chore(deps): bump fgumi to 0.5.0

[nh13]

Bumps the sort engine from fgumi 0.3.1 to 0.5.0, released today. One-line dependency bump plus one test update for an upstream behavior change; no original sort code added here.

What mako picks up

Sort-relevant highlights from the 0.4.0 + 0.5.0 changelogs:

• Single-pass --verify (#568, #594) — verify now streams its input once (header parsed through a tee, consumed bytes replayed) instead of re-opening the path, and bounds run comparison to O(order divergence).
• Performance — records stream into the sort instead of via an intermediate BAM (#593), narrower radix passes with the bound derived inside the first counting pass (#606, #622), and a pool-integrated coordinate --write-index merge (#647).
• New flags, which surface automatically through the flattened Sort — --sort-threads / --merge-threads for per-phase thread control (#608), --max-temp-files to tune spill-file consolidation (#643), --index-threshold always|never (#632).
• Input handling — uncompressed SAM and stdin accepted anywhere BAM is (#644).
• Correctness fixes — BAI sidecar written to the samtools path rather than an extension-replaced one (#607), BAM bin field recomputed after raw POS/CIGAR mutation (#591), placed-but-unmapped reads position-binned in the BAI (#651), @HD SS written as <sort-order>:<sub-sort> with the spec spelling lexicographical (#514, #567), and secondary/supplementary reads placed at the exact template coordinate via tc (#529).

0.5.0 carries several [breaking] entries, but all of them are in commands mako does not expose (group, dedup, consensus, codec, compare). MSRV is unchanged at 1.93.0, matching rust-toolchain.toml, so no toolchain bump.

Test change

Test 10 asserted the old behavior: that --verify rejected a non-seekable stdin stream up front with a message mentioning stdin, because verify needed to re-read its input. Single-pass verify removes that limitation, so the assertion is now testing behavior that no longer exists — it failed with BAM file is NOT correctly sorted by Coordinate: 1 violations found, i.e. verify had happily consumed the pipe and done its job.

verify_rejects_stdin_input is replaced with verify_accepts_stdin_input, which locks in the new behavior: verify from a piped stdin exits 0 on a sorted stream and non-zero on an unsorted one. Both directions are asserted deliberately — together they prove the stream was actually consumed and its order evaluated, rather than verify short-circuiting on an empty read (a one-sided pass-only assertion would be satisfied by a verify that read nothing). Both - and /dev/stdin are still exercised, since the latter is a real existing path.

Verification

cargo ci-fmt   # clean
cargo ci-lint  # clean (clippy --all-targets -D warnings)
cargo ci-test  # 11 passed; 0 failed

Version banner confirms the resolved dependency:

$ mako --version
mako 0.1.3 (rev fa11a2d)
powered by fgumi 0.5.0

Follow-up (not fixed here)

fgumi's --sort-threads help text contains a bwa mem -t 32 ... | fgumi sort -@ 8 --sort-threads 4 example, so mako --help now shows one fgumi sort invocation. Cosmetic, and it belongs upstream in fgumi's arg doc comment rather than in this repo.

Summary by CodeRabbit

• New Features
  • mako --verify now supports validating BAM data streamed through standard input.
  • Sorted piped input is accepted, while unsorted input still reports a verification failure.

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Organization UI

Review profile: CHILL

Plan: Pro Plus

Run ID: 907298af-6d6a-4c4a-b788-87038d1e416e

📥 Commits

Reviewing files that changed from the base of the PR and between fa11a2d and 6ab85aa.

⛔ Files ignored due to path filters (1)

• Cargo.lock is excluded by !**/*.lock

📒 Files selected for processing (2)

• Cargo.toml
• tests/cli.rs

---

📝 Walkthrough

Walkthrough

The fgumi dependency is upgraded from 0.3.1 to 0.5.0. CLI tests now verify sorted piped stdin succeeds and unsorted piped stdin fails for both - and /dev/stdin.

Changes

Stdin verification

Layer / File(s)	Summary
Verify piped stdin behavior Cargo.toml, tests/cli.rs	The fgumi dependency is upgraded, and piped stdin verification tests now cover sorted and unsorted BAM streams through both stdin forms.

Estimated code review effort: 2 (Simple) | ~10 minutes

Possibly related PRs

• fg-labs/mako#10: Also updates the fgumi dependency version in Cargo.toml.

Poem

A rabbit pipes BAM through the burrow so bright,
Sorted streams pass verification right.
Unsorted ones hop to a non-zero fate,
With - and /dev/stdin both at the gate.
Thump, thump—dependencies update!

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title clearly matches the main change: bumping fgumi from 0.3.1 to 0.5.0.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.

✨ Finishing Touches

📝 Generate docstrings

• Create stacked PR
• Commit on current branch

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}