Merge branch 'main' into PYI050

.github/workflows/release.yaml

@@ -34,6 +34,7 @@ jobs:
           args: --out dist
       - name: "Test sdist"
         run: |
+          rustup default $(cat rust-toolchain)
           pip install dist/${{ env.PACKAGE_NAME }}-*.tar.gz --force-reinstall
           ruff --help
           python -m ruff --help

.pre-commit-config.yaml

@@ -1,4 +1,12 @@
 fail_fast: true
+
+exclude: |
+  (?x)^(
+    crates/ruff/resources/.*|
+    crates/ruff_python_formatter/resources/.*|
+    crates/ruff_python_formatter/src/snapshots/.*
+  )$
+
 repos:
   - repo: https://github.com/abravalheri/validate-pyproject
     rev: v0.12.1
@@ -19,7 +27,7 @@ repos:
       - id: markdownlint-fix
 
   - repo: https://github.com/crate-ci/typos
-    rev: v1.14.8
+    rev: v1.14.12
     hooks:
       - id: typos
 

Cargo.lock

@@ -691,7 +691,7 @@ dependencies = [
 
 [[package]]
 name = "flake8-to-ruff"
-version = "0.0.270"
+version = "0.0.271"
 dependencies = [
  "anyhow",
  "clap",
@@ -1725,7 +1725,7 @@ dependencies = [
 
 [[package]]
 name = "ruff"
-version = "0.0.270"
+version = "0.0.271"
 dependencies = [
  "annotate-snippets 0.9.1",
  "anyhow",
@@ -1818,7 +1818,7 @@ dependencies = [
 
 [[package]]
 name = "ruff_cli"
-version = "0.0.270"
+version = "0.0.271"
 dependencies = [
  "annotate-snippets 0.9.1",
  "anyhow",

README.md

@@ -139,7 +139,7 @@ Ruff can also be used as a [pre-commit](https://pre-commit.com) hook:
 ```yaml
 - repo: https://github.com/astral-sh/ruff-pre-commit
   # Ruff version.
-  rev: v0.0.270
+  rev: v0.0.271
   hooks:
     - id: ruff
 ```

_typos.toml

@@ -1,5 +1,5 @@
 [files]
-extend-exclude = ["snapshots", "black"]
+extend-exclude = ["resources", "snapshots"]
 
 [default.extend-words]
 trivias = "trivias"

crates/flake8_to_ruff/Cargo.toml

@@ -1,6 +1,6 @@
 [package]
 name = "flake8-to-ruff"
-version = "0.0.270"
+version = "0.0.271"
 edition = { workspace = true }
 rust-version = { workspace = true }
 

crates/ruff/Cargo.toml

@@ -1,6 +1,6 @@
 [package]
 name = "ruff"
-version = "0.0.270"
+version = "0.0.271"
 authors.workspace = true
 edition.workspace = true
 rust-version.workspace = true

crates/ruff/resources/test/fixtures/pyflakes/F522.py

@@ -2,3 +2,6 @@
 "{bar}{}".format(1, bar=2, spam=3)  # F522
 "{bar:{spam}}".format(bar=2, spam=3)  # No issues
 "{bar:{spam}}".format(bar=2, spam=3, eggs=4, ham=5)  # F522
+# Not fixable
+(''
+ .format(x=2))

crates/ruff/resources/test/fixtures/pyflakes/F841_3.py

@@ -83,6 +83,11 @@ def f():
         pass
 
 
+def f():
+    with (Nested(m)) as (cm):
+        pass
+
+
 def f():
     toplevel = tt = lexer.get_token()
     if not tt:

crates/ruff/resources/test/fixtures/pylint/iteration_over_set.py

@@ -3,12 +3,6 @@
 for item in {"apples", "lemons", "water"}:  # flags in-line set literals
     print(f"I like {item}.")
 
