reverse hex dump

Install zig.

$ mkdir build && cd build
$ zig build-exe --library c ../src/hex2bin.zig

Executable binary is at build/hex2bin.

hex2bin INPUT.hex OUTPUT.bin

The input file must use unix line endings, and the last line must be blank.

Here is an example of a complete file:

; comment
:0x00 ; offset directive
68 65 6c 6c 6f 20 77 6f 72 6c 64 0a ; the text "hello world\n"
:0x0c ; offset directive

Comments start with ; and end at the end of the line.

; this is a comment

A byte is encoded as a sequence of two nibbles, where a nibble is 0-9, A-F, a-f.

05 0a 0f 14 ; four bytes with decimal values 5, 10, 15, 20

The nibbles that encode a byte must be immediately adjacent to each other.

0 0 ; error

Nibbles from different bytes must not be immediately adjacent to each other. This is to make sure there's no confusion about big/little endian byte order.

0000 ; error

The output of the program simply the sequence of bytes encoded by these nibble sequences.

An offset directive is optional and simply asserts the byte offset of the output at the current position in the input file. An offset directive starts with :0x, followed by 1 or more nibbles, and terminated by the first non-nibble character.

:0x00 ; offset starts at 0
68 65 6c 6c 6f 20 77 6f 72 6c 64 0a ; the text "hello world\n"
:0x0c ; we've encountered 12 bytes so far
:0x0d ; this is an error; we're still at offset 12

TODO: add support for cool stuff like utf8"foo" and u32le decimal literals.

This project produces hexdumps in the format consumed by this project:

• hexdump-zip