planner: fix the join scope with explicit parents (#33311) #33418
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
cherry-pick #33311 to release-6.0
You can switch your code base to this Pull Request by using git-extras:
# In tidb repo: git pr https://github.com/pingcap/tidb/pull/33418
After apply modifications, you can push your change to this PR via:
Signed-off-by: AilinKid [email protected]
What problem does this PR solve?
Issue Number: close #31770
Problem Summary:
What is changed and how it works?
after this pingcap/parser#1129, tidb parser will reorganize join table regardless the explicit parents
Eg: t1 join (t2 join t3) will be reorged as (t1 join t2) join t3
parser/1129 seek to solve the problem
1: with no explicit parents join sequence like: t1 join t2 join t3, make them left-associative
2: with explicit parents and can be reorder the join sequece, like: a join b right join c, we find it here, change the right join as left join, then with the logical semantic same, rewrite it as a join c left join b
Although it's a nice idea, we still has something to do before that, like name resolution in parents scope. like:
select t1.* from t t0 cross join (t t1 join t t2 on 100=t0.a)
t0.a here shouldn't be seen by the join inside. But after it's rewrite it's different
select t1.* from (t t0 cross join t t1) join t t2 on 100=t0.a
this kind of join reorder can be detected and improved in logical optimization.
Check List
Tests
Side effects
Documentation
Release note