-for item in set(("apples", "lemons", "water")):  # flags set() calls
-    print(f"I like {item}.")
-
-for number in {i for i in range(10)}:  # flags set comprehensions
-    print(number)
-
 numbers_list = [i for i in {1, 2, 3}]  # flags sets in list comprehensions
 
 numbers_set = {i for i in {1, 2, 3}}  # flags sets in set comprehensions
@@ -36,3 +30,9 @@ numbers_set = {i for i in (1, 2, 3)}  # tuples in comprehensions are fine
 numbers_dict = {str(i): i for i in [1, 2, 3]}  # lists in dict comprehensions are fine
 
 numbers_gen = (i for i in (1, 2, 3))  # tuples in generator expressions are fine
+
+for item in set(("apples", "lemons", "water")):  # set constructor is fine
+    print(f"I like {item}.")
+
+for number in {i for i in range(10)}:  # set comprehensions are fine
+    print(number)

crates/ruff/resources/test/fixtures/pylint/nonlocal_without_binding.py

@@ -17,3 +17,30 @@ def f():
 
     def f():
         nonlocal y
+
+
+def f():
+    x = 1
+
+    def g():
+        nonlocal x
+
+    del x
+
+
+def f():
+    def g():
+        nonlocal x
+
+    del x
+
+
+def f():
+    try:
+        pass
+    except Exception as x:
+        pass
+
+    def g():
+        nonlocal x
+        x = 2

crates/ruff/src/checkers/ast/mod.rs

@@ -307,9 +307,18 @@ where
                                 stmt.range(),
                                 ExecutionContext::Runtime,
                             );
-                        } else {
+                        }
+
                         // Ensure that every nonlocal has an existing binding from a parent scope.
                         if self.enabled(Rule::NonlocalWithoutBinding) {
+                            if self
+                                .semantic_model
+                                .scopes
+                                .ancestors(self.semantic_model.scope_id)
+                                .skip(1)
+                                .take_while(|scope| !scope.kind.is_module())
+                                .all(|scope| !scope.declares(name.as_str()))
+                            {
                                 self.diagnostics.push(Diagnostic::new(
                                     pylint::rules::NonlocalWithoutBinding {
                                         name: name.to_string(),
@@ -4042,7 +4051,7 @@ where
                         let name_range =
                             helpers::excepthandler_name_range(excepthandler, self.locator).unwrap();
 
-                        if self.semantic_model.scope().defines(name) {
+                        if self.semantic_model.scope().has(name) {
                             self.handle_node_store(
                                 name,
                                 &Expr::Name(ast::ExprName {
@@ -4067,7 +4076,7 @@ where
 
                         if let Some(binding_id) = {
                             let scope = self.semantic_model.scope_mut();
-                            scope.remove(name)
+                            scope.delete(name)
                         } {
                             if !self.semantic_model.is_used(binding_id) {
                                 if self.enabled(Rule::UnusedVariable) {
@@ -4697,13 +4706,8 @@ impl<'a> Checker<'a> {
         }
 
         let scope = self.semantic_model.scope_mut();
-        if scope.remove(id.as_str()).is_some() {
-            return;
-        }
-        if !self.enabled(Rule::UndefinedName) {
-            return;
-        }
-
+        if scope.delete(id.as_str()).is_none() {
+            if self.enabled(Rule::UndefinedName) {
                 self.diagnostics.push(Diagnostic::new(
                     pyflakes::rules::UndefinedName {
                         name: id.to_string(),
@@ -4711,6 +4715,8 @@ impl<'a> Checker<'a> {
                     expr.range(),
                 ));
             }
+        }
+    }
 
     fn check_deferred_future_type_definitions(&mut self) {
         while !self.deferred.future_type_definitions.is_empty() {
@@ -4977,7 +4983,7 @@ impl<'a> Checker<'a> {
                         .collect();
                     if !sources.is_empty() {
                         for (name, range) in &exports {
-                            if !scope.defines(name) {
+                            if !scope.has(name) {
                                 diagnostics.push(Diagnostic::new(
                                     pyflakes::rules::UndefinedLocalWithImportStarUsage {
                                         name: (*name).to_string(),

crates/ruff/src/rules/pyflakes/rules/strings.rs

@@ -386,7 +386,9 @@ pub struct StringDotFormatExtraNamedArguments {
     missing: Vec<String>,
 }
 
