docs(storage): correct sidecar WAL removal timing

[teipsum]

Problem

The single-file storage docs claimed the sidecar WAL directory is "removed after each checkpoint". It is not: periodic checkpoints fold in-flight mutations into the .grafeo file but retain the sidecar; remove_sidecar_wal is invoked only on a clean close(). Retaining the sidecar across checkpoints is exactly what lets an unexpected restart recover in-flight mutations — only an orderly close discards it.

Fix

Correct three docstrings to match actual behaviour: the file/mod.rs module doc, the GrafeoFileManager lifecycle steps, and the remove_sidecar_wal doc comment (all under crates/grafeo-storage/src/file/). Docs-only; no functional change.

Note: flush.rs already documents the periodic-checkpoint WAL correctly as "Kept" on current main, so it is deliberately left untouched — only the remaining stale docstrings are corrected.

Tests

cargo fmt --all -- --check                                 -> clean
cargo test -p grafeo-storage --all-features file::         -> 54 passed; 0 failed
RUSTDOCFLAGS="-D rustdoc::broken_intra_doc_links" \
  cargo doc -p grafeo-storage --no-deps --all-features     -> no broken links

Single commit, based on main @ 4ebae02.

---

Summary by cubic

Fix storage docs to reflect the correct sidecar WAL lifecycle: periodic checkpoints keep the sidecar; only a clean close() calls remove_sidecar_wal. Updates the GrafeoFileManager lifecycle comments and the module docs in crates/grafeo-storage/src/file/; no behavior change.

^{Written for commit 18e59df. Summary will update on new commits.}

[cubic-dev-ai]

No issues found across 2 files

_{Re-trigger cubic}

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.

📢 Thoughts on this report? Let us know!

[codspeed-hq]

Merging this PR will not alter performance

✅ 70 untouched benchmarks

---

_{Comparing teipsum:u3b-storage-docstring-fix (18e59df) with main (4ebae02)}