Dumpy

Dumpy is a binary parser with a declarative syntax. All you need to do is telling the framework what your data look like, and let Dumpy do the rest.

Dumpy supports python versions >= 3.0.

Installing

Just use pip or tools alike:

pip install dumpy

Usage

To parse a pre-defined binary structure, you need to define a class for that structure. For example, a class for pascal strings :

import dumpy.types as dt

class PString(dict, metaclass=dt.DumpyMeta):
    __field_specs__ = (
        dt.field('len', dt.UInt8, default=dt.count_of('data')),
        dt.field('data', dt.UInt8, count=dt.counted_by('len')),
    )

s = PString()
s['data'] = b'\x01\x02\x03\x04'

# The length field is calculated automatically.
assert s['len'] == 4

assert s.pack() == b'\x04\x01\x02\x03\x04'

s2 = s.unpack(s.pack())
assert s2['len'] == 4
assert bytes(s2['data']) == b'\x01\x02\x03\x04'

See demo/png_packer.py for a real-world format parser.

Known Issues

1. The type checking and conversion code in Dumpy is kind of naive, It does not check element types in a sequence.

2. Dumpy is very slow at current stage. You may not want to use it to parse network messages or huge data structures.