Improve Ruff Formatter Interoperability

[magic-akari]

Summary

Thank you for developing such an amazing project.

I am currently trying to port the ruff python formatter to wasm. See https://github.com/wasm-fmt/ruff_fmt
I ran into some issues that prevented me from moving forward.

Here are some changes to make it more interoperable for downstream developers:

1. Change the visibility of some structs. This is a straightforward task, adding some missing pub keywords.
2. Export some default config functions.
3. Implement FromStr for some structs, so that users can easily parse configuration from strings.
4. Fix with_quote_style
5. Fix indent_style serde default value

Test Plan

Most of the changes don't involve logic changes, except for a couple of FromStrs, and I'll add test cases if necessary.

[magic-akari]

@MichaReiser
I need to adapt some configuration of dprint.
It looks like ruff's playground formatter is not configurable.

ruff/crates/ruff_wasm/src/lib.rs

Lines 270 to 276 in ac5c8bb

	pub fn format(&self, contents: &str) -> Result<String, Error> {
	// TODO(konstin): Add an options for py/pyi to the UI (1/2)
	let options = PyFormatOptions::from_source_type(PySourceType::default());
	let printed = format_module(contents, options).map_err(into_error)?;
	Ok(printed.into_code())
	}

ruff/crates/ruff_python_formatter/src/options.rs

Lines 54 to 59 in ac5c8bb

	pub fn from_source_type(source_type: PySourceType) -> Self {
	Self {
	source_type,
	..Self::default()
	}
	}

[github-actions]

PR Check Results

Benchmark

Linux

group                                      main                                   pr
-----                                      ----                                   --
formatter/large/dataset.py                 1.00      8.2±0.09ms     4.9 MB/sec    1.00      8.2±0.02ms     5.0 MB/sec
formatter/numpy/ctypeslib.py               1.00  1650.7±37.17µs    10.1 MB/sec    1.03  1699.3±65.30µs     9.8 MB/sec
formatter/numpy/globals.py                 1.00    184.8±4.56µs    16.0 MB/sec    1.03    191.0±8.14µs    15.5 MB/sec
formatter/pydantic/types.py                1.00      3.5±0.08ms     7.3 MB/sec    1.02      3.5±0.08ms     7.2 MB/sec
linter/all-rules/large/dataset.py          1.00     10.2±0.05ms     4.0 MB/sec    1.04     10.6±0.03ms     3.8 MB/sec
linter/all-rules/numpy/ctypeslib.py        1.00      2.7±0.01ms     6.1 MB/sec    1.02      2.8±0.00ms     6.0 MB/sec
linter/all-rules/numpy/globals.py          1.01    389.0±0.80µs     7.6 MB/sec    1.00    386.9±0.37µs     7.6 MB/sec
linter/all-rules/pydantic/types.py         1.00      5.3±0.04ms     4.8 MB/sec    1.04      5.5±0.02ms     4.7 MB/sec
linter/default-rules/large/dataset.py      1.00      5.3±0.02ms     7.7 MB/sec    1.08      5.7±0.01ms     7.1 MB/sec
linter/default-rules/numpy/ctypeslib.py    1.00   1155.4±2.47µs    14.4 MB/sec    1.05   1209.1±9.37µs    13.8 MB/sec
linter/default-rules/numpy/globals.py      1.00    134.0±2.83µs    22.0 MB/sec    1.02    136.1±0.38µs    21.7 MB/sec
linter/default-rules/pydantic/types.py     1.00      2.4±0.07ms    10.4 MB/sec    1.04      2.6±0.03ms    10.0 MB/sec

Windows

group                                      main                                   pr
-----                                      ----                                   --
formatter/large/dataset.py                 1.00     10.0±0.05ms     4.1 MB/sec    1.00      9.9±0.06ms     4.1 MB/sec
formatter/numpy/ctypeslib.py               1.00  1876.0±10.87µs     8.9 MB/sec    1.01  1903.5±18.15µs     8.7 MB/sec
formatter/numpy/globals.py                 1.00    190.2±1.33µs    15.5 MB/sec    1.04    198.8±5.75µs    14.8 MB/sec
formatter/pydantic/types.py                1.00      4.2±0.02ms     6.1 MB/sec    1.02      4.3±0.04ms     6.0 MB/sec
linter/all-rules/large/dataset.py          1.01     12.5±0.04ms     3.2 MB/sec    1.00     12.5±0.05ms     3.3 MB/sec
linter/all-rules/numpy/ctypeslib.py        1.00      3.5±0.01ms     4.7 MB/sec    1.00      3.5±0.02ms     4.7 MB/sec
linter/all-rules/numpy/globals.py          1.00    361.7±3.62µs     8.2 MB/sec    1.01    365.5±7.61µs     8.1 MB/sec
linter/all-rules/pydantic/types.py         1.00      6.5±0.03ms     3.9 MB/sec    1.00      6.5±0.04ms     3.9 MB/sec
linter/default-rules/large/dataset.py      1.00      6.8±0.02ms     6.0 MB/sec    1.01      6.8±0.05ms     6.0 MB/sec
linter/default-rules/numpy/ctypeslib.py    1.00   1402.6±6.12µs    11.9 MB/sec    1.00   1407.9±7.61µs    11.8 MB/sec
linter/default-rules/numpy/globals.py      1.00    143.6±4.34µs    20.5 MB/sec    1.00    143.4±1.14µs    20.6 MB/sec
linter/default-rules/pydantic/types.py     1.00      3.1±0.04ms     8.3 MB/sec    1.00      3.0±0.02ms     8.4 MB/sec

[magic-akari]

There are multiple commits here,
I think it's fine use squash merge or I squash locally and then force push to the branch.

[MichaReiser]

These changes look good to me. Thanks for contributing them upstream.

I will be interested to see where it goes with your dprint implementation. Just a reminder, the formatter API isn't stable, and we may change it in minor releases without notice.

[magic-akari]

You can try it:

1. instal dprint: https://dprint.dev/install/
2. config the plugin

dprint config add wasm-fmt/ruff_fmt

---

I am using ruff formatter as git dependency.
I was wondering if there was a roadmap. I'm waiting for it to be published on crates.

[MichaReiser]

I was wondering if there was a roadmap. I'm waiting for it to be published on crates.

We don't have any near-term plans on publishing our crates to crates.io because people will (not unjustified) assume that the APIs are stable and we aren't ready to give that guarantee yet.