feat(rpc): introduce auth api taikoAuth_lastCertainL1OriginByBatchID

[YoGhurt111]

No description provided.

[claude]

Claude finished @YoGhurt111's task —— View job

---

Code Review Complete ✓

I've reviewed this PR implementing the new taikoAuth_lastCertainL1OriginByBatchID API. Here's my feedback:

---

Summary

This PR adds a new auth API endpoint lastCertainL1OriginByBatchID that retrieves the cached L1 origin for a given batch ID without falling back to chain scanning. This complements the existing lastL1OriginByBatchID endpoint which does perform fallback scanning.

Overall assessment: The implementation is solid with good test coverage. The code is clean, well-tested, and follows the existing patterns in the codebase.

---

Positive Observations

1. Consistent API design - The new endpoint follows the same pattern as the recently added lastCertainBlockIDByBatchID endpoint (lines 83-87, 228-237 in mod.rs)

2. Comprehensive test coverage - Two new tests specifically cover the new endpoint:

   • returns_none_for_last_certain_l1_origin_without_mapping (lines 457-471 in tests.rs)
   • returns_last_certain_l1_origin_from_mapping (lines 475-509 in tests.rs)

3. Correct implementation - The implementation properly:

   • Returns None when no cached mapping exists (early return)
   • Only queries the L1 origin if the batch mapping exists
   • Reuses the existing read_l1_origin_by_block_id helper

4. Clear separation of concerns - The "certain" (cached-only) vs regular (with fallback) distinction is well-maintained

---

Code Quality Issues

Minor: Redundant early return check

crates/rpc/src/eth/auth/mod.rs:232-234

The early return for None could be simplified using ? operator:

// Current
let Some(block_id) = self.read_cached_last_block_number_by_batch_id(batch_id)? else {
    return Ok(None);
};

// Cleaner
let block_id = match self.read_cached_last_block_number_by_batch_id(batch_id)? {
    Some(id) => id,
    None => return Ok(None),
};

// Or even simpler with combinators
return self.read_cached_last_block_number_by_batch_id(batch_id)?
    .map(|block_id| self.read_l1_origin_by_block_id(block_id))
    .transpose();

The last option is most idiomatic Rust and avoids intermediate variables entirely.

---

Test Coverage

Strong points:

• Both success and failure cases are tested
• Integration with the full database stack is covered
• Verification that no fallback scanning occurs (lines 512-548 in tests.rs)

Complete coverage: No additional test cases needed.

---

Performance Considerations

✓ Efficient - This endpoint explicitly avoids expensive chain scanning by design, making it suitable for high-frequency queries.

✓ Database access - Single read-only transaction, no iteration.

---

Security Concerns

✓ No security issues identified

• Read-only operation
• No user-controlled database writes
• Proper error handling (no information leakage)
• Database errors mapped through internal_eth_error

---

Documentation

Missing: The PR description is empty. Consider adding:

• Why this endpoint is needed (use case)
• How it differs from lastL1OriginByBatchID
• When clients should use "certain" vs regular endpoints

---

Recommendations

1. Low priority: Consider using the combinator-based approach for the implementation (see code quality section)
2. PR description: Add context about why this endpoint is needed
3. Before merge: Run just clippy and just test to ensure no regressions

---

Final Verdict

Approve with suggestions ✓

The implementation is correct, well-tested, and ready to merge. The suggested improvements are minor style optimizations that would improve code clarity but aren't blocking issues.

--- | Branch: feature/last_certain_l1_origin_by_batch_id