-impl AlwaysAutofixableViolation for StringDotFormatExtraNamedArguments {
+impl Violation for StringDotFormatExtraNamedArguments {
+    const AUTOFIX: AutofixKind = AutofixKind::Sometimes;
+
     #[derive_message_formats]
     fn message(&self) -> String {
         let StringDotFormatExtraNamedArguments { missing } = self;
@@ -394,10 +396,10 @@ impl AlwaysAutofixableViolation for StringDotFormatExtraNamedArguments {
         format!("`.format` call has unused named argument(s): {message}")
     }
 
-    fn autofix_title(&self) -> String {
+    fn autofix_title(&self) -> Option<String> {
         let StringDotFormatExtraNamedArguments { missing } = self;
         let message = missing.join(", ");
-        format!("Remove extra named arguments: {message}")
+        Some(format!("Remove extra named arguments: {message}"))
     }
 }
 

crates/ruff/src/rules/pyflakes/rules/undefined_export.rs

@@ -51,7 +51,7 @@ impl Violation for UndefinedExport {
 pub(crate) fn undefined_export(name: &str, range: TextRange, scope: &Scope) -> Vec<Diagnostic> {
     let mut diagnostics = Vec::new();
     if !scope.uses_star_imports() {
-        if !scope.defines(name) {
+        if !scope.has(name) {
             diagnostics.push(Diagnostic::new(
                 UndefinedExport {
                     name: (*name).to_string(),

crates/ruff/src/rules/pyflakes/rules/undefined_local.rs

@@ -47,7 +47,7 @@ impl Violation for UndefinedLocal {
 pub(crate) fn undefined_local(checker: &mut Checker, name: &str) {
     // If the name hasn't already been defined in the current scope...
     let current = checker.semantic_model().scope();
-    if !current.kind.is_any_function() || current.defines(name) {
+    if !current.kind.is_any_function() || current.has(name) {
         return;
     }
 

crates/ruff/src/rules/pyflakes/rules/unused_variable.rs

@@ -1,5 +1,5 @@
 use itertools::Itertools;
-use ruff_text_size::TextRange;
+use ruff_text_size::{TextRange, TextSize};
 use rustpython_parser::ast::{self, Ranged, Stmt};
 use rustpython_parser::{lexer, Mode, Tok};
 
@@ -61,21 +61,37 @@ impl Violation for UnusedVariable {
     }
 }
 
-/// Return the [`TextRange`] of the token after the next match of
-/// the predicate, skipping over any bracketed expressions.
-fn match_token_after<F, T>(located: &T, locator: &Locator, f: F) -> TextRange
+/// Return the [`TextRange`] of the token before the next match of the predicate
+fn match_token_before<F>(location: TextSize, locator: &Locator, f: F) -> Option<TextRange>
 where
     F: Fn(Tok) -> bool,
-    T: Ranged,
 {
-    let contents = locator.after(located.start());
+    let contents = locator.after(location);
+    for ((_, range), (tok, _)) in lexer::lex_starts_at(contents, Mode::Module, location)
+        .flatten()
+        .tuple_windows()
+    {
+        if f(tok) {
+            return Some(range);
+        }
+    }
+    None
+}
+
+/// Return the [`TextRange`] of the token after the next match of the predicate, skipping over
+/// any bracketed expressions.
+fn match_token_after<F>(location: TextSize, locator: &Locator, f: F) -> Option<TextRange>
+where
+    F: Fn(Tok) -> bool,
+{
+    let contents = locator.after(location);
 
     // Track the bracket depth.
     let mut par_count = 0u32;
     let mut sqb_count = 0u32;
     let mut brace_count = 0u32;
 
