opt: support tuples, make index constraints for IN

[RaduBerinde]

opt: support Tuple and DTuple when building a scalar

Adding support for tuples, using orderedListOp.

Release Notes: None

opt: index constraints for IN

Release note: None

[cockroach-teamcity]

This change is 

[petermattis]

I'm having deja-vu looking at some of this code. I'm curious how much of it you are basing on the existing span generation code.

---

Review status: 0 of 6 files reviewed at latest revision, 2 unresolved discussions, some commit checks pending.

---

pkg/sql/opt/testdata/build-scalar, line 147 at r2 (raw file):

build-scalar int
(@1) IN (1, 2)

Is there a reason for the parentheses around @1? Ditto below.

---

pkg/sql/opt/testdata/build-scalar, line 155 at r2 (raw file):

      └── const (2) (type: int)

build-scalar,normalize int

What difference are you intending to see between this test and the previous one?

---

Comments from Reviewable

[RaduBerinde]

I'm not actively basing it on the existing code (which has gotten pretty messy over time), but I am familiar with it so I'm probably using some understanding from it. The approach is somewhat different, we generate logical spans directly from expressions, and then manipulate the spans, whereas the existing code extracts a bunch of ComparisonExprs and converts those to spans later.

---

Review status: 0 of 6 files reviewed at latest revision, 2 unresolved discussions.

---

pkg/sql/opt/testdata/build-scalar, line 147 at r2 (raw file):

Previously, petermattis (Peter Mattis) wrote…

Is there a reason for the parentheses around @1? Ditto below.

Fixed

---

pkg/sql/opt/testdata/build-scalar, line 155 at r2 (raw file):

Previously, petermattis (Peter Mattis) wrote…

What difference are you intending to see between this test and the previous one?

Removed, they were left over from an earlier version where I was normalizing orderedListOp to listOp in this case (but I removed that because there's no benefit, at least not at this time).

---

Comments from Reviewable