ldb dump_wal: add LOG_DATA, TIMED_PUT, PUT_BLOB_INDEX

[evanj]

Previously, LOG_DATA would be ignored and would print like an empty record, and TIMED_PUT and PUT_BLOB_INDEX would cause ldb to exit with an unimplemented error.

Add a unit test to cover these cases.

[evanj]

Rebased on main. This is still relevant.

[github-actions]

✅ clang-tidy: No findings on changed lines

Completed in 341.2s.

[github-actions]

✅ Claude Code Review

Auto-triggered after CI passed — reviewing commit 2072d5c

---

Code Review: ldb dump_wal: add LOG_DATA, TIMED_PUT, PUT_BLOB_INDEX

APPROVE with suggestions. The implementation is correct, follows existing patterns, and solves a real user-facing problem.

Findings

F1. Test only asserts LOG_DATA — missing TIMED_PUT and PUT_BLOB_INDEX assertions (MEDIUM)

The test writes all three record types but only verifies LOG_DATA. The other two handlers could be broken and the test would still pass. Add:

ASSERT_NE(std::string::npos, captured_output.find("TIMED_PUT("))
    << "ldb output:\n\n" << captured_output;
ASSERT_NE(std::string::npos, captured_output.find("PUT_BLOB_INDEX("))
    << "ldb output:\n\n" << captured_output;

F2. stdout capture not restored on early test exit (LOW)

The rdbuf swap pattern is not exception/early-return safe. An RAII guard would be more robust. This is a novel pattern in the RocksDB test suite.

F3. TimedPutCF format includes write_time in parentheses (SUGGESTION)

TIMED_PUT(cf, write_time) is mildly inconsistent with PUT(cf), MERGE(cf), etc. However, COMMIT_WITH_TIMESTAMP(xid, ts) provides precedent. Current format is acceptable.

F4. Consider decoding blob index for readability (SUGGESTION)

BlobIndex::DebugString() could provide human-readable output instead of raw hex. Nice-to-have, not required.

Full review written to review-findings.md.

---

ℹ️ About this response

Generated by Claude Code.
Review methodology: claude_md/code_review.md

Limitations:

• Claude may miss context from files not in the diff
• Large PRs may be truncated
• Always apply human judgment to AI suggestions

Commands:

• /claude-review [context] — Request a code review
• /claude-query <question> — Ask about the PR or codebase