Replace `assert!(a==b)` with `assert_eq!(a,b)` as part of `bool_assert_comparison` lint #13333

nyurik · 2024-09-02T02:09:39Z

Expand lint bool_assert_comparison to catch assert!(a == b) and assert!(a != b) (or their debug versions), and suggest to replace them with the corresponding assert_eq!(a, b) or assert_ne!(a, b).

TODO:

Decide if this lint should expand bool_assert_comparison or be a new lint (name?). Note that existing lint description fits well to the new functionality.
Are we ok to keep this as style? Note that IntelliJ has a similar lint built-in, and it shows up similar to styling guide.

Closes #13252

changelog: [bool_assert_comparison]: now also converts assert!(a==b) into assert_eq!(a,b) and similar

rustbot · 2024-09-02T02:09:44Z

r? @xFrednet

rustbot has assigned @xFrednet.
They will have a look at your PR within the next two weeks and either review your PR or reassign to another reviewer.

Use r? to explicitly pick a reviewer

workingjubilee · 2024-09-02T04:15:47Z

does this lint on assert!(x == y) in const contexts, where they are not replaceable with assert_eq! in today's Rust?

workingjubilee · 2024-09-02T04:32:45Z

tests/ui/assertions_on_constants.rs

@@ -1,4 +1,4 @@
-#![allow(non_fmt_panics, clippy::needless_bool, clippy::eq_op)]
+#![allow(non_fmt_panics, clippy::needless_bool, clippy::eq_op, clippy::bool_assert_comparison)]


oh, it does!

given the sheer amount of pushback you gave me when you had to reorganize your code for linting reasons, I find it frankly astonishing that you would introduce a lint that forces every user of const fn to do the exact same thing you were complaining about having to do: rearchitect their code purely for linting reasons if they want this lint to remain useful for their code that doesn't use const assertions.

lol, worry not, this lint needs feedback. So i guess you point is that i should use is_in_const_context check in this lint so that it doesn't get triggered for const context.

Yes, that seems like it would be sufficient, I think?

workingjubilee · 2024-09-02T04:38:42Z

rustc has reliably gotten the user feedback of "these two things should not be the same lint" for basically every single lint expansion into a new category of code patterns. the fact that this new lint on some instances of assert! would have to be allowed every time someone wrote a const fn makes treating this as the same thing as assert_eq!(expr, false) especially baffling.

nyurik · 2024-09-02T04:56:29Z

any suggestions for the new lint name? And yes, if this (or any other) lint shows when its not applicable, its a false positive and should be fixed

workingjubilee · 2024-09-02T05:34:47Z

why not just assert_on_eq?

nyurik · 2024-09-02T05:48:58Z

Perhaps. Usecase could be allow assert_on_eq to mean that assert!(foo==bar) is ok...

smoelius · 2024-09-02T10:46:15Z

While I very much like this idea, Clippy had a lint for this before. See #2156, especially the issues/PRs referenced at the bottom.

nyurik · 2024-09-04T14:14:23Z

Reading the other discussion, it seems there was hope that assert!(foo==bar) would somehow document the values of foo and bar, and later these efforts did not get too far (if i am reading it correctly). If so, we should merge related PRs into one lint and proceed with assert! -> assert_eq! suggestions. If on the other hand there is a way to improve assert! machinery in some foreseeable future, let's postpone this work.

xFrednet · 2024-09-09T20:30:35Z

Decide if this lint should expand bool_assert_comparison or be a new lint (name?). Note that existing lint description fits well to the new functionality.

That's a good question. I've created https://rust-lang.zulipchat.com/#narrow/stream/257328-clippy/topic/New.20lint.20or.20expand.20.60bool_assert_comparison.60.20rust-clippy.231333/near/468909875

xFrednet

Excellent start, I left comments, mostly related to your TODO comments :D

xFrednet · 2024-09-09T20:52:03Z

clippy_lints/src/bool_assert_comparison.rs

+    // TODO: is `cond.span.from_expansion()` correct / needed?
+    if (node != BinOpKind::Eq && node != BinOpKind::Ne) || cond.span.from_expansion() {


// TODO: is cond.span.from_expansion() correct / needed?

It should be correct 👍

xFrednet · 2024-09-09T20:53:21Z

clippy_lints/src/bool_assert_comparison.rs

+
+    let new_name = format!("{macro_name}_{}", if node == BinOpKind::Eq { "eq" } else { "ne" });
+    let msg = format!("replace it with `{new_name}!(..)`");
+    span_lint_and_then(


is_in_const_context should be checked some time before here, and it would be good to have tests for it.

xFrednet · 2024-09-09T20:58:59Z

clippy_lints/src/bool_assert_comparison.rs

+        |diag| {
+            let macro_name_span = cx.sess().source_map().span_until_char(macro_call.span, '!');
+            // TODO: should this be `cond.span`, expanded to include preceding whitespace? If so, how?
+            let equality_span = Span::new(lhs.span.hi(), rhs.span.lo(), span.ctxt(), span.parent());


We know that the == actually comes from the source code, so I would probably use cond.span.with_lo(lhs.span.hi()).with_high(rhs.span.lo()). But this should result in the same span, so either way should be fine :)

bors · 2024-12-08T10:54:33Z

☔ The latest upstream changes (presumably ec37eb6) made this pull request unmergeable. Please resolve the merge conflicts.

