Make clickbench query IDs 1-based

[pepijnve]

Using 1 as the basis index makes it easier to correlate benchmark runs with query files in an editor

Which issue does this PR close?

None

Rationale for this change

Clickbench query IDs are currently zero-based, while most (if not all) text editors are 1-based when it comes to line numbering. Using 1-based ids makes it easier to correlate benchmark output with the query files.

Additionally TPCH IDs are already 1-based, so this aligns the two.

What changes are included in this PR?

Change clickbench runner to start counting at 1.

Are these changes tested?

Manual testing

Are there any user-facing changes?

Clickbench query IDs in benchmark output are all incremented by one

[AdamGS]

But Clickbench treats/displays them as 0-based

[pepijnve]

Ok, that’s probably a stronger argument for keeping things as is

[pepijnve]

@AdamGS would a PR that splits queries.sql into a file per query be acceptable? There's precedent for that in some of the other benchmarks, and that seems to be a reasonable compromise between maintaining consistent query IDs on the one hand and enabling trivial lookup on the other hand.

[AdamGS]

I'm not a maintainer but I feel your pain always to do that small mental jump when looking for a specific query. I think that's a fine solution.

[pepijnve]

Ok, thanks for your opinion. Little developer quality of life improvements like this are worth it imo. I'll make a PR that contains the necessary code changes and a little bash script that pulls the query file from the upstream clickbench repo and splits it out.

[pepijnve]

Take 2 in #16476