Sourcing::Plugin

The Sourcing library uses two plugin roles to handle different aspects of event sourcing:

Plugin Roles

Sourcing provides two plugin roles for different storage concerns:

Sourcing::Plugin::EventStore

Handles event storage operations:

• emit — persisting events to storage
• get-events — fetching events by ID criteria
• get-events-after — fetching events after a version for catch-up
• supply — streaming events in real-time

Sourcing::Plugin::StateCache

Handles projection state caching:

• store-cached-data — caching projection state
• get-cached-data — retrieving cached state for fast recovery

Architecture

This separation enables flexible architectures:

• Different backends — Use PostgreSQL for events but Redis for state cache
• Independent scaling — Scale event store and cache separately
• Hybrid approaches — Mix memory and persistent storage as needed

Combined Implementation

For convenience, Sourcing::Plugin::Memory implements both roles in a single class. This is useful for testing and development.

See Also

• Sourcing::Plugin::EventStore — event storage interface
• Sourcing::Plugin::StateCache — state caching interface
• Sourcing::Plugin::Memory — combined implementation
• Writing a Plugin Guide — how to implement your own plugin