refactor(optimizer): type-safe plan base with compile-time convention check

[BugenZhao]

Signed-off-by: Bugen Zhao [email protected]
I hereby agree to the terms of the RisingWave Labs, Inc. Contributor License Agreement.

What's changed and what's your intention?

Reviewing the following files should be sufficient:

• batch.rs
• stream.rs
• mod.rs
• plan_base.rs

---

Follow-up of #12980. This PR further makes PlanBase over a generic ConventionMarker.

risingwave/src/frontend/src/optimizer/plan_node/plan_base.rs

Line 112 in 35172ec

pub struct PlanBase<C: ConventionMarker> {

As the result, the base field in the plan node struct will specify the corresponding ConventionMarker.

risingwave/src/frontend/src/optimizer/plan_node/stream_hash_agg.rs

Lines 30 to 32 in 35172ec

	#[derive(Debug, Clone, PartialEq, Eq, Hash)]
	pub struct StreamHashAgg {
	pub base: PlanBase<Stream>,

Any attempts to access fields with a mismatched convention (like logical_agg.emit_on_window_close() or stream_hash_agg.plan_base().order()) will be a compile error.

Accessing the PlanBase through PlanNodeMeta trait will now also get a type with specific convention, which makes the trait not object-safe anymore.

risingwave/src/frontend/src/optimizer/plan_node/mod.rs

Lines 95 to 102 in 35172ec

	/// The trait for accessing the meta data and [`PlanBase`] for plan nodes.
	pub trait PlanNodeMeta {
	type Convention: ConventionMarker;
	const NODE_TYPE: PlanNodeType;
	/// Get the reference to the [`PlanBase`] with corresponding convention.
	fn plan_base(&self) -> &PlanBase<Self::Convention>;

However, there're still a lot of places where PlanRef are directly worked on. To still allow erasing the type of a plan node, the original PlanNodeMeta turns into AnyPlanNodeMeta and is automatically implemented.

risingwave/src/frontend/src/optimizer/plan_node/mod.rs

Lines 115 to 140 in 35172ec

	/// The object-safe version of [`PlanNodeMeta`], used as a super trait of [`PlanNode`].
	///
	/// Check [`PlanNodeMeta`] for more details.
	pub trait AnyPlanNodeMeta {
	fn node_type(&self) -> PlanNodeType;
	fn plan_base(&self) -> PlanBaseRef<'_>;
	fn convention(&self) -> Convention;
	}
	/// Implement [`AnyPlanNodeMeta`] for all [`PlanNodeMeta`].
	impl<P> AnyPlanNodeMeta for P
	where
	P: PlanNodeMeta,
	{
	fn node_type(&self) -> PlanNodeType {
	P::NODE_TYPE
	}
	fn plan_base(&self) -> PlanBaseRef<'_> {
	PlanNodeMeta::plan_base_ref(self)
	}
	fn convention(&self) -> Convention {
	P::Convention::value()
	}
	}

Future work may further explore reducing the usage of type-erased PlanRef. For example, we can at least constrain the convention for it.

Checklist

• I have written necessary rustdoc comments
• I have added necessary unit tests and integration tests
• All checks passed in ./risedev check (or alias, ./risedev c)

Documentation

• My PR needs documentation updates. (Please use the Release note section below to summarize the impact on users)

Release note

If this PR includes changes that directly affect users or other significant modifications relevant to the community, kindly draft a release note to provide a concise summary of these changes. Please prioritize highlighting the impact these changes will have on users.

Signed-off-by: Bugen Zhao

[TennyZhuang]

LGTM!

[stdrc]

LGTM

[stdrc]

Redundant import

[chenzl25]

Will take a look later. Compile error rather than runtime panic sounds pretty cool!

[chenzl25]

Rest LGTM!

[chenzl25]

I am wondering whether we can expose a different property fetching interface for PlanRef to let developers know which (logical, batch or stream) properties they want to access explicitly. For example, plan_ref.batch_property().unwrap() to access the batch property with an explicit unwrap.