-    for ((tok, _), (_, range)) in lexer::lex_starts_at(contents, Mode::Module, located.start())
+    for ((tok, _), (_, range)) in lexer::lex_starts_at(contents, Mode::Module, location)
         .flatten()
         .tuple_windows()
     {
@@ -91,97 +107,83 @@ where
             }
             Tok::Rpar => {
                 par_count = par_count.saturating_sub(1);
-                // If this is a closing bracket, continue.
-                if par_count == 0 {
-                    continue;
-                }
             }
             Tok::Rsqb => {
                 sqb_count = sqb_count.saturating_sub(1);
-                // If this is a closing bracket, continue.
-                if sqb_count == 0 {
-                    continue;
-                }
             }
             Tok::Rbrace => {
                 brace_count = brace_count.saturating_sub(1);
-                // If this is a closing bracket, continue.
-                if brace_count == 0 {
-                    continue;
-                }
             }
             _ => {}
         }
+
         // If we're in nested brackets, continue.
         if par_count > 0 || sqb_count > 0 || brace_count > 0 {
             continue;
         }
 
         if f(tok) {
-            return range;
+            return Some(range);
         }
     }
-    unreachable!("No token after matched");
+    None
 }
 
-/// Return the [`TextRange`] of the token matching the predicate,
-/// skipping over any bracketed expressions.
-fn match_token<F, T>(located: &T, locator: &Locator, f: F) -> TextRange
+/// Return the [`TextRange`] of the token matching the predicate or the first mismatched
+/// bracket, skipping over any bracketed expressions.
+fn match_token_or_closing_brace<F>(location: TextSize, locator: &Locator, f: F) -> Option<TextRange>
 where
     F: Fn(Tok) -> bool,
-    T: Ranged,
 {
-    let contents = locator.after(located.start());
+    let contents = locator.after(location);
 
     // Track the bracket depth.
-    let mut par_count = 0;
-    let mut sqb_count = 0;
-    let mut brace_count = 0;
+    let mut par_count = 0u32;
+    let mut sqb_count = 0u32;
+    let mut brace_count = 0u32;
 
-    for (tok, range) in lexer::lex_starts_at(contents, Mode::Module, located.start()).flatten() {
+    for (tok, range) in lexer::lex_starts_at(contents, Mode::Module, location).flatten() {
         match tok {
             Tok::Lpar => {
-                par_count += 1;
+                par_count = par_count.saturating_add(1);
             }
             Tok::Lsqb => {
-                sqb_count += 1;
+                sqb_count = sqb_count.saturating_add(1);
             }
             Tok::Lbrace => {
-                brace_count += 1;
+                brace_count = brace_count.saturating_add(1);
             }
             Tok::Rpar => {
-                par_count -= 1;
-                // If this is a closing bracket, continue.
                 if par_count == 0 {
-                    continue;
+                    return Some(range);
                 }
+                par_count = par_count.saturating_sub(1);
             }
             Tok::Rsqb => {
-                sqb_count -= 1;
-                // If this is a closing bracket, continue.
                 if sqb_count == 0 {
-                    continue;
+                    return Some(range);
                 }
+                sqb_count = sqb_count.saturating_sub(1);
             }
             Tok::Rbrace => {
-                brace_count -= 1;
-                // If this is a closing bracket, continue.
                 if brace_count == 0 {
-                    continue;
+                    return Some(range);
                 }
+                brace_count = brace_count.saturating_sub(1);
             }
             _ => {}
         }
+
         // If we're in nested brackets, continue.
         if par_count > 0 || sqb_count > 0 || brace_count > 0 {
             continue;
         }
 
         if f(tok) {
-            return range;
+            return Some(range);
         }
     }
-    unreachable!("No token after matched");
+    None
 }
 
