Expressive allows you to write PSR-7 middleware applications for the web.

PSR-7 is a standard defining HTTP message interfaces; these are the incoming request and outgoing response for your application. By using PSR-7, we ensure that your applications will work in other PSR-7 contexts.

Middleware is any code sitting between a request and a response; it typically analyzes the request to aggregate incoming data, delegates it to another layer to process, and then creates and returns a response. Middleware can and should be relegated only to those tasks, and should be relatively easy to write and maintain.

Middleware is also designed for composability; you should be able to nest middleware and re-use middleware.

With Expressive, you can build PSR-7-based middleware applications:

  • APIs
  • Websites
  • Single Page Applications
  • and more.


Expressive builds on zend-stratigility to provide a robust convenience layer on which to build applications. The features it provides include:

  • Routing

Stratigility provides limited, literal matching only. Expressive allows you to utilize dynamic routing capabilities from a variety of routers, providing much more fine-grained matching capabilities. The routing layer also allows restricting matched routes to specific HTTP methods, and will return "405 Not Allowed" responses with an "Allow" HTTP header containing allowed HTTP methods for invalid requests.

Routing is abstracted in Expressive, allowing the developer to choose the routing library that best fits the project needs. By default, we provide wrappers for Aura.Router, FastRoute, and the zend-mvc router.

  • PSR-11 Container

Expressive encourages the use of Dependency Injection, and defines its Application class to compose a PSR-11 ContainerInterface instance. The container is used to lazy-load middleware, whether it is piped (Stratigility interface) or routed (Expressive).

  • Templating

While Expressive does not assume templating is being used, it provides a templating abstraction. Developers can write middleware that typehints on this abstraction, and assume that the underlying adapter will provide layout support and namespaced template support.

  • Error Handling

Applications should handle errors gracefully, but also handle them differently in development versus production. Expressive provides both basic error handling via Stratigility's own ErrorHandler implementation, providing specialized error response generators that can perform templating or use Whoops.

Flow Overview

Below is a diagram detailing the workflow used by Expressive.

Expressive Architectural Flow

The Application acts as an "onion"; in the diagram above, the top is the outer-most layer of the onion, while the bottom is the inner-most.

The Application dispatches each middleware. Each middleware receives a request and a delegate for handing off processing of the request should the middleware not be able to fully process it itself. Internally, the delegate composes a queue of middleware, and invokes the next in the queue when invoked.

Any given middleware can return a response, at which point execution winds its way back out the onion.


The terminology "pipeline" is often used to describe the onion. One way of looking at the "onion" is as a queue, which is first-in-first-out (FIFO) in operation. This means that the first middleware on the queue is executed first, and this invokes the next, and so on (and hence the "next" terminology). When looked at from this perspective:

  • In most cases, the entire queue will not be traversed.
  • The inner-most layer of the onion represents the last item in the queue, and should be guaranteed to return a response; usually this is indicative of a malformed request (HTTP 400 response status) and/or inability to route the middleware to a handler (HTTP 404 response status).
  • Responses are returned back through the pipeline, in reverse order of traversal.

Double pass middleware

The system described above is what is known as lambda middleware. Each middleware receives the request and the delegate, and you pass only the request to the delegate when wanting to hand off processing:

php function (ServerRequestInterface $request, DelegateInterface $delegate) { $response = $delegate->process($request); return $response->withHeader('X-Test', time()); }

In Expressive 1.X, the default middleware style was what is known as double pass middleware. Double pass middleware receives both the request and a response in addition to the delegate, and passes both the request and response to the delegate when invoking it:

php function (ServerRequestInterface $request, ResponseInterface $response, callable $next) { $response = $next($request, $response); return $response->withHeader('X-Test', time()); }

It is termed "double pass" because you pass both the request and response when delegating to the next layer.

Expressive 2.X still supports double-pass middleware, though we recommend the lambda style.

Stratigility 1.X Error Middleware

The Expressive 1.X series was based on Stratigility 1.X, which had a concept of error middleware. This was middleware that either implemented Zend\Stratigility\ErrorMiddlewareInterface, or the "error middleware" signature of:

php function ( $error, ServerRequestInterface $request, ResponseInterface $response, callable $next )

Such middleware would be invoked by calling $next() with a third argument representing the error:

php return $next($request, $response, $error);

Because this middleware would only be invoked in these special conditions, we recommended piping such middleware last in the queue.

Starting with Expressive 1.1, this feature was deprecated, in favor of using standard middleware with try/catch blocks internally for error handling. As such, we still document it, for our 1.X users, but recommend migrating to the suggested error handling practices.

The Application allows arbitrary middleware to be injected, with each being executed in the order in which they are attached; returning a response from middleware prevents any middleware attached later from executing.

You can attach middleware manually, in which case the pipeline is executed in the order of attachment, or use configuration. When you use configuration, you will specify a priority integer to dictate the order in which middleware should be attached. Middleware specifying high integer prioritiess are attached (and thus executed) earlier, while those specifying lower and/or negative integers are attached later. The default priority is 1.

Expressive provides default implementations of "routing" and "dispatch" middleware, which you either attach to the middleware pipeline manually, or via configuration. These are implemented as the classes Zend\Expressive\Middleware\RouteMiddleware and Zend\Expressive\Middleware\DispatchMiddleware, respectively.

Routing within Expressive consists of decomposing the request to match it to middleware that can handle that given request. This typically consists of a combination of matching the requested URI path along with allowed HTTP methods:

  • map a GET request to the path /api/ping to the PingMiddleware
  • map a POST request to the path /contact/process to the HandleContactMiddleware
  • etc.

Dispatching is simply the act of calling the middleware mapped by routing. The two events are modeled as separate middleware to allow you to act on the results of routing before attempting to dispatch the mapped middleware; this can be useful for implementing route-based authentication or validation.

The majority of your application will consist of routing rules that map to routed middleware.

Middleware piped to the application earlier than routing should be middleware that you wish to execute for every request. These might include:

  • bootstrapping
  • parsing of request body parameters
  • addition of debugging tools
  • embedded Expressive applications that you want to match at a given literal path
  • etc.

Such middleware may decide that a request is invalid, and return a response; doing so means no further middleware will be executed! This is an important feature of middleware architectures, as it allows you to define application-specific workflows optimized for performance, security, etc.

Middleware piped to the application after the routing and dispatch middleware will execute in one of two conditions:

  • routing failed
  • routed middleware called on the next middleware instead of returning a response.

As such, the largest use case for such middleware is to provide a "default" error response for your application, ucually as an HTTP 404 Not Found response.

The main points to remember are:

  • The application is a queue, and operates in FIFO order.
  • Each middleware can choose whether to return a response, which will cause the queue to unwind, or to traverse to the next middleware.
  • Most of the time, you will be defining routed middleware, and the routing rules that map to them.
  • You get to control the workflow of your application by deciding the order in which middleware is queued.