caesura 📜

Mid-execution distributed locking with zero boilerplate.

Standard distributed locks force you to wrap your entire function in a with block, even if you need heavy parsing or database lookups just to find out what needs locking. caesura lets you safely acquire locks mid-execution: yield signals exactly where the blocking I/O happens, and the decorator manages teardown—even on crash.

uv add caesura

Why

Usage

import caesura

@caesura.serialize
def process_webhook(payload):
    parsed_data = heavy_json_parse(payload)  # lock-free work first

    yield redis.lock(f"user:{parsed_data.user_id}")  # hand the lock to the decorator

    db.update_balance(parsed_data.user_id)  # resumes safely locked
    return True  # locks released on return — or on crash — via Python's ExitStack

Deadlock Prevention

Yield multiple resources as a tuple. caesura sorts them by their name (falling back to str()) and acquires them in an identical, deterministic order across all workers — making deadlocks mathematically impossible.

@caesura.serialize
def transfer_funds(sender_id, receiver_id):
    yield (redis.lock(sender_id), redis.lock(receiver_id))  # sorted before acquiring
    db.execute_transfer(sender_id, receiver_id)