xxchan

hmmm but that might be complex if we want to preserve comment in the AST

Maybe we can: - change `validator` to sth like `normalizer` and output normalized results. - Output normalized output if `default_validator` is used.

Original added by @Xuanwo https://github.com/risinglightdb/sqllogictest-rs/pull/15 As indicated by its signature, the actual result is separated by columns `Vec`, but the expected result is not. So we need to normalize the...

Cool! I think this is a nice-to-have feature. But the problem 1/2 (clean-up) can be alternatively solved (or bypassed) by `-j1` 😄. For the `.part` problem, this solution can partly...

> > But the problem 1/2 (clean-up) can be alternatively solved (or bypassed) by `-j1` 😄. > > But we may also want to test the `drop` statement. 🥵 `drop`...

Yes, so I said your proposal is a nice-to-have feature 🤤

btw, this includes slt's internal config (`control sortmode`), and DB session's config (`SET xxx`) 🤔 It may be confusing if we only let one of them file-level. But the latter...

Actually `include` by design propagates side-effect (for `CREATE`, `INSERT`). So if the user forgets to clean up (`DELETE`) in `a.slt`, they will also have problem in `b.slt`. But this isn't...

We can have a style suggestion document first 🥸

When file-level behavior is expected, I guess a different semantics for `include` (described here https://github.com/singularity-data/risingwave/issues/3629#issuecomment-1215625871) is actually expected.