Creating GridgoContext from configuration

It is very convenient to create GridgoContext using configuration, whether it is HOCON, JSON or YAML. The code is very simple:

// create a configurator using TypeSafe, which supports JSON and HOCON
var configurator = TypeSafeConfigurator.ofResource("gridgo-context.conf");

// create the context using the configurator
var context = new ConfiguratorContextBuilder().setRegistry(registry)
                                              .setConfigurator(configurator)
                                              .build();

Following are an example of a conf file:

# the application name
applicationName = "helloworld"

# list of gateways
gateways {

    # define a gateway with name "test" and one subscriber
    test.subscribers += "class:io.gridgo.example.TestProcessor"

    # add another gateway with name "another" and one subscriber
    # with a custom execution strategy and condition
    another.subscribers += {
        processor = "class:io.gridgo.example.AnotherProcessor"
        executionStrategy = "bean:myExecutionStrategy"
        condition = "payload.body.data not empty"
    }

    # add another gateway with 2 connectors and autoStart false
    alsoAnother {
        autoStart = false
        connectors += "kafka:topic1?brokers=localhost:9092"

        # this connector will have a custom ConnectorContextBuilder
        connectors += {
            endpoint = "vertx:http://localhost:8080"
            contextBuilder = "bean:myConnectorContextBuilder"
        }
        subscribers += "class:io.gridgo.example.AlsoAnotherProcessor"
    }
}

# list of components
components += "bean:myComponent"

This is the same configuration using YAML

applicationName: "helloworld"

gateways:
    test:
        subscribers:
            - "class:io.gridgo.example.TestProcessor"
    another:
        subscribers:
            - processor: "class:io.gridgo.example.AnotherProcessor"
              executionStrategy: "bean:myExecutionStrategy"
              condition: "payload.body.data not empty"
    alsoAnother:
        autoStart: false
        connectors:
            - "kafka:topic1?brokers=localhost:9092"
            - endpoint: "vertx:http://localhost:8080"
              contextBuilder: "bean:myConnectorContextBuilder"
        subscribers:
            - "class:io.gridgo.example.AlsoAnotherProcessor"
components:
    - "bean:myComponent"

As you can see the syntax very flexible: for subscribers, connectors you can use either String or Object. The syntax is based on HOCON format, a superset of JSON, which is optimized for human.

For processor, you can either use a bean or a class. If you use a bean, it must be registered in the Registry object passed to the ConfiguratorContextBuilder. If you use a class, it must have exactly one constructor, preferrably non-args. If you use args constructor, the arguments will be looked up from the Registry based on their types. If there are multiple beans applicable for one arguments, it will throw a AmbigiousException.