planner: fix index range intersection for in-list and other predicates

[ghazalfamilyusa]

What problem does this PR solve?

Issue Number: close #57694

Problem Summary:

This is a limitation for optimal multi-column index range derivation. It is similar to #54166 but different where point ranges are derived from IN list on the prefix column. A simple example is below where the range is based on "t1.a1 IN ( 44, 70, 76)" and not the full predicate "t1.a1 IN ( 44, 70, 76) AND (t1.a1 > 70 OR (t1.a1 = 70 AND t1.b1 > 41))"

CREATE TABLE t1 (
  a1 int,
  b1 int,
  c1 varbinary(767) DEFAULT NULL,
  KEY twoColIndex (a1,b1)
)

EXPLAIN  format = brief
SELECT 1
     FROM t1 FORCE INDEX (twoColIndex)
     WHERE
       t1.a1 IN ( 44, 70, 76)
       AND (t1.a1 > 70 OR (t1.a1 = 70 AND t1.b1 > 41))
--------------

id      estRows task    access object   operator info
Projection      10.07   root            1->Column#5
└─IndexReader   10.10   root            index:Selection
  └─Selection   10.10   cop[tikv]               or(gt(test.t1.a1, 70), and(eq(test.t1.a1, 70), gt(test.t1.b1, 41)))
    └─IndexRangeScan    30.00   cop[tikv]       table:t1, index:twoColIndex(a1, b1)     range:[44,44], [70,70], [76,76]

The optimal solution is below where the range intersection of all predicates is: (70 41,70 +inf], [76,76]

id      estRows task    access object   operator info
Projection      43.33   root            1->Column#5
└─IndexReader   54.17   root            index:IndexRangeScan
  └─IndexRangeScan      54.17   cop[tikv]       table:t1, index:twoColIndex(a1, b1)     range:(70 41,70 +inf], [76,76]

We already have the logic of intersection through prior 54166 PR and it is computed for this case as well. The problem is that we do not choose the intersection for this case.

What changed and how does it work?

The fix is to recognize this case by comparing the conjunctive normal form, or CNF result (applying all predicates) and IN list ranges and pick the most selective. The CNF result is more selective if the covers a superset of filters. In the above example, the CNF result access conditions is "t1.a1 IN ( 44, 45) AND (t1.a1 > 70 OR (t1.a1 = 70 AND t1.b1 > 41))" which is a superset of the IN list access conditions "

Check List

Tests

• Unit test
• Integration test
• Manual test (add detailed scripts or steps below)
• No need to test

  • I checked and no code files have been changed.

Side effects

• Performance regression: Consumes more CPU
• Performance regression: Consumes more Memory
• Breaking backward compatibility

Documentation

• Affects user behaviors
• Contains syntax changes
• Contains variable changes
• Contains experimental features
• Changes MySQL compatibility

Release note

Please refer to Release Notes Language Style Guide to write a quality release note.

None

[tiprow]

Hi @ghazalfamilyusa. Thanks for your PR.

PRs from untrusted users cannot be marked as trusted with /ok-to-test in this repo meaning untrusted PR authors can never trigger tests themselves. Collaborators can still trigger tests on the PR using /test all.

I understand the commands that are listed here.

Details

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes-sigs/prow repository.

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 73.9687%. Comparing base (9812d85) to head (61274f3).
Report is 37 commits behind head on master.

Additional details and impacted files

@@               Coverage Diff                @@
##             master     #58029        +/-   ##
================================================
+ Coverage   73.1839%   73.9687%   +0.7847%     
================================================
  Files          1671       1675         +4     
  Lines        460731     471156     +10425     
================================================
+ Hits         337181     348508     +11327     
+ Misses       102827     101849       -978     
- Partials      20723      20799        +76     

Flag	Coverage Δ
integration	43.8313% <23.8095%> (?)
unit	72.8339% <100.0000%> (+0.5287%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

Components	Coverage Δ
dumpling	52.6910% <ø> (ø)
parser	∅ <ø> (∅)
br	46.2693% <ø> (+0.2649%)	⬆️

[ti-chi-bot]

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: qw4990, winoros

The full list of commands accepted by this bot can be found here.

The pull request process is described here

Details

Needs approval from an approver in each of these files:

• OWNERS [qw4990,winoros]
• pkg/planner/OWNERS [qw4990,winoros]

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

[ti-chi-bot]

[LGTM Timeline notifier]

Timeline:

• 2024-12-10 06:20:59.173926521 +0000 UTC m=+333049.262729061: ☑️ agreed by qw4990.
• 2024-12-10 13:33:29.589299483 +0000 UTC m=+358999.678102025: ☑️ agreed by winoros.

[tiprow]

@ghazalfamilyusa: The following test failed, say /retest to rerun all failed tests or /retest-required to rerun all mandatory failed tests:

Test name	Commit	Details	Required	Rerun command
fast_test_tiprow	61274f3	link	true	/test fast_test_tiprow

Full PR test history. Your PR dashboard.

Details

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes-sigs/prow repository. I understand the commands that are listed here.