sql: improved partial index implication with virtual columns

[RaduBerinde]

The issue tracks figuring out how to prove implication (and thus enable the use of the partial index) for cases like:

CREATE TABLE t (
  k INT PRIMARY KEY,
  a INT,
  b INT,
  c INT AS (b + 10) VIRTUAL,
  INDEX (a) WHERE c IN (10, 20, 30)
)

SELECT k FROM t WHERE c = 20

Jira issue: CRDB-2993

[github-actions]

We have marked this issue as stale because it has been inactive for
18 months. If this issue is still relevant, removing the stale label
or adding a comment will keep it active. Otherwise, we'll close it in
10 days to keep the issue queue tidy. Thank you for your contribution
to CockroachDB!

[yuzefovich]

cc @mgartner has this been addressed?

[mgartner]

This was fixed for some cases in #123163. However, the specific case mentioned in this issue has not been fixed. See: https://github.com/mgartner/cockroach/blob/3aa3d6f135cf09bae622747d670a22c94e781785/pkg/sql/opt/partialidx/testdata/implicator/false-negative#L53-L60

# We do not prove implication in this case because a + 10 > 4 is normalized to
# a > -6 before a + 10 can be replaced with b.
predtest vars=(a int, b int as (a + 10) virtual)
a + 10 > 4
=>
a + 10 IS NOT NULL
----
false