is easily changed to print single bytes (just change the parameter type to `char`) if that's what you need? – Some programmer dude Jul 20 '12 at 14:13
yes but in a main program that read the file 4 bytes at a time – Manuel Selva Jul 20 '12 at 14:23
is the result of a right-shift for signed integers well-defined? – moooeeeep Jul 20 '12 at 15:05
@ManuelSelva `int` should always be 32 bits (i.e. 4 bytes) even on 64-bit machines. `long` can be different depending on 32- or 64-bit architectures, and other things. – Some programmer dude Jul 23 '12 at 06:37

score 0 · Answer 4 · answered Jul 20 '12 at 13:25

0

your compiler could support this utility function ltoa

char *ltoa( long value, char * buffer, int radix );

that could simplify the display with radix = 2

answered Jul 20 '12 at 13:25

CapelliC

59,646
5
47
90

score 0 · Answer 5 · answered Jul 20 '12 at 13:30

0

od -t x1 FILE | sed 's:0:0000:g;s:1:0001:g;\
s:2:0010:g;s:3:0011;
...[replace every hex-digit up to F with the bin representation] '

answered Jul 20 '12 at 13:30

alinsoar

15,386
4
57
74

the syntax of the command is not goo. can you fix it please, this is exactly what I need – Manuel Selva Jul 20 '12 at 13:35

score 0 · Answer 6 · answered Jul 20 '12 at 13:53

Just write a little program:

#include <stdio.h>

void binarywrite (unsigned char c)
{
    int i = 0;
    for (i = 0; i < 8; i++)
    {
        printf("%1d",(c>>(7-i)) & 1 );
    }
    printf(" ");
}

int main(int argc, char* argv[])
{
    char c = 0;

    if(argc >=2)
    {
        char* pfilename = argv[1];
        FILE *fp;
        fp = fopen(pfilename,"rb");
        char ch;
        while((ch = getc(fp)) != EOF)
        {
            binarywrite(ch);
        }
        fclose(fp);
    }

    return 0;
}

And you can use it as './binarywite filename'

Thanks for the answer. but why I didn't want to do that is that your solution doesn't print the same result that the other C hand coded solution proposed in an other answer ;-) — Manuel Selva, Jul 20 '12 at 14:10

score 0 · Answer 7 · answered Aug 05 '12 at 09:31

xxd (which usually comes with vim) can convert to binary representation directly:

xxd -b FILE

To cut off the formatting of the output, a bit of awk can be appended so you get a clean stream of zeros and ones only:

xxd -b FILE | \
    awk '{ for (i=0; i<NF; i=i+1)
           { if ($i ~ /^[01][01][01][01][01][01][01][01]$/) printf $i }
         }; END { print "" }
    '

Print a file in binary form (1 and 0)

7 Answers7