The lib is built around a single entry point, the Stream class. This is the only import required. This lib must be trivial to use, all the behaviors it exposes should be carefully named and designed to make them self-explanatory as much as possible. A stream declaration should resemble natural language:
events = Stream(events).foreach(print).truncate(when=lambda event: event["year"] > "2023").catch(KeyError)is relatively close to
Stream the events and for each one, print it, truncate the stream when an event's date is after 2023, and catch
KeyErrors along the way.
Applying an operation just performs some arguments validation and returns a new instance of a Stream sub-class corresponding to the operation. Defining a stream is basically constructing a composite structure where each node is a Stream instance: The above events variable holds a CatchStream[int] instance, whose .upstream attribute points to a TruncateStream[int] instance, whose .upstream attribute points to a ForeachStream[int] instance, whose .upstream points to a Stream[int] instance. Each node's .source attribute points to the same range(10).
Each node in this composite structure exposes an .accept method enabling traversal by a visitor. Both Stream.__iter__ and Stream.__repr__ rely on visitors defined in the package streamable.visitors.
A Stream[T] both inherits from Iterable[T] and holds an Iterable[T] as its .source: when you instantiate a stream from an iterable you decorate it with a fluent interface.
git clone [email protected]:ebonnal/streamable
cd streamable
python -m venv .venv
source .venv/bin/activate
python -m pip install -r requirements-dev.txtpython -m coverage run -m unittest && coverage reportpython -m mypy streamable testspython -m autoflake --in-place --remove-all-unused-imports --remove-unused-variables --ignore-init-module -r streamable tests \
&& python -m isort streamable tests \
&& python -m black .