Kata

Needs true random tests. The current "random" tests just consists of different values of the same route.

Also needs more fixed tests. There is only ever 1 route hash tested for the kata.

The methological and design of the kata is incorrect: segments refer to URL segments, which is the path of the URL further delimited by dashes. Hence a method named .segments should not return the named hash, it should return an array of the URL segments. The named segments would be the params captured by the route.

Even then, storing the params in the object as a mutable state on Router is insane design, because it doesn't belong to Router, but specific to a request itself. Instead of returning ["CommentsController", :edit] and storing { slug: "test-post", id: "12" } into .segments the method should really return ["CommentsController", :edit, { slug: "test-post", id: "12" }] (with empty hash or nil as the third element when no named segments exist). Every MVC framework does this so the action will actually get the params as it is invoked by the framework.

Test.expect should not be used.

Took me a bit. but figured it out. thank you!

Thank you for the fun Kata!

The explanation could use a little work. Maybe break down the methods into separate sections.

I also think that the fact that you have to save the request_info in an instance variable in the route? method is kinda ganky.

In retrospect, I am guessing that the intention is for the solution to handle various different ROUTES hashes in this format rather than the example ROUTES hash exclusively. If my interpretation is correct, some additional test cases would be beneficial to weed out solutions that do not handle variations.