fix `ProveField` warnings in `results`

[etan-status]

Nim emits ProveField warnings with if on case object discriminator. Replace with case instead to avoid those warnings. Note we currently have ProveField disabled but it keeps showing up sometimes when compiling with nim c instead of make.

[etan-status]

Note, this doesn't get rid of all ProveFields warnings, as some need to suppress it at call side (e.g., var).

However, it still, overall, reduces the amount of warnings quite a lot, so is possibly a net gain overall.

[etan-status]

Nim devel 🤔 (seems to be alright on master)

/home/runner/.nimble/pkgs2/unittest2-0.0.7-b6d4a5cbe28b43c166d6442ba6804aafd4abe368/unittest2.nim(900) test_interval_set
/home/runner/work/nim-stew/nim-stew/tests/test_interval_set.nim(129) runTest`gensym123
/home/runner/work/nim-stew/nim-stew/stew/interval_set.nim(1164) verify
/home/runner/work/nim-stew/nim-stew/stew/interval_set.nim(345) new
/home/runner/work/nim-stew/nim-stew/stew/interval_set.nim(295) right
/home/runner/work/nim-stew/nim-stew/stew/interval_set.nim(289) blk
/home/runner/work/nim-stew/nim-stew/nim/lib/system/arc.nim(43) nimIncRef
SIGSEGV: Illegal storage access. (Attempt to read from nil?)

[etan-status]

@araq re Nim devel issue. Something in this PR seems to mess with orc.

[Araq]

Same problem as https://github.com/nim-lang/Nim/issues/22058 ?

[etan-status]

Same problem as nim-lang/Nim#22058 ?

Not sure. The linked problem also occurs on Nim 1.6, this one here is devel only.

There is also no {.noSideEffects.} here

[tersec]

Nim devel issues should not block development progress. I stand by what I wrote in https://github.com/status-im/nim-ssz-serialization/pull/47#issuecomment-1584890664

But it remains that the primary purpose of the nim-ssz-serialization CI is to make nim-ssz-serialization in the context of supported Nim versions (of which constantly-changing devel is not one), so if devel in the future causes too-regular CI regressions, it might prove warranted to reconsider this situation.

s/nim-ssz-serialization/nim-stew/g, but otherwise identical.

[etan-status]

@ringabout Reran on Nim fda8b6f193e2e229488f76f18089f01eb08272fb which includes https://github.com/nim-lang/Nim/pull/22088

It's still broken here :) So, likely a separate bug.

[ringabout]

Seems to be a 2.0 to devel regression.

[arnetheduck]

Nim devel issues should not block development progress.

they shouldn't, but we can also use the opportunity to rid upstream of debilitating bugs that might come to bite us later - ie this is not an urgent PR in the sense that there's no runtime bug being fixed so we can merge it when convenient

[tersec]

Nim devel issues should not block development progress.

they shouldn't, but we can also use the opportunity to rid upstream of debilitating bugs that might come to bite us later - ie this is not an urgent PR in the sense that there's no runtime bug being fixed so we can merge it when convenient

These can happen simultaneously: if/as useful, create a branch, destined never to be merged, which can function as a repro for the bug in question. Then fix master. Then both activities can happen in parallel: upstream Nim can fix as best suits them, and meanwhile, nim-stew development isn't hindered.

[ringabout]

I'm fixing it, which seems to be a piece of cake => https://github.com/nim-lang/Nim/pull/22093

[etan-status]

Looks good now on devel