Add ARM64 test; Translate non‑benchmark tests to run on GitHub Hosted Runners; Fix some bugs

[euvre]

What problem does this PR solve?

1. Fix cleanup transaction skip logic to be exclusive to cleanup statements

Previously, when a null transaction was detected after BeginTxn(), the code
assumed it was a cleanup skip and returned success. However, this assumption
was not verified, which could mask other errors if a null transaction occurred
for non-cleanup statements.
The fix adds explicit type checking to ensure the skip logic only applies to
cleanup statements (StatementType::kCommand with CommandType::kCleanup). For
any other statement type with a null transaction, an UnrecoverableError is
raised to properly report the unexpected condition.

2. Fix SegmentTreeForTopnDistance member initialization order

Fixed a critical bug where inner_index_ array was always allocated with size 0
due to C++ member initialization order. In the original code:

• topn_ was default-initialized to 0 at declaration
• inner_index_ immediately used topn_ << 1 (evaluating to 0) for allocation
• The constructor then set topn_ to the correct value, but too late

This caused undefined behavior when accessing inner_index_ elements in Build()
and Update(), leading to non-deterministic search results.
The fix moves inner_index_ initialization into the constructor, using the
function parameter topn instead of the member topn_ to ensure correct
allocation size.

Type of change

• Test cases
• Bug Fix

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 50.62%. Comparing base (8ff61e2) to head (78f3ada).
⚠️ Report is 3 commits behind head on main.

Additional details and impacted files

@@ Coverage Diff @@ ## main #3338 +/- ## ========================================== - Coverage 50.67% 50.62% -0.06%  ========================================== Files 793 793 Lines 155947 155949 +2 Branches 28580 28580 ========================================== - Hits 79024 78944 -80  - Misses 66596 66639 +43  - Partials 10327 10366 +39 

Flag	Coverage Δ
debug http test	27.80% <ø> (+0.13%)	⬆️
debug parallel test	18.89% <ø> (-0.08%)	⬇️
debug pysdk test	29.60% <ø> (+<0.01%)	⬆️
debug sqllogical test	32.95% <ø> (+0.02%)	⬆️
debug unit test	38.14% <ø> (-0.03%)	⬇️

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

☔ View full report in Codecov by Sentry.
📢 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.