Awesome

HTTP Handling For JSON Event Sourcing

The JSON Event Sourcing solution works with Kafka and MongoDB. The library in this package provides the HTTP handling for that environment. You can use it to easily enable that logic in a Java HTTP server solution. An example is the pincette-jes-http repository, which uses a Netty HTTP server and delegates everything to this library.

The URL path for an aggregate always has the form [/context][/application]/aggregate_type[/id]. When the identifier is set, one specific aggregate instance is addressed. Without it the complete collection of aggregates of that type is addressed. For example, a POST would do a search on all the aggregate instances.

When aggregates change because of a command the result is sent back through the Kafka reply topic. It is possible to push this to the client using Server-Sent Events. This is done with the fanout.io service. The endpoint /sse is where the client should connect to. It will be redirected to the fanout.io service with the encrypted username in a URL parameter. Then fanout.io comes back to the /sse-setup endpoint, where the channel is created.

Requests need to have a valid JSON Web Token. They are taken from the bearer token in the Authorization header. Then the cookie access_token is tried.

Methods

Path	Method	Description
[/app]/type/id	GET	Fetches an aggregate instance.
	POST	Sends a command to an aggregate instance.
	PUT	Replaces an aggregate instance.
[/app]/type	POST	Performs a search in a collection of aggregates of some type. The request body should be a JSON array containing MongoDB aggregation pipeline stages. If the aggregate never uses the `_acl` field, then you can add the optional URL-parameter `noacl=true` to simplify the query. Otherwise the full ACL expression will be added to the incoming query.
/sse	GET	The Server-Sent Events endpoint for clients.
/sse-setup	GET	The Server-Sent Events endpoint for fanout.io.