UNIX for Dummies Questions & Answers

Hi,

Im wondering how I can convert a binary file to a text file?
I have ran the following command to output which type of binary file coding it is

Code:

od -t x1 -c eHat.data0 | head -20

and that gives me the following output;

Code:

0000000 5c 00 00 00 cd 06 f2 41 00 00 00 c6 00 00 00 c6 \ \0 \0 \0 315 006 362 A \0 \0 \0 306 \0 \0 \0 306
0000020 42 8f 95 c0 19 d5 5a c2 8e 99 1f c1 91 03 94 41 B 217 225 300 031 325 Z 302 216 231 037 301 221 003 224 A
0000040 f7 7e cd 40 c4 cc d2 be 00 00 00 c6 00 00 00 c6 367 ~ 315 @ 304 314 322 276 \0 \0 \0 306 \0 \0 \0 306
0000060 00 00 00 c6 6c d9 91 3e b6 9c 2c 3e 5b 98 63 3e \0 \0 \0 306 l 331 221 > 266 234 , > [ 230 c >
0000100 84 61 c1 3e 29 ec 6d 3e 10 9e 4d 44 00 00 00 c6 204 a 301 > ) 354 m > 020 236 M D \0 \0 \0 306
0000120 00 00 00 c6 e2 c3 79 44 c2 50 b6 43 bd ab 9a 40 \0 \0 \0 306 342 303 y D 302 P 266 C 275 253 232 @
0000140 5c 00 00 00 5c 00 00 00 00 00 00 c6 00 00 00 c6 \ \0 \0 \0 \ \0 \0 \0 \0 \0 \0 306 \0 \0 \0 306
0000160 00 00 00 c6 00 00 00 c6 00 00 00 c6 00 00 00 c6 \0 \0 \0 306 \0 \0 \0 306 \0 \0 \0 306 \0 \0 \0 306
*
0000300 00 00 00 c6 5c 00 00 00 5c 00 00 00 00 00 00 c6 \0 \0 \0 306 \ \0 \0 \0 \ \0 \0 \0 \0 \0 \0 306
0000320 00 00 00 c6 00 00 00 c6 83 08 a8 c2 82 d2 25 c3

Is there a way I can convert this to a text file?
Many thanks!

Moderator's Comments:

edit by bakunin: please use CODE-tags not only for code but also for terminal output and file contents. Thank you.

I'm afraid above is NOT the binary representation of a well-formed text file (many zero bytes, mal-formed multibyte chars, malformed single byte chars, ...).
While

Code:

awk '{for(i=2; i<=NF; i++) printf "%c", 0+sprintf ("%d", "0x"$i)}' file
Col1 Col2 Col3 Col4
ABC 00000001 15-Dec-15 13000
ABC 00000001 31-Jan-16 13500
.
.
.

would reconstruct an ASCII file, the result of it applied to every other line of your file yields

Code:

\��A��B��ZŽ����A�~�@��Ҿ���lّ>��,>[�c>�a�>)�m>�MD����yD�P�C���@\\������c

A few chars are recognizable ("\", "A", "B", etc), but the main part is garbage.

What a "binary file" and what a "text file" is is primarily determined by the files contents. So there is no "conversion" between the two, like "convert CSV to Excel format". The latter is just two representations of the same thing - tabular data - but the former are simply different types of contents.

There are (today rather unusual) meanings of "text file" as containing compiled code (calling a relocatible object file before linking "text" was the IBM speak of the seventies) and if you are using the term this way you will have to use a linker. If you are using "text" in the more common meaning of "sequences of characters transporting meaning to a reading human" then there is no (common!) way to transform one into the other.

There are of course programs which use binary representations for text files (simpliest example is a packing program which packs a text file into something binary). In such a case you need to specifically reverse the algorithm (i.e use an unpack-program to restore the text file), but this means to know the used algorithm first. And even this only works if a text file was packed beforehand. Put a binary file in and another binary file will come out, packing or no packing.

If you want to search for text parts inside a binary file, this is possible: use the strings command to list readable strings within otherwise binary files:

Code:

strings /path/to/file

I hope this helps.

bakunin

This looks like packed decimal to me.
The fields that look like '00 00 00 c6' appear to be cobol S9(7) COMP 3.
And on your machine, the hd program reverses the digits in each byte so that it is probably
0000006c on disk.