[chenzl25]

Because accessing a property in PlanRef could panic underlying potentially might require a very high code coverage rate to ensure them work well. Exposing an Option or Result can help developers realize what are they doing right now and the potential result. At least we can throw an Error instead of panicking the whole system.

[BugenZhao]

let developers know which (logical, batch or stream) properties they want to access explicitly

Yeah, this is a good point. Several reasons for not doing this now:

• I want to make the changes minimal in this PR.
• We may first further explore if it's possible to constrain PlanRef to something like PlanRef<Convention>, to reduce the usages of accessing on the convention-erased plan.

Because accessing a property in PlanRef could panic underlying potentially might require a very high code coverage rate to ensure them work well.

Agree.

At least we can throw an Error instead of panicking the whole system.

That's not a big problem. We always catch the panic when handling a SQL statement, so panicking will not affect other active sessions.

[gitguardian]

⚠️ GitGuardian has uncovered 1 secret following the scan of your pull request.

Please consider investigating the findings and remediating the incidents. Failure to do so may lead to compromising the associated services or software components.

🔎 Detected hardcoded secret in your pull request

GitGuardian id	Secret	Commit	Filename
7648795	Generic CLI Secret	b57e395	integration_tests/iceberg-cdc/run_test.sh	View secret

🛠 Guidelines to remediate hardcoded secrets

1. Understand the implications of revoking this secret by investigating where it is used in your code.
2. Replace and store your secret safely. Learn here the best practices.
3. Revoke and rotate this secret.
4. If possible, rewrite git history. Rewriting git history is not a trivial act. You might completely break other contributing developers' workflow and you risk accidentally deleting legitimate data.

To avoid such incidents in the future consider

• following these best practices for managing and storing secrets including API keys and other credentials
• install secret detection on pre-commit to catch secret before it leaves your machine and ease remediation.

---

<sup>🦉 GitGuardian detects secrets in your source code to help developers and security teams secure the modern development process. You are seeing this because you or someone else with access to this repository has authorized GitGuardian to scan your pull request.

Our GitHub checks need improvements? Share your feedbacks!</sup>

[codecov]

Codecov Report

Merging #13000 (b57e395) into main (942e99d) will decrease coverage by 0.03%.
Report is 2 commits behind head on main.
The diff coverage is 87.24%.

@@            Coverage Diff             @@
##             main   #13000      +/-   ##
==========================================
- Coverage   68.33%   68.31%   -0.03%     
==========================================
  Files        1496     1498       +2     
  Lines      252058   252102      +44     
==========================================
- Hits       172253   172217      -36     
- Misses      79805    79885      +80     

Flag	Coverage Δ
rust	68.31% <87.24%> (-0.03%)	⬇️

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

Files	Coverage Δ
src/frontend/src/handler/explain.rs	79.19% <ø> (ø)
src/frontend/src/optimizer/mod.rs	92.21% <ø> (ø)
...c/frontend/src/optimizer/plan_node/batch_delete.rs	75.00% <100.00%> (ø)
...frontend/src/optimizer/plan_node/batch_exchange.rs	69.56% <ø> (ø)
...c/frontend/src/optimizer/plan_node/batch_expand.rs	46.34% <ø> (ø)
...c/frontend/src/optimizer/plan_node/batch_filter.rs	97.61% <ø> (ø)
...ontend/src/optimizer/plan_node/batch_group_topn.rs	58.97% <ø> (ø)
...frontend/src/optimizer/plan_node/batch_hash_agg.rs	96.84% <ø> (ø)
...rontend/src/optimizer/plan_node/batch_hash_join.rs	90.54% <ø> (ø)
...ontend/src/optimizer/plan_node/batch_hop_window.rs	75.49% <ø> (ø)
... and 84 more

... and 5 files with indirect coverage changes

📣 We’re building smart automated test selection to slash your CI/CD build times. Learn more