feat(db): add audit log for database changes

[whitdog47]

This pull request introduces an audit log table to track changes within the database. Note: as part of this implementation, we have refactored the Base class by moving it from database.core to a new file, database.base, to resolve a circular import issue with the Audit class.

Details:

• Audit Log Implementation:

  • A new audit log table has been created to record changes to specified tables. This is achieved through the track_changes function and the use of SQLAlchemy's event system.
  • The track_changes function is registered as an event listener for SQLAlchemy's "before_flush" event, ensuring that changes are logged before they are committed to the database.
  • Only updates to the specified list of objects are tracked, allowing us to focus on changes that matter most.

• Refactoring:

  • The Base class has been moved to database.base to eliminate a circular import issue caused by the Audit class, which subclasses Base.
  • All relevant imports have been updated across the codebase to reflect this change.

• Code Requirements:

  • To enable auditing, each views.py file must include the addition of the .user attribute to the db_session. This is necessary to associate changes with the current user. For example:

    db_session.user = current_user

  • To get the current_user, the parameter list in each views.py function must be updated to include this line before any database processing:

    current_user: CurrentUser

    This ensures that the current user is passed to the function and can be used for auditing purposes. It may also be necessary add this import for CurrentUser

    from dispatch.auth.service import CurrentUser

  • Lastly, the class name needs to be added to the tracked_classes list in src/dispatch/database/core.py

Impact:

• The introduction of the audit log enhances the system's ability to track and review changes, improving accountability and traceability.
• The refactoring of the Base class should not affect existing functionality but will improve the code's modularity and maintainability.

Testing:

• Existing tests have been run to ensure no functionality is broken.
• New tests have been added to verify the auditing functionality and ensure the correct operation of the track_changes mechanism.