Run cargo fmt on the whole code base by bkchr · Pull Request #9394 · paritytech/substrate

bkchr · 2021-07-20T08:52:50Z

No description provided.

athei · 2021-07-20T09:09:15Z

IMHO we should add a cargo fmt -- --check CI job in this PR. Because otherwise differently formatted code will slip in and make running cargo fmt a pain, again.

gilescope

I've looked over 1/4 of these changes and am happy with them.

athei · 2021-07-20T09:51:05Z

One thing I noticed is that the auto generated weight files are reformatted, too. This will break the CI check whenever those are updated (by the bot).

I think we should either run rustfmt automatically on the generated weight files (not sure how) or skip those files in cargo fmt. Not sure how to do this either. Maybe we can add some #[rustfmt(skip)] to the weight template?

expenses

This might cause some issues for PR authors, but it needs to be done anyway

andresilva · 2021-07-20T11:39:22Z

paritytech/polkadot@2a49df1

@bkchr Do you mind adding the same changes to rustfmt.toml here? It's explained here.

Maybe we can add some #[rustfmt(skip)] to the weight template?

This is probably the easiest and most sensible thing to do.

andresilva

I went through the first 60 files. Overall I think rustfmt does a very good job and there's not much to fix but I found some issues with weird formatting around: multi-line strings and really long expression lines (e.g. matching on the result of a function call). For multi-line strings maybe the format_strings option should help, for other cases introducing some temporary variables helps with making the code more readable. If we could somehow coordinate people to actually look at the diffs then we could probably catch all regressions, we only need 10 people to look at 100 files each :)

andresilva · 2021-07-20T11:47:48Z

bin/node-template/node/src/command.rs

+				You can enable it with `--features runtime-benchmarks`."
+					.into())


This looks weird. Wouldn't format_string = true help?

Yes format_strings would help but it messes up things like client/cli/src/commands/utils.rs:

println!( "Network ID/version: {}\n \ Public key (hex): {}\n \ Public key (SS58): {}\n \ Account ID: {}\n \ SS58 Address: {}",

So our choice is to manually indent multiline strings (keeping control) or to add rustfmt::skips where we need more control. On the whole I think doing multi-line strings manually means less surprises.

There's about 5 places in the codebase where it messes up (mostly snippits of json). Personally I'd not format strings but given there's only 5 places where it causes harm 🤷 ...

andresilva · 2021-07-20T11:48:30Z

bin/node-template/pallets/template/src/benchmarking.rs

-	crate::mock::new_test_ext(),
-	crate::mock::Test,
-);
+impl_benchmark_test_suite!(Template, crate::mock::new_test_ext(), crate::mock::Test,);


Suggested change

impl_benchmark_test_suite!(Template, crate::mock::new_test_ext(), crate::mock::Test,);

impl_benchmark_test_suite!(Template, crate::mock::new_test_ext(), crate::mock::Test);

I suppose this will happen in a bunch of other places. Maybe should be reported as an issue to rustfmt?

rust-lang/rustfmt#2918

ok so we can in general remove those commas as we find them and it won't put them back.
Maybe we have a separate follow-up PR that nurfs those extra commas at the end of macros?

andresilva · 2021-07-20T12:01:30Z

bin/node/bench/src/main.rs

+		]
+		.iter()
+		{
 			for block_type in [
 				BlockType::RandomTransfersKeepAlive,
 				BlockType::RandomTransfersReaping,
 				BlockType::Noop,
-			].iter() {
-				for database_type in [BenchDataBaseType::RocksDb, BenchDataBaseType::ParityDb].iter() {
-					import_benchmarks.push((profile, size.clone(), block_type.clone(), database_type));
+			]
+			.iter()
+			{
+				for database_type in
+					[BenchDataBaseType::RocksDb, BenchDataBaseType::ParityDb].iter()
+				{
+					import_benchmarks.push((
+						profile,
+						size.clone(),
+						block_type.clone(),
+						database_type,
+					));


Maybe introducing some variables for the arrays would make this code less ugly.

Ideally we want to keep this PR as small as possible in terms of manual code changes. Beautification should be a followup PR.

andresilva · 2021-07-20T12:03:25Z

bin/node/cli/src/command.rs

 				Err("Benchmarking wasn't enabled when building the node. \
-				You can enable it with `--features runtime-benchmarks`.".into())
-			}
-		}
+				You can enable it with `--features runtime-benchmarks`."
+					.into())


Same as the earlier issue with formatting multi-line strings.

andresilva · 2021-07-20T12:03:34Z

bin/node/cli/src/command.rs

+		Some(Subcommand::TryRuntime) => Err("TryRuntime wasn't enabled when building the node. \
+				You can enable it with `--features try-runtime`."
+			.into()),


It's leaving untouched the leading spaces. If we manually fix the indenting of these lines then rustfmt will retain them.

