Filter stale rows in constrained bounded scans

[fereidani]

Fixes following example's panic:

(sort V)
(constructor K (i64) V)
(constructor Add (i64 V V) V :cost 100000)
(constructor Shl (i64 V V) V :cost 100000)
(constructor LEA (V V i64 i64) V :cost 1)
(constructor ADD64 (V V) V :cost 1)

(ruleset algebra)
(birewrite (Add w (Add w a b) c) (Add w a (Add w b c)) :ruleset algebra)
(rewrite (Add w a a) (Shl w a (K one))
    :when ((= one (+ w 1))) :ruleset algebra)

(ruleset x64)
(birewrite (ADD64 a b) (Add 64 a b) :ruleset x64)
(rewrite (Add 64 b i) (LEA b i 1 0) :ruleset x64)
(rewrite (LEA b i 1 d) (Add 64 b (Add 64 i (K dv)))
    :when ((= dv (+ d 0))) :ruleset x64)
(rewrite (Add 64 b (Add 64 i (K dv))) (LEA b i 1 d)
    :when ((= d (+ dv 0))) :ruleset x64)

(let $root (ADD64 (K 21) (K 21)))

(run-schedule (repeat 6 (seq (run x64) (run algebra))))

[yihozhang]

Great catch, thank you!

Could you add the reproduction example to the tests folder and add a small comment on RowBuffer::get_row_unchecked that it should not be used if stale rows are undesired?

[codecov-commenter]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 86.59%. Comparing base (53b9721) to head (99c3118).

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #964      +/-   ##
==========================================
- Coverage   86.59%   86.59%   -0.01%     
==========================================
  Files          95       95              
  Lines       29676    29676              
==========================================
- Hits        25699    25698       -1     
- Misses       3977     3978       +1     

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

[codspeed-hq]

Merging this PR will not alter performance

✅ 37 untouched benchmarks
⏩ 227 skipped benchmarks¹

---

_{Comparing fereidani:main (99c3118) with main (53b9721)²}

Footnotes

1. 227 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. ↩

2. No successful run was found on main (99c3118) during the generation of this report, so 53b9721 was used instead as the comparison base. There might be some changes unrelated to this pull request in this report. ↩

[fereidani]

Great project! Happy to contribute! I applied the changes.

[yihozhang]

Thank you!

[yihozhang]

Pushed two small follow-ups:

• Fixed the safety doc comment on RowBuffer::get_row_unchecked (trailing whitespace + a broken sentence with a dangling "either").
• Added the missing cargo insta snapshot for tests/repro-stale-rows.egg — this is why the "test" CI check was failing; make test passes locally now.