Flexibly Represented Interactive Data


License
MIT
Install
pip install frid==0.6.1

Documentation

FRID: Flexibly Represented Interactive Data

This Python package is a tool for data manipulation.

Supported data types include:

  • All JSON types: string, integer, floating point, boolean, null, array (as Python lists), and object (as Python dictionaries).
  • Additional data types: binary types (bytes/bytearray/memoryview) and date types (date/time/datetime).
  • Base classes are provided for user-extensible data structures, allowing users to convert between any customized data structures and string representations.

Current key features include:

  • A data serialization/deserialization tool:
    • Data can be dumped into and loaded from a string representation that is more concise than the JSON format.
    • The library is capable of encoding data in fully JSON- or JSON5-compatible formats, including escape sequences in strings to support additional data types.
  • A number of utilities functions and classes, including some asyncio utilities.
  • A key/value store supporting memory, file system, redis, and sqlalchemy as backends.
  • Web application support tools, such as:
    • Converting data from HTTP request bodies based on content type.
    • Converting data types to HTTP requests and setting the correct headers.
    • Sending streaming responses if the data of the body comes from an asynchronous iterator.
    • A mimimal web server routing framework, supporting both WSGi and ASGi.
    • Websocket support for ASGi.

Dependencies:

The package do not have required dependencies, but it needs a few optional dependencies to run some features:

  • redis: for Redis support; also a system command redis-server to run unit tests.
  • sqlalchemy: for SQL support, with some additional packages depending on backend; to run async key/value store with Sqlite, one need aiosqlite and greenlet to run the unit tests. For Postgres, the following are needed:
    • psycopg[binary] for sync Postgres support,
    • asyncpg for async Postgres support.
  • For web servers:
    • gunicorn to run WSGi unit tests.
    • uvicorn to run ASGi unit tests.
  • For websocket support (ASGi only):
    • websockets with uvicorn (or install uvicorn[standard]); it also allows us to run websockets' threading-based sync client for unit tests.
    • websocket-client to run websocket-client-based unit tests.

Other possibilities that haven't been tested:

  • For ASGi: daphne and hypercorn;
  • For ASGi with uvicorn: wsproto is an alternative to websockets.