Creative Racket Competition 2022 https://racket.discourse.group/t/creative-racket-competition-2022-january-1-2022-february-28-2022/454?u=spdegabrielle
Is it possible with raco test to run only a specific test suite (e.g. specified by it’s name)? I haven’t found anything in the docs.
No, there isn’t any functionality like that, although it would be nice to have.
I got the hint. :wink: I’m putting it on my todo list. :slightly_smiling_face:
I wasn’t trying to hint, just saying that it wasn’t intentionally omitted. :wink:
Sorry, I couldn’t help it. :smile:
But if you wanted to do it, then I think a good approach would be (a) having raco test take an argument that sets some parameter that (b) rackunit or other testing frameworks could read and decide to run/not run some tests.
I added a ticket to avoid that I forget your suggested approach (and also as a reference for me, of course): https://github.com/racket/racket/issues/4107
Hi @badkins! @soegaard2 told me that you found in-inclusive-range to be slower than in-range sometimes. Could you provide some examples?
<sigh> I didn’t even realize there was in-inclusive-range :( I was using my own naive function, inclusive-range, which simply returns a list :) Thanks for the tip!!
Ah!
Yeah, in-inclusive-range should be as fast as in-range. FWIW, inclusive-range is also an actual function already pre-defined in Racket, too!
Of course, it’s even faster to code the loop manually, which is what I’m doing now…
Another thing is, even though range and inclusive-range return a list, if it appears directly within a for clause, it should be as efficient as in-range/in-inclusive-range too.
I have changed some APIs in a module and now want to update the Scribble documentation. Since the documentation contains several executable code snippets, the document generation fails. Is there a way to get the documentation partially generated on a “best effort” basis, so it’s easier for me to see where I still need to make changes?
For now, I’ll comment out everything from the broken location until the end of the document, so I can at least generate the documentation up to the comment.
> a “best effort” basis, so it’s easier for me to see where I still need to make changes What happens currently? Does it refuse to render the document?
I think this should not be happening for run-time error—Scribble can already cope with that, but for compile-time error, I’m not sure. I thought it can cope with it too, but I might misremember
Without the commenting, it fails because I changed the name of a function that’s used in one of the code examples. So I get an error for an unbound identifier.
> but for compile-time error, I’m not sure. I thought it can cope with it too, but I might misremember Yes, I think my problem qualifies as a compile-time error (assuming the processing is like in #lang racket).
OK, I remember it wrong. Looks like both run-time and compile-time error result in failure alike
I wrote a package that does something similar, except that you need to modify the run-tests invocation instead of using the command line to omit test suites. You could use that package as a starting point for what you want to do;
I’ll try it tomorrow because I’m too tired now, but from the documentation the package looks great! :slightly_smiling_face:
@sschwarzer add this to your Scribble file at the top
@(let ([this-handler (scribble-eval-handler)])
(scribble-eval-handler
(λ (f show? expr)
(with-handlers ([exn:fail? values])
(this-handler f show? expr)))))