Deepmage

Introudction

Deepmage is a terminal-based hex editor for non-octet-oriented data. Its goal is to easily allow you to operate at the level of individual bits and n-bit words. This is the functionality I missed in most of the free hex editors, which were obviously made with 8-bit bytes in mind. Of course 8-bit bytes are good enough for most users, but people who do a lot of security research and/or reversing might know of a few frustrating situations in which the assumption that everything divides neatly into octets was proven false. For me, it was first with baseband (7-bit packed GSM alphabet) and with codecs (variable-length bitfields).

Status

Deepmage is currently in the alpha version. The code needs some refactoring, there's not real documentation and few comments and it works really slow if your terminal windows is big - but the basic functionality is there.

Deepmage allows you to view and edit files like in a normal hex editor, but you can also directly edit bits or change word (note that 'word' here means a basic unit, not machine word) size to something different than the usual 8 bits. It supports large file by only keeping in memory the parts of the file that have unsaved changes or are currently displayed to the user.

Dependencies

• asciimatics
• bitstring
• hy

How to use

You can download this repo

git clone https://github.com/mmiszczyk/deepmage

or get deepmage from PyPi

pip3 install deepmage

Then, to use the editor:

./deepmage [PATH_TO_FILE]

Use arrow keys, page up/page down and home/end to navigate.

Switch between hex and bit modes with F2.

Save changes with F5.

Exit with F10 or CTRL+C.

Usage for dumping/parsing scripts:

usage: deepmage-hexdump [-h] [-w wordsize] [-c cols] [-b] file

positional arguments:
  file                  Path to a file to edit

optional arguments:
  -h, --help            show this help message and exit
  -w wordsize, --wordsize wordsize
                        Size of words (in bits)
  -c cols, --cols cols
  -b, --bits

usage: deepmage-hexdump [-h] [-w wordsize] [--write_buffer_size bytes] [-b]
                        [-i]
                        [input-filename] output-filename

positional arguments:
  input-filename        Path to a hexdump file
  output-filename       Path to an output file

optional arguments:
  -h, --help            show this help message and exit
  -w wordsize, --wordsize wordsize
                        Size of words (in bits)
  --write_buffer_size bytes
  -b, --bits
  -i, --ignore-errors   Skip invalid characters

TODO

Planned

• fixing any bugs that will come out during testing

Maybe

• performance improvements
• UI code refactoring
• two-column display with approximate ASCII representation
• search and replace
• endianness toggle
• insert/delete bytes

Long-term (unlikely)

• parsing user-defined structure with words of different length, including variable-length bitfields