I define custom errors in appErrors.ts.

I create an AppError class and use it as super generic error object for my application. Why don't I use just Error? Cause I want the generic error to hold status codes.

Then I use AppError to create more specific error classes, if I am sure that I will use them again and again. For example, I use EmptyFieldError every time when I check objects for properties.

Check recordsAPI.ts to see how I throw these errors.

If I use trycatch, in a catch statement I always put an error inside next(). That way it is always going to be caught by handleError middleware.

catch (error) {
    next(error);
}

handleErrors.ts – is a typical express error handler middleware, since it takes an error as a first argument. It's going to catch any error passed to the next() as an argument.

→ Check express guide: Writing error handlers for more information about express error handler middleware

handleErrors is mounted the last in app.ts, so that it would be able to catch errors after controllers throw them.

Inside this middleware you can check caught Error for error name or status code and decide what to do with it depending on these values.

Let's go to the life circle of an error.

1. Throw an error some where

   if (typeof dateStr !== "string")
         throw new WrongTypeError({ date: dateStr }, "string");

2. Catch-statement catches this error and sends it to the next()

   catch (error) {
       next(error);
   }

3. handleError (error handler middleware) gets this error and decides what to do with it depending in its name or/and status code: logs it or/and sends appropriate response.