Expand lint `bool_assert_comparison` to catch `assert!(a == b)` and `assert!(a != b)` (or their debug versions), and suggest to replace them with the corresponding `assert_eq!(a, b)`.

xFrednet · 2025-03-16T14:44:46Z

I'm unassigning myself from this PR. @nyurik if you want to continue this work, please request a new reviewer with r? clippy

rustbot · 2025-03-31T23:42:09Z

☔ The latest upstream changes (possibly d28d234) made this pull request unmergeable. Please resolve the merge conflicts.

Jarcho · 2025-09-17T00:07:00Z

Ping @nyurik from triage. Do you plan to return to working on this? Since the PR is marked as a draft I assume it's incomplete.

Jarcho · 2025-10-01T20:21:49Z

Closing since there's been no response. @nyurik if you wish to return to working on this feel free to reopen or create a new PR.

rustbot assigned xFrednet Sep 2, 2024

rustbot added the S-waiting-on-review Status: Awaiting review from the assignee but also interested parties label Sep 2, 2024

nyurik force-pushed the asserts-to-eq branch from 1cc3977 to 9fd422f Compare September 2, 2024 02:28

workingjubilee reviewed Sep 2, 2024

View reviewed changes

xFrednet reviewed Sep 9, 2024

View reviewed changes

nyurik force-pushed the asserts-to-eq branch 2 times, most recently from 8f4d34e to 42eb926 Compare September 22, 2024 22:59

nyurik marked this pull request as draft September 22, 2024 23:02

nyurik added 2 commits January 31, 2025 01:12

Replace assert!(a==b) with assert_eq!

60abd3b

Expand lint `bool_assert_comparison` to catch `assert!(a == b)` and `assert!(a != b)` (or their debug versions), and suggest to replace them with the corresponding `assert_eq!(a, b)`.

wip

e3627c2

nyurik force-pushed the asserts-to-eq branch from 42eb926 to e3627c2 Compare January 31, 2025 06:14

xFrednet removed their assignment Mar 16, 2025

rustbot added S-waiting-on-author Status: This is awaiting some action from the author. (Use `@rustbot ready` to update this status) and removed S-waiting-on-review Status: Awaiting review from the assignee but also interested parties labels Mar 31, 2025

nyurik mentioned this pull request Apr 8, 2025

chore: use assert_eq! and assert_ne! instead of assert! uutils/coreutils#7698

Merged

Jarcho closed this Oct 1, 2025

		@@ -1,4 +1,4 @@
		#![allow(non_fmt_panics, clippy::needless_bool, clippy::eq_op)]
		#![allow(non_fmt_panics, clippy::needless_bool, clippy::eq_op, clippy::bool_assert_comparison)]

		// TODO: is `cond.span.from_expansion()` correct / needed?
		if (node != BinOpKind::Eq && node != BinOpKind::Ne) \|\| cond.span.from_expansion() {

Replace assert!(a==b) with assert_eq!(a,b) as part of bool_assert_comparison lint #13333

Replace assert!(a==b) with assert_eq!(a,b) as part of bool_assert_comparison lint #13333

Uh oh!

Conversation

nyurik commented Sep 2, 2024 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

TODO:

Uh oh!

rustbot commented Sep 2, 2024

Uh oh!

workingjubilee commented Sep 2, 2024

Uh oh!

workingjubilee Sep 2, 2024

Choose a reason for hiding this comment

Uh oh!

nyurik Sep 2, 2024

Choose a reason for hiding this comment

Uh oh!

workingjubilee Sep 3, 2024

Choose a reason for hiding this comment

Uh oh!

workingjubilee commented Sep 2, 2024

Uh oh!

nyurik commented Sep 2, 2024

Uh oh!

workingjubilee commented Sep 2, 2024

Uh oh!

nyurik commented Sep 2, 2024

Uh oh!

smoelius commented Sep 2, 2024

Uh oh!

nyurik commented Sep 4, 2024

Uh oh!

xFrednet commented Sep 9, 2024

Uh oh!

xFrednet left a comment

Choose a reason for hiding this comment

Uh oh!

xFrednet Sep 9, 2024

Choose a reason for hiding this comment

Uh oh!

xFrednet Sep 9, 2024

Choose a reason for hiding this comment

Uh oh!

xFrednet Sep 9, 2024

Choose a reason for hiding this comment

Uh oh!

bors commented Dec 8, 2024

Uh oh!

xFrednet commented Mar 16, 2025

Uh oh!

rustbot commented Mar 31, 2025

Uh oh!

Jarcho commented Sep 17, 2025

Uh oh!

Jarcho commented Oct 1, 2025

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

7 participants

Replace `assert!(a==b)` with `assert_eq!(a,b)` as part of `bool_assert_comparison` lint #13333

Replace `assert!(a==b)` with `assert_eq!(a,b)` as part of `bool_assert_comparison` lint #13333

nyurik commented Sep 2, 2024 •

edited

Loading