Binding

In Camel terms a binding is a way of wrapping an Endpoint in a contract; such as a Data Format, a Content Enricher or validation step. Bindings are completely optional and you can choose to use them on any camel endpoint.

Bindings are inspired by the work of SwitchYard project adding service contracts to various technologies like Camel and many others. But rather than the SwitchYard approach of wrapping Camel in SCA, Camel Bindings provide a way of wrapping Camel endpoints with contracts inside the Camel framework itself; so you can use them easily inside any Camel route.

Using Bindings

A Binding is currently a bean which defines the contract (though we’ll hopefully add bindings to the Camel DSL).

There are a few approaches to defining a bound endpoint (i.e. an endpoint bound with a Binding).

Using the binding URI

You can prefix any endpoint URI with binding:nameOfBinding: where nameOfBinding is the name of the Binding bean in your registry.

from("binding:jaxb:activemq:myQueue").to("binding:jaxb:activemq:anotherQueue")
java

Here we are using the "jaxb" binding which may, for example, use the JAXB Data Format to marshal and unmarshal messages.

Using a BindingComponent

There is a Component called BindingComponent which can be configured in your Registry by dependency injection which allows the creation of endpoints which are already bound to some binding.

For example if you registered a new component called "jsonmq" in your registry using code like this

        JacksonDataFormat format = new JacksonDataFormat(MyBean.class);
        context.bind("jsonmq", new BindingComponent(new DataFormatBinding(format), "activemq:foo."));
java

Then you could use the endpoint as if it were any other endpoint.

from("jsonmq:myQueue").to("jsonmq:anotherQueue")
java

This would use the queues "foo.myQueue" and "foo.anotherQueue" and would use the given Jackson Data Format to marshal on and off the queue.

When to use Bindings

If you only use an endpoint once in a single route; a binding may actually be more complex and more work than just using the 'raw' endpoint directly and using explicit marshalling and validation in the camel route as normal.

However bindings can help when you are composing many routes together; or using a single route as a 'template' that is configured input and output endpoints; bindings then provide a nice way to wrap up a contract and endpoint together.

Another good use case for bindings is when you are using many endpoints which use the same binding; rather than always having to mention a specific data format or validation rule, you can just use the BindingComponent to wrap the endpoints in the binding of your choice.

So bindings are a composition tool really; only use them when they make sense - the extra complexity may not be worth it unless you have lots of routes or endpoints.