Add plan optimizer rules and push expressions

[joseph-isaacs]

Summary

• add typed PlanParentReduceRule and type-erased adapter APIs for child-driven physical-plan rewrites
• add an ordered static PlanParentRuleSet whose first successful rewrite wins
• register concrete Eval reductions for Concat, Take, RowIdx, and Pack
• drive optimization top-down so a rewrite can discard unused children before those children are optimized
• partition bound expressions directly, pushing independent struct-field work into referenced fields while retaining cross-field residuals
• push safe boolean expressions through dictionary values and optimize heterogeneous chunks independently
• preserve global row-index semantics and reduce root-independent expressions through RowIdx, including the empty pack() used by COUNT(*)
• keep recursive traversal in the optimizer: rules return one unoptimized replacement and the optimizer continues from it

Why these are one change

The rule interfaces have no useful behavior without registered rules, and the pushdown implementation defines the invariants the interfaces must support. Keeping them together makes the API reviewable against its real consumers and leaves one complete optimizer commit. This PR therefore absorbs the former #9166, which is closed with no remaining implementation diff.

Safety

• debug builds verify every successful rewrite preserves row count and dtype
• dictionary pushdown requires boolean output, a root reference, strictness, and infallibility
• nullable struct expressions remain above parent validity
• cross-field expressions remain residual expressions over a struct pruned to referenced fields
• chunk pushdown rejects expressions that reference global row indices
• a same-child-type guard prevents a residual expression from immediately reapplying the same rule

Stack

This is the first active PR after merged plan-model PR #9142. It is followed directly by #9187.

Checks

• RUSTC_WRAPPER= cargo test -p vortex-layout — 229 passed
• RUSTC_WRAPPER= cargo clippy -p vortex-layout --all-targets --all-features -- -D warnings
• cargo +nightly fmt --all -- --check
• git diff --check

[codspeed-hq]

Merging this PR will improve performance by 20.44%

⚠️ Unknown Walltime execution environment detected

Using the Walltime instrument on standard Hosted Runners will lead to inconsistent data.

For the most accurate results, we recommend using CodSpeed Macro Runners: bare-metal machines fine-tuned for performance measurement consistency.

⚡ 1 improved benchmark
✅ 1993 untouched benchmarks
⏩ 89 skipped benchmarks¹

Performance Changes

	Mode	Benchmark	BASE	HEAD	Efficiency
⚡	Simulation	cold_misaligned[(64, 256)]	5.3 ms	4.4 ms	+20.44%

Tip

Curious why this is faster? Comment @codspeedbot explain why this is faster on this PR, or directly use the CodSpeed MCP with your agent.

---

_{Comparing vortex-plan-rules (5361eb1) with develop (15c73f3)}

Footnotes

1. 89 benchmarks were skipped, so the baseline results were used instead. If they were deleted from the codebase, click here and archive them to remove them from the performance reports. ↩

[robert3005]

This doesn't work for nested field access since the next level is the same planid, not sure if this is handled in a follow up but you likely want to return some kind of continuation token