unasync

Welcome to unasync, a project that can transform your asynchronous code into synchronous code.

See urllib3/urllib3#1323 for more details about unasync.

Users

• The official Elasticsearch Python client: https://github.com/elastic/elasticsearch-py
• Hip, a HTTP client: https://github.com/python-trio/hip
• httpcore, another low-level HTTP client uses a different implementation of the same idea: https://github.com/encode/httpcore

Installation

pip install unasync

Usage

To use the unasync project you need to install the package and then create a _async folder where you will place the asynchronous code that you want to transform into synchronous code.

And then in your setup.py place the following code.

import unasync

setuptools.setup(
    ...
    cmdclass={'build_py': unasync.cmdclass_build_py()},
    ...
)

And when you will build your package you will get your synchronous code in _sync folder.

If you'd like to customize where certain rules are applied you can pass customized unasync.Rule instances to unasync.cmdclass_build_py()

import unasync

setuptools.setup(
    ...
    cmdclass={'build_py': unasync.cmdclass_build_py(rules=[
        # This rule transforms files within 'ahip' -> 'hip'
        # instead of the default '_async' -> '_sync'.
        unasync.Rule("/ahip/", "/hip/"),

        # This rule's 'fromdir' is more specific so will take precedent
        # over the above rule if the path is within /ahip/tests/...
        # This rule adds an additional token replacement over the default replacements.
        unasync.Rule("/ahip/tests/", "/hip/tests/", additional_replacements={"ahip": "hip"}),
    ])},
    ...
)

Documentation

https://unasync.readthedocs.io/en/latest/

License: Your choice of MIT or Apache License 2.0