exhaustive | An annotation and Kotlin compiler plugin | Code Analyzer library

This works, but since setZoom and setCenter aren't used, it feels like an anti-pattern:

I'm not very familiar with fancy match expressions, but, in this very specific case, it seems it would be easier to just use Some _ in the match. You could prove, if you need this, a lemma stating that f1 No is always Some No_parse.

You could also instead define f2 to take an argument of type option (exp No), and delegate the issue to the function caller.

Of course, all this may be stating the obvious...

It looks like you have defined custom paths in your TypeScript config (usually tsconfig.json). The import plugin doesn't know about the correct location of the TypeScript config and hence cannot resolve those paths. What you need to do, is to specify the correct path to your TypeScript config via the project parameter in the resolver options:

The version of base is tied to the version of GHC. It looks like you're on GHC 9.2, which a lot of libraries still don't support yet. You can try passing --allow-newer to cabal, but I expect that to cause other problems. Downgrade to an older major version of GHC instead.

Matching with a singleton list ([x], a list with exactly one element) here is not necessary. You can work with the empty list [] and the "cons" (x:xs):

I modified your query a little, and then got group ids using RANK():

I think option 2 is the way to go. But to make it work you have to disable the warning selectively. This is supported starting with Scala 2.13.2 and 2.12.13

I originally came up with something that seemed overcomplicated so I asked Titian Cernicova-Dragomir to look it over. He was able to simplify it (as I suspected). It turns out the key was something I'd done quite late in the process of building my original: Using & {initialState?: undefined} in one of two overload signatures to add in undefined to the possible types that the data member of the returned object could have.

Here's the result, with explanatory comments. There's an assumption in there: That you want the setData function not to accept undefined even when there's no initialState (and so TData has undefined in it). But there are instructions for removing that if you want setData to accept TData (even when it includes undefined).

Currently, this is a bug.

It seems that the useRouter methods changes useRouter itself. So every time you call one of these methods, useRouter is changing and that leads to this loop.

And the other problem with this is that Next.js is not memorizing useRouter, so it changes even if the value is the same.

Currently, the closest workaround I have found comes from a comment on this open issue https://github.com/vercel/next.js/issues/18127#issuecomment-950907739.

And what it does is that it "converts" useRouter into a useRef and exports the push method. So every time you use this method, this reference won't change if the value didn't change.

Workaround:

I quickly came up with this, which seems to have worked: