StateView

StateView

One role of the StateView component is to provide the state of an aggregate using the cache, with the event store as a fallback, leveraging memoization techniques.

There are two ways to get the state of an aggregate:

getAggregateFreshState and getAggregateFreshStateAsync

Single-node, non-distributed applications typically use the StateView to retrieve the state of an aggregate. However, it is also possible to run independent nodes that implement a “StateView as a service”—listening to events on a message bus, rebuilding aggregate states, and providing a fallback mechanism to resynchronize with the event store. Some experimental examples demonstrate this using RabbitMQ. Generally, we can provide a StateViewer object when constructing the application layer, deciding which state viewer implementation to adopt. To use the standard cache+event store state viewer, we can use getAggregateStorageFreshStateViewer or getAggregateStorageFreshStateViewerAsync from the CommandHandler component. These can be called in a curried style, passing the event store to return the required stateReader function. Many existing examples use this pattern.

The StateView also provides functions to dispatch calls that update the details cache. By managing dependencies, it ensures the details cache is aware of changes to aggregate states and triggers refreshes of dependent details.

A more detailed explanation will come soon.