 /// Generate a [`Edit`] to remove an unused variable assignment, given the
@@ -201,10 +203,10 @@ fn remove_unused_variable(
                 {
                     // If the expression is complex (`x = foo()`), remove the assignment,
                     // but preserve the right-hand side.
-                    let edit = Edit::deletion(
-                        target.start(),
-                        match_token_after(target, checker.locator, |tok| tok == Tok::Equal).start(),
-                    );
+                    let start = target.start();
+                    let end =
+                        match_token_after(start, checker.locator, |tok| tok == Tok::Equal)?.start();
+                    let edit = Edit::deletion(start, end);
                     Some(Fix::suggested(edit))
                 } else {
                     // If (e.g.) assigning to a constant (`x = 1`), delete the entire statement.
@@ -232,10 +234,10 @@ fn remove_unused_variable(
             return if contains_effect(value, |id| checker.semantic_model().is_builtin(id)) {
                 // If the expression is complex (`x = foo()`), remove the assignment,
                 // but preserve the right-hand side.
-                let edit = Edit::deletion(
-                    stmt.start(),
-                    match_token_after(stmt, checker.locator, |tok| tok == Tok::Equal).start(),
-                );
+                let start = stmt.start();
+                let end =
+                    match_token_after(start, checker.locator, |tok| tok == Tok::Equal)?.start();
+                let edit = Edit::deletion(start, end);
                 Some(Fix::suggested(edit))
             } else {
                 // If (e.g.) assigning to a constant (`x = 1`), delete the entire statement.
@@ -258,15 +260,20 @@ fn remove_unused_variable(
         for item in items {
             if let Some(optional_vars) = &item.optional_vars {
                 if optional_vars.range() == range {
-                    let edit = Edit::deletion(
-                        item.context_expr.end(),
-                        // The end of the `Withitem` is the colon, comma, or closing
-                        // parenthesis following the `optional_vars`.
-                        match_token(&item.context_expr, checker.locator, |tok| {
-                            tok == Tok::Colon || tok == Tok::Comma || tok == Tok::Rpar
-                        })
-                        .start(),
-                    );
+                    // Find the first token before the `as` keyword.
+                    let start =
+                        match_token_before(item.context_expr.start(), checker.locator, |tok| {
+                            tok == Tok::As
+                        })?
+                        .end();
+
+                    // Find the first colon, comma, or closing bracket after the `as` keyword.
+                    let end = match_token_or_closing_brace(start, checker.locator, |tok| {
+                        tok == Tok::Colon || tok == Tok::Comma
+                    })?
+                    .start();
+
+                    let edit = Edit::deletion(start, end);
                     return Some(Fix::suggested(edit));
                 }
             }

crates/ruff/src/rules/pyflakes/snapshots/ruff__rules__pyflakes__tests__F522_F522.py.snap

@@ -33,6 +33,7 @@ F522.py:2:1: F522 [*] `.format` call has unused named argument(s): spam
   2 |+"{bar}{}".format(1, bar=2, )  # F522
 3 3 | "{bar:{spam}}".format(bar=2, spam=3)  # No issues
 4 4 | "{bar:{spam}}".format(bar=2, spam=3, eggs=4, ham=5)  # F522
+5 5 | # Not fixable
 
 F522.py:4:1: F522 [*] `.format` call has unused named argument(s): eggs, ham
   |
@@ -40,6 +41,8 @@ F522.py:4:1: F522 [*] `.format` call has unused named argument(s): eggs, ham
 5 | "{bar:{spam}}".format(bar=2, spam=3)  # No issues
 6 | "{bar:{spam}}".format(bar=2, spam=3, eggs=4, ham=5)  # F522
   | ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ F522
+7 | # Not fixable
+8 | (''
   |
   = help: Remove extra named arguments: eggs, ham
 
@@ -49,5 +52,19 @@ F522.py:4:1: F522 [*] `.format` call has unused named argument(s): eggs, ham
 3 3 | "{bar:{spam}}".format(bar=2, spam=3)  # No issues
 4   |-"{bar:{spam}}".format(bar=2, spam=3, eggs=4, ham=5)  # F522
   4 |+"{bar:{spam}}".format(bar=2, spam=3, )  # F522
+5 5 | # Not fixable
+6 6 | (''
+7 7 |  .format(x=2))
+
+F522.py:6:2: F522 `.format` call has unused named argument(s): x
+  |
+6 |   "{bar:{spam}}".format(bar=2, spam=3, eggs=4, ham=5)  # F522
+7 |   # Not fixable
+8 |   (''
+  |  __^
+9 | |  .format(x=2))
+  | |_____________^ F522
+  |
+  = help: Remove extra named arguments: x
 
 

crates/ruff/src/rules/pyflakes/snapshots/ruff__rules__pyflakes__tests__F841_F841_3.py.snap

