fix: persist frag reuse index external file on local filesystem

[wjones127]

Previously, when FragReuseIndexDetails exceeded 204800 bytes (triggered by large compactions with many fragments), the code wrote the details to an external file (details.binpb). On local filesystems, ObjectStore::create returns a LocalWriter that atomically renames a temp file to the final path in Writer::shutdown. However, frag_reuse.rs imported tokio::io::AsyncWriteExt but not lance_io::traits::Writer, so writer.shutdown() resolved to AsyncWriteExt::shutdown (flush/close only) — the temp file was deleted on drop without being persisted. Any subsequent load_indices call would fail with Not found: .../details.binpb.

Fixed by using UFCS Writer::shutdown(writer.as_mut()).await? to explicitly call the lance trait method, matching the existing pattern in ivf.rs and blob.rs.

Fixes #6161

[github-actions]

Review

Clean, well-targeted bugfix. The root cause analysis is spot-on: AsyncWriteExt::shutdown (flush/close) was being resolved instead of lance_io::traits::Writer::shutdown (which persists the temp file via atomic rename on local FS). The UFCS fix Writer::shutdown(writer.as_mut()).await? matches the established pattern used throughout the codebase (ivf.rs, blob.rs, writer.rs, etc.).

The regression test is well-constructed — it creates enough fragments to exceed the 204800-byte inline threshold and verifies the external file round-trips correctly.

No issues found. LGTM.

[codecov]

Codecov Report

❌ Patch coverage is 83.76068% with 19 lines in your changes missing coverage. Please review.

Files with missing lines	Patch %	Lines
rust/lance-io/src/object_writer.rs	73.61%	16 Missing and 3 partials ⚠️

📢 Thoughts on this report? Let us know!

[Xuanwo]

Thank you!

[westonpace]

Yikes, I wonder if there is any way we can rework our code to avoid this? Maybe LocalWriter can have a customer Drop impl which logs a warning if shutdown isn't called? Though maybe there are legitimate reasons to drop a writer before we finish writing 😦

[westonpace]

Or maybe we can put the rename inside of poll_shutdown? That actually seems like a more natural fit.

[wjones127]

Or maybe we can put the rename inside of poll_shutdown? That actually seems like a more natural fit.

@westonpace You are right, we can remove this footgun. I've just pushed a fix that does that.