flexdi

Flex DI

.. image:: https://img.shields.io/pypi/v/flexdi.svg :target: https://pypi.org/project/flexdi/

.. image:: https://img.shields.io/endpoint.svg?url=https%3A%2F%2Factions-badge.atrox.dev%2Fcal-pratt%2Fflexdi%2Fbadge%3Fref%3Dmain&style=flat :target: https://github.com/cal-pratt/flexdi/actions

.. image:: https://readthedocs.org/projects/flexdi/badge/?version=latest :target: https://flexdi.readthedocs.io

flexdi is a yet another Dependency Injection library for Python. flexdi provides a lightweight alternative to common DI solutions with minimal setup to be included in your projects. This library is intended for use with type annotated Python libraries, as it leverages these type annotations to perform injection.

.. note:: | Flex DI is still a work in progress and is not yet intended for production use-cases. | Be aware that APIs are likely to change in upcoming releases.

| | The full documentation is available at flexdi.readthedocs.io <https://flexdi.readthedocs.io>_.

Goals

• | Minimal Setup | Minimize boilderplate by leveraging type annotations to resolve arguments. Allows user code to remain generic and reusable.

• | Inject Any Callable | Provide any typed callable as an input to be invoked. Supports sync and async callables.

• | Resource Management | Define dependencies as context managers and have their startup and shutdown logic invoked in a reliable order. Supports sync and async context managers.

• | Scoped Dependencies | Clearly define dependency lifetimes. Use "application" scoped dependencies to provide singleton like objects. Use "request" scoped dependencies to allow short-term isolated usage.

Overview

flexdi offers the FlexGraph, used to manage dependencies and invoke callables. The graph is a representation of what dependencies exist, and what providers can be used to fulfil them.

When determining dependencies for a callable, flexdi will examine the type annotations of the arguments to populate the graph with dependencies that can satisfy the callable. A callable can be anything from a class, to functions, class methods, generators, etc.

flexdi allows binding helper functions to the graph as a providers of types, determined by their return annotations. Bindings can themselves be injected with dependencies, determined by their parameter annotations. A Binding can be a simple function or a generator with custom teardown logic.

Example Usage

A simple example of an application with SQLAlchemy dependencies:

.. code:: python

from typing import Iterator

from sqlalchemy import Engine, create_engine, text
from sqlalchemy.orm import Session

from flexdi import FlexGraph

# The FlexGraph keeps track of what dependencies different
# providers require, and will later be used to resolve them.
graph = FlexGraph()


# Let's add a binding for an Engine.
# The binding will be used for anything that requires an Engine.
# FlexGraph uses the return type annotation to create bindings.
@graph.bind
def provide_engine() -> Engine:
    return create_engine("sqlite:///mydb.db")


# Generator responses can also be used. e.g.
# - A function returning Iterator[T] binds to T
# - A function returning Generator[T, U, V] binds to T
# - A function returning AsyncIterator[T] binds to T
# - A function returning AsyncGenerator[T, U] binds to T
@graph.bind
def provide_connection(engine: Engine) -> Iterator[Session]:
    with Session(engine) as session:
        yield session


# An entrypoint is a convenience method for a creating no argument
# version of a function or coroutine. You should typically only
# have one entrypoint used in your applications.
@graph.entrypoint
def main(conn: Session) -> None:
    statement = text("SELECT name FROM sqlite_master;")
    for [table_name] in conn.execute(statement):
        print(table_name)


# Notice that we call main with no arguments!
if __name__ == "__main__":
    main()

The same example, but using async code:

.. code:: python

from typing import AsyncIterator

from sqlalchemy import text
from sqlalchemy.ext.asyncio import AsyncConnection, AsyncEngine, create_async_engine

from flexdi import FlexGraph

graph = FlexGraph()


@graph.bind
async def provide_engine() -> AsyncIterator[AsyncEngine]:
    engine = create_async_engine("sqlite+aiosqlite:///mydb.db")
    try:
        yield engine
    finally:
        await engine.dispose()


@graph.bind
async def provide_connection(engine: AsyncEngine) -> AsyncIterator[AsyncConnection]:
    async with engine.begin() as conn:
        yield conn


@graph.entrypoint
async def main(conn: AsyncConnection) -> None:
    statement = text("SELECT name FROM sqlite_master;")
    for [table_name] in await conn.execute(statement):
        print(table_name)


if __name__ == "__main__":
    main()

Alternatives

Although there are many, many other dependency injection libraries, I found that I was still left looking for more lightweight/minimal solutions to this problem. My thoughts on some of the popular alternatives I have used in the past:

• | dependency-injector <https://github.com/ets-labs/python-dependency-injector>_ | This library is probably the most mature out of all the alternatives. Its main driving principal is that "Explicit is better than implicit", in that you need to specify explicitly how to assemble/ inject the dependencies. flexdi is still explicit in the sense that dependencies are directly referenced from their type annotations, and by leveraging them we can avoid a lot of the more verbose setup required in DeclarativeContainer structures.

• | fastapi <https://github.com/tiangolo/fastapi>_ | This web framework provides an excellent way to perform dependency injection, but it does not provide a way to perform dependency injection outside the context of web request. When configuring the injection, you must also provide default values to arguments, which ties application code to the web framework, making it more difficult to re-use code in other contexts. Additionally, it does not provide rich support for lifetime/singleton scoped dependencies, making the setup of some dependencies increasingly awkward.

• | pinject <https://github.com/google/pinject>_ | This library allows you to perform DI with minimal setup, but its major downfall is that it relies on the names of arguments to perform injection. If the name of the argument does not match the name of the class, then you are forced to bind it explicitly. If there are multiple objects that specify a dependency of a particular type, but use different names, then you need to bind them all individually as well. And sadly, this project has now been archived and is read-only.

Want to make a contribution?

See CONTRIBUTING.md <https://github.com/cal-pratt/flexdi/blob/main/CONTRIBUTING.md>_