andresilva · 2021-07-20T12:05:01Z

bin/node/executor/benches/bench.rs

+	node_runtime::WASM_BINARY.expect(
+		"Development wasm binary is not available. \
+									  Testing is only supported with the flag disabled.",
+	)


andresilva · 2021-07-20T12:05:20Z

bin/node/executor/tests/basic.rs

+		"Development wasm binary is not available. \
+											 Testing is only supported with the flag disabled.",


andresilva · 2021-07-20T12:06:20Z

bin/node/executor/tests/common.rs

 	let header = match executor_call::<NeverNativeValue, fn() -> _>(
 		env,
 		"BlockBuilder_finalize_block",
-		&[0u8;0],
+		&[0u8; 0],
 		true,
 		None,
-	).0.unwrap() {
+	)
+	.0
+	.unwrap()
+	{


Creating a temp var for the result of executor_call would probably make this less weird.

andresilva · 2021-07-20T12:08:09Z

bin/node/runtime/src/lib.rs

+		"Development wasm binary is not available. This means the client is \
 						built with `SKIP_WASM_BUILD` flag and it is only usable for \
-						production chains. Please rebuild with the flag disabled.")
+						production chains. Please rebuild with the flag disabled.",
+	)


TriplEight · 2021-07-20T17:41:53Z

.gitlab-ci.yml

  allow_failure:                   true

+cargo-fmt:
+  stage:                           test


as a debug step, just to avoid running everything else, move it to the previous stage

Suggested change

stage: test

stage: .pre

just do not forget to revert this at the end :)

Ty, but that works ;)

gilescope · 2021-07-21T12:00:04Z

test-utils/client/src/lib.rs

+			mem,
+			rx,
+		)
+		.unwrap_err();


Again here if we pull these strings out into vars they will format fine.

andresilva

I agree that it's best not to do any manual changes in this PR otherwise it will make it harder to deal with conflicts on existing PRs.

.

bkchr · 2021-07-21T14:02:22Z

bot merge

ghost · 2021-07-21T14:02:27Z

Waiting for commit status.

Run cargo fmt on the whole code base

19bd756

bkchr requested review from andresilva, athei, kianenigma, pepyakin, shawntabrizi, sorpaas and tomusdrw as code owners July 20, 2021 08:52

bkchr added 2 commits July 20, 2021 11:17

Second run

1e5c254

Add CI check

6dbe33d

bkchr requested a review from a team as a code owner July 20, 2021 09:19

gilescope approved these changes Jul 20, 2021

View reviewed changes

expenses approved these changes Jul 20, 2021

View reviewed changes

Fix compilation

cbd3e84

More unnecessary braces

66dd0a5

andresilva suggested changes Jul 20, 2021

View reviewed changes

bkchr added 6 commits July 20, 2021 15:15

Handle weights

759fa7b

Use --all

50395db

Use correct attributes...

e87733a

Fix UI tests

af67ab7

AHHHHHHHHH

f5b1b69

🤦

b947586

TriplEight previously requested changes Jul 20, 2021

View reviewed changes

bkchr added 7 commits July 20, 2021 22:04

Merge remote-tracking branch 'origin/master' into bkchr-cargo-fmt

332b916

Docs

cd0cfbb

Fix compilation

4d51706

🤷

1d97cd2

Please stop

6d80a73

🤦 x 2

168541b

More

a5686da

gilescope reviewed Jul 21, 2021

View reviewed changes

make rustfmt.toml consistent with polkadot

cbe6d1a

andresilva approved these changes Jul 21, 2021

View reviewed changes

shawntabrizi approved these changes Jul 21, 2021

View reviewed changes

ghost merged commit 1d5abf0 into master Jul 21, 2021

ghost deleted the bkchr-cargo-fmt branch July 21, 2021 14:32

emostov mentioned this pull request Jul 21, 2021

Run cargo +nightly fmt #9406

Merged

Satoshi-Kusumoto mentioned this pull request Jul 29, 2021

Feature/rust fmt litentry/heima#62

Merged

ordian mentioned this pull request Aug 1, 2021

cargo +nightly fmt paritytech/polkadot#3540

Merged

This pull request was closed.

		You can enable it with `--features runtime-benchmarks`."
		.into())

	impl_benchmark_test_suite!(Template, crate::mock::new_test_ext(), crate::mock::Test,);
	impl_benchmark_test_suite!(Template, crate::mock::new_test_ext(), crate::mock::Test);

		"Development wasm binary is not available. \
		Testing is only supported with the flag disabled.",

Conversation

bkchr commented Jul 20, 2021

Uh oh!

athei commented Jul 20, 2021

Uh oh!

gilescope left a comment

Choose a reason for hiding this comment

Uh oh!

athei commented Jul 20, 2021 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

expenses left a comment

Choose a reason for hiding this comment

Uh oh!

andresilva commented Jul 20, 2021

Uh oh!

andresilva left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

andresilva Jul 20, 2021 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

andresilva left a comment

Choose a reason for hiding this comment

Uh oh!

bkchr commented Jul 21, 2021

Uh oh!

ghost commented Jul 21, 2021

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

7 participants

athei commented Jul 20, 2021 •

edited

Loading

andresilva Jul 20, 2021 •

edited

Loading