Add cargo-udeps in CI

[JonathanPlasse]

• Close CI: Add CI step to check for unused dependencies #3643

[github-actions]

PR Check Results

Ecosystem

✅ ecosystem check detected no changes.

Benchmark

Linux

group                                      main                                    pr
-----                                      ----                                    --
linter/all-rules/large/dataset.py          1.03     16.8±0.08ms     2.4 MB/sec     1.00     16.4±0.08ms     2.5 MB/sec
linter/all-rules/numpy/ctypeslib.py        1.01      4.3±0.02ms     3.9 MB/sec     1.00      4.3±0.01ms     3.9 MB/sec
linter/all-rules/numpy/globals.py          1.00    587.9±1.26µs     5.0 MB/sec     1.01    590.9±2.12µs     5.0 MB/sec
linter/all-rules/pydantic/types.py         1.02      7.4±0.05ms     3.5 MB/sec     1.00      7.2±0.06ms     3.5 MB/sec
linter/default-rules/large/dataset.py      1.01      8.8±0.03ms     4.6 MB/sec     1.00      8.7±0.02ms     4.7 MB/sec
linter/default-rules/numpy/ctypeslib.py    1.01  1974.1±117.34µs     8.4 MB/sec    1.00   1954.8±2.42µs     8.5 MB/sec
linter/default-rules/numpy/globals.py      1.00    213.6±0.52µs    13.8 MB/sec     1.03    219.4±0.68µs    13.5 MB/sec
linter/default-rules/pydantic/types.py     1.00      4.1±0.01ms     6.2 MB/sec     1.00      4.1±0.01ms     6.2 MB/sec

Windows

group                                      main                                   pr
-----                                      ----                                   --
linter/all-rules/large/dataset.py          1.06     17.9±0.88ms     2.3 MB/sec    1.00     16.9±0.56ms     2.4 MB/sec
linter/all-rules/numpy/ctypeslib.py        1.01      4.7±0.28ms     3.5 MB/sec    1.00      4.7±0.22ms     3.5 MB/sec
linter/all-rules/numpy/globals.py          1.05   644.9±38.36µs     4.6 MB/sec    1.00   612.5±37.68µs     4.8 MB/sec
linter/all-rules/pydantic/types.py         1.04      7.9±0.34ms     3.2 MB/sec    1.00      7.6±0.38ms     3.4 MB/sec
linter/default-rules/large/dataset.py      1.02      9.8±0.46ms     4.1 MB/sec    1.00      9.6±0.48ms     4.2 MB/sec
linter/default-rules/numpy/ctypeslib.py    1.03      2.1±0.09ms     7.8 MB/sec    1.00      2.1±0.11ms     8.0 MB/sec
linter/default-rules/numpy/globals.py      1.02   243.6±19.81µs    12.1 MB/sec    1.00   238.3±18.62µs    12.4 MB/sec
linter/default-rules/pydantic/types.py     1.01      4.5±0.17ms     5.7 MB/sec    1.00      4.4±0.23ms     5.7 MB/sec

[MichaReiser]

Oh wow... I didn't expect this to be picked up and be implemented so quickly! Thank you.

Can you add a test plan to your PR summary that shows that the job fails if there's an unused dependency (you can push a change to this PR and then add a link to the action-run to PR summary).

[JonathanPlasse]

Successful cargo-udeps CI

No unused dependencies found

Failed cargo-udeps CI

Unused dependencies found

`ruff_diagnostics v0.0.0 (/home/runner/work/ruff/ruff/crates/ruff_diagnostics)`
└─── dependencies
     └─── "clap"
`ruff_formatter v0.0.0 (/home/runner/work/ruff/ruff/crates/ruff_formatter)`
└─── dev-dependencies
     └─── "insta"
`ruff_python_formatter v0.0.0 (/home/runner/work/ruff/ruff/crates/ruff_python_formatter)`
└─── dev-dependencies
     └─── "test-case"
`ruff_wasm v0.0.0 (/home/runner/work/ruff/ruff/crates/ruff_wasm)`
└─── dev-dependencies
     └─── "wasm-bindgen-test"
Note: They might be false-positive.
      For example, `cargo-udeps` cannot detect usage of crates that are only used in doc-tests.
      To ignore some dependencies, write `package.metadata.cargo-udeps.ignore` in Cargo.toml.

[MichaReiser]

Nice polished summary! I'm unsure about adding the PR result to the comment action because of the added complexity, and that I don't think it is necessary. My inline comment goes into details why. What was your motivation when you added it?

[JonathanPlasse]

Nice polished summary! I'm unsure about adding the PR result to the comment action because of the added complexity, and that I don't think it is necessary. My inline comment goes into details why. What was your motivation when you added it?

It was to make the cargo-udeps summary more visible.
But, as you said, it would cause noise and it should not happen very often.