Cfrgtkky

In our code we use Spring Cloud Gateway to dynamically configure routes to services using the fluent java routes API. Lets assume we want to filter on the path of an incoming request. We can implement this in two manners and my question is about the difference between the two:

@Bean

public RouteLocator customRouteLocator(RouteLocatorBuilder builder, ThrottleGatewayFilterFactory throttle) {

    return builder.routes()

        .route(r -> 

           r.host("**.abc.org")

            .and()

            .path("/image/png")

            .uri("http://httpbin.org:80")

        )

        .route(r -> 

           r.host("**.abc.org")

            .uri("http://httpbin.org:80")

            .and(ServerWebExchangeUtils.toAsyncPredicate(

                       new PathRoutePredicateFactory().apply(config -> config.setPattern("/images/jpeg")))

        )

        .build();

}

• In the first route, we configure the path predicate by calling the PredicateSpec.path().




• In the second example we configure the path using Route.AsyncBuilder.and() (Route.AsyncBuilder is the type returned by .uri() method).

Obviously, the first manner is more readable and easy. But I guess there is a use case for defining predicates after the uri() method? What the difference between defining predicates before or after the call to uri()?