Skip to content

Middleware

Middlewares are a mechanism to execute logic before and/or after the request handler.
Middlewares are often used to implement cross-cutting functionality, such as:

  • telemetry (e.g. structured logging, metrics, etc.)
  • load-shedding (e.g. timeouts, rate-limiting, etc.)
  • access control (e.g. authentication, authorization, etc.)
  • etc.

Middleware types

Pavex provides three types of middlewares: pre-processing, post-processing, and wrapping middlewares.
As the naming suggests, they differ in when they start and complete their execution, making them suitable for different use cases.

Request processing pipeline

In this guide we'll often talk about the request processing pipeline.
This term refers to the sequence of components that handle a request, from the moment it arrives to the moment the response is sent back to the caller.
It includes, in particular, the request handler and all the middlewares that apply to that route.

At a glance:

Type Starts Completes Suitable for
Pre-processing Before the request handler Before the request handler Skipping the remaining processing, returning an early response.
Example: rejecting unauthenticated requests.
Post-processing After the request handler After the request handler Modifying the response and/or performing side-effects based on the its contents.
Examples: logging the response's status code, injecting headers.
Wrapping Before the request handler After the request handler Accessing the future representing the rest of the request processing pipeline.
Examples: enforcing a timeout, attaching a tracing span to the request processing pipeline.

Each middleware type has a dedicated section in this guide. Check them out for more details!