Famedly Control Synapse

---

Table of Contents

• Documentation
• Installation
• Configuration
• Testing
• Code Quality
• License

Documentation

This repository contains a Synapse module which allows for creating and managing "managed rooms". Those rooms have the following properties: invite-only, non-federated, cannot be left. Users are associated to groups in an external service and those same groups are associated to rooms in this module. Users are then automatically added to the corresponding rooms. This module is only "half of the work", the other half being in the external service.

The openapi specification is available at openapi-spec.yaml. You can use tools like swagger.io for a more readable format.

Worker routing

The HTTP endpoints exposed by this module are only registered on the background task synapse worker (or the main process if no dedicated worker is configured). Requests to these endpoints on any other worker will return 404 NOT_FOUND.

This must be taken into account when configuring a reverse proxy for example; all requests under /_famedlyControl/ must be routed to the background task worker.

Installation

pip install famedly-control-synapse

Configuration

Here are the available configuration options:

# the outer modules section is just provided for completeness, the config block is the actual module config.
modules:
  - module: "famedly_control_synapse.FamedlyControl"
    config:
        famedly_control:
          api_url: str = "", # Prefix of the current famedly control http API
          access_token: str = "", # Access token to authenticate against famedly control
        sync_enabled: bool = true, # Whether to run the background group membership sync loop
        sync_polling_interval_seconds: int = 30, # Interval in seconds between polling requests
        auth_provider: str = "", # The unique, internal ID of the external identity provider.

Testing

To create a virtual development environment and install dependencies:

hatch shell

The tests use pytest, with the development environment managed by Hatch. Running the tests can be done like this:

hatch test

Additional optional testing arguments:

Run the tests in parallel: -p

Collect coverage data (automatically output as lcov.info): -c

Running a specific test:

Selecting a specific test to run can be as easy as providing the path to the test. All tests start from the base test directory, tests. If running all tests, this can be left out. For specific tests, see the pytest usage docs for more information.

Code Quality

Use hatch fmt to automatically format code, enforce style rules, and check types using:

• black and isort for formatting
• ruff for linting
• mypy for static type checking

Check Code Without Modifying It

To check code quality without modifying files:

• Check formatting with isort and black:

  hatch fmt --check -f

• Check types and linting with mypy and ruff:

  hatch fmt --check -l

• Check all of above, formatting, linting, and typing:

  hatch fmt --check

Auto-formatting Code

To automatically fix issues in the code:

• Format only using black and isort:

  hatch fmt -f

• Type checks(mypy) and lint, fixing autofixable ruff issues:

  hatch fmt -l

• Run all tools, format, lint, type-check:

  hatch fmt

License

famedly-control-synapse is distributed under the terms of the AGPL-3.0 license.