Carl Mäsak
Carl Mäsak
My new favorite slow code (which should probably be turned into a test): ``` (set N 13) (set indent "") (for row 1 N (if indent (pr indent)) (for col...
And maybe one that does effectively a tree traversal, but breaks out of it early, returning the result-so-far.
Another candidate for this is to have a `(breakpoint)` function. [Python has the prior art](https://docs.python.org/3/library/functions.html#breakpoint). Maybe we should call it `(bp)` to better align with Bel's own penchant for brevity.
I just created and ran a version of the code without the `for` loops but with all the rest of the code. It does indeed take 25 minutes. I'm now...
Looking into #200 for evidence of the slowness of arrays, I found [a similar comment](https://github.com/masak/bel/issues/200#issuecomment-1165371973) I had made about four months ago. But back then I wasn't using arrays, so...
This issue could help finally merge #216.
Not a bug. The final `(1 2 3 4 5)` is correct. It would be _nice_ if `(set (some ...) ...)` worked like that, but it simply doesn't. Recall the...
Going to re-open this one, because after #288 there are again some higher-order fastfuncs. Two things: * Have tests for every higher-order function, testing that continuations can go into and...
Looking at this with fresh eyes, the whole problem is never going to be fixed as long as the `Language::Bel` package still has the `call` method. That whole method is...
PR #400 gets us rid of `call`. I'll consider this issue closable when we get our higher-order fastfuncs back.