djb2-pure

A single-file pure-Python implementation of djb2, the classic non-cryptographic 32-bit hash function designed by Daniel J. Bernstein (1991).

Pure stdlib. Zero dependencies. Single file.

What is djb2?

djb2 is one of the simplest and most widely used string-hashing functions. It appears in Berkeley DB, SDBM-derived libraries, the crc32-substitute discussions in comp.lang.c, and Bernstein's own historical notes. Its algebraic form is:

h = 5381
for c in bytes:
    h = (h * 33 + c) mod 2**32
return h

The constant 33 and seed 5381 are the canonical Bernstein choices — no derivation or "improved" variant. This module implements that exact algorithm verbatim.

Install

pip install -e .

Quick start

from djb2_pure import djb2, djb2_hex

djb2(b"")           # 5381          (the canonical seed)
djb2(b"a")          # 177670
djb2(b"abc")        # 193485963
djb2(b"hello world")  # 894552257
djb2_hex(b"hello world")  # "3551c8c1"

Public API

• djb2(data: bytes) -> int — returns the 32-bit hash as an unsigned integer
• djb2_hex(data: bytes) -> str — returns the 8-character lowercase hex string
• Djb2 — streaming class for incremental hashing of large/chunked inputs

Limitations

• NOT cryptographic. djb2 has known collisions and bias; do not use for security, integrity, or anywhere adversarial.
• NOT 64-bit. djb2 returns a 32-bit hash. For a 64-bit variant see djb2a (Bernstein's later proposal: seed 5381 with hash h = h * 33 ^ c, mod 2^64).
• Streaming class wraps the same algorithm. No novel variant; the class exists for parity with hashlib ergonomics.

License

CC0-1.0 — public domain dedication.