El uso de midje proporcionado en una cláusula let no agrega métodos

When I make a test using an outer let clause to structure some definitions and calls, the stubs don't work the way I'd expect. For example:

Esta prueba falla 

(fact "blah"
  (let [x (meth1 123)]
    x => 246
    (provided
      (meth2 123) => 246)))

Con este código 

(defn meth2 [x]
  (prn "meth2" x)
  (* 3 x))

(defn meth1 [x]
  (let [y (meth2 x)]
    y))

Am I not supposed to use let statements with midje? I can't understand how to get these to pass without removing the let.

preguntado el 27 de noviembre de 13 a las 05:11

1 Respuestas

First of all, your test would even fail if meth2 was stubbed correctly since (fn [] x) returns a function, thus the fact (fn [] x) => 246 will never hold.

Secondly, I think that provided is used to stub function calls only when ejecución the left/right-hand-side of facts. In your case that's (fn [] x) (y x is already evaluated at that point), as well as 246 (which is constant). meth1 never gets called in the context of your fact, only beforehand.

To change that, you could do one of two things. Either you make the let part of the left-hand-side of your fact:

...
(let [x (meth1 123)] x) => 246
...

O haces x a function that is evaluated when the fact is tested:

...
(let [x #(meth1 123)] (x)) => 246
...

I don't think there is a way to really ver provided in action, i.e. like this:

(let [x (meth1 123)]
   x => 369
  (let [...]
    x => 246
    (provided 
      (meth2 123) => 246))))

A let wrapping facts seems to be executed before the first fact is touched.

respondido 27 nov., 13:09

This is what I ended up going with. Along with using contexts for setup and cleanup, instead of a let statement - aciniglio

No es la respuesta que estás buscando? Examinar otras preguntas etiquetadas or haz tu propia pregunta.