Atomicity

When working with the atomic interface provided by gyver.context, you can take advantage of the atomic utility function. This function simplifies the creation of atomic contexts in both synchronous and asynchronous scenarios.

Using the atomic Utility Function

Asyncio

Note that the atomic function also works for the AtomicAsyncAdapter

The atomic utility function allows you to create atomic contexts with ease. It wraps a given context with the atomic interface, ensuring proper resource management and atomic operations.

Example:

from gyver.context import atomic

# Create an atomic context using the atomic utility function
atomic_context = atomic(my_context)

# Inside the atomic context, the atomicity is managed automatically
with atomic_context as resource:
    # Perform operations within the atomic context
    do_something()

# Upon exiting the atomic context, the adapter's `rollback` or `commit` methods are invoked
# if any errors occurred. This happens only at the outermost frame.

The atomic function also supports an optional parameter bound, which determines whether the resulting context will be the owner of the resource or will attach to the current context. By default, bound is set to True, allowing reuse of already active resources from the provided context. It handles all context management from that context. When set to False, the new context manages resources independently, opening and closing them based on its own stack count.

Example:

# Assume there's an active resource from the outer context
client = None

# Wrapping an existing context with the atomic interface
with context as client:
    pass

# Creating an atomic context that reuses the existing resource (bound=True)
with atomic(context) as atomic_client:
    print(atomic_client is client)  # True

# Creating an independent atomic context with its own resource (bound=False)
with atomic(context, bound=False) as unbound_client:
    print(unbound_client is client)  # False