@@ -377,126 +377,145 @@ F841_3.py:77:25: F841 [*] Local variable `cm` is assigned to but never used
 79 79 | 
 80 80 | 
 
-F841_3.py:87:5: F841 [*] Local variable `toplevel` is assigned to but never used
+F841_3.py:87:26: F841 [*] Local variable `cm` is assigned to but never used
    |
 87 | def f():
-88 |     toplevel = tt = lexer.get_token()
-   |     ^^^^^^^^ F841
-89 |     if not tt:
-90 |         break
+88 |     with (Nested(m)) as (cm):
+   |                          ^^ F841
+89 |         pass
    |
-   = help: Remove assignment to unused variable `toplevel`
+   = help: Remove assignment to unused variable `cm`
 
 ℹ Suggested fix
 84 84 | 
 85 85 | 
 86 86 | def f():
-87    |-    toplevel = tt = lexer.get_token()
-   87 |+    tt = lexer.get_token()
-88 88 |     if not tt:
-89 89 |         break
+87    |-    with (Nested(m)) as (cm):
+   87 |+    with (Nested(m)):
+88 88 |         pass
+89 89 | 
 90 90 | 
 
-F841_3.py:93:5: F841 [*] Local variable `toplevel` is assigned to but never used
+F841_3.py:92:5: F841 [*] Local variable `toplevel` is assigned to but never used
    |
-93 | def f():
-94 |     toplevel = tt = lexer.get_token()
+92 | def f():
+93 |     toplevel = tt = lexer.get_token()
+   |     ^^^^^^^^ F841
+94 |     if not tt:
+95 |         break
+   |
+   = help: Remove assignment to unused variable `toplevel`
+
+ℹ Suggested fix
+89 89 | 
+90 90 | 
+91 91 | def f():
+92    |-    toplevel = tt = lexer.get_token()
+   92 |+    tt = lexer.get_token()
+93 93 |     if not tt:
+94 94 |         break
+95 95 | 
+
+F841_3.py:98:5: F841 [*] Local variable `toplevel` is assigned to but never used
+   |
+98 | def f():
+99 |     toplevel = tt = lexer.get_token()
    |     ^^^^^^^^ F841
    |
    = help: Remove assignment to unused variable `toplevel`
 
 ℹ Suggested fix
-90 90 | 
-91 91 | 
-92 92 | def f():
-93    |-    toplevel = tt = lexer.get_token()
-   93 |+    tt = lexer.get_token()
-94 94 | 
 95 95 | 
-96 96 | def f():
+96 96 | 
+97 97 | def f():
+98    |-    toplevel = tt = lexer.get_token()
+   98 |+    tt = lexer.get_token()
+99 99 | 
+100 100 | 
+101 101 | def f():
 
-F841_3.py:93:16: F841 [*] Local variable `tt` is assigned to but never used
+F841_3.py:98:16: F841 [*] Local variable `tt` is assigned to but never used
    |
-93 | def f():
-94 |     toplevel = tt = lexer.get_token()
+98 | def f():
+99 |     toplevel = tt = lexer.get_token()
    |                ^^ F841
    |
    = help: Remove assignment to unused variable `tt`
 
 ℹ Suggested fix
-90 90 | 
-91 91 | 
-92 92 | def f():
-93    |-    toplevel = tt = lexer.get_token()
-   93 |+    toplevel = lexer.get_token()
-94 94 | 
 95 95 | 
-96 96 | def f():
-
-F841_3.py:97:5: F841 [*] Local variable `toplevel` is assigned to but never used
-   |
-97 | def f():
-98 |     toplevel = (a, b) = lexer.get_token()
-   |     ^^^^^^^^ F841
-   |
-   = help: Remove assignment to unused variable `toplevel`
-
-ℹ Suggested fix
-94 94 | 
-95 95 | 
-96 96 | def f():
-97    |-    toplevel = (a, b) = lexer.get_token()
-   97 |+    (a, b) = lexer.get_token()
-98 98 | 
+96 96 | 
+97 97 | def f():
+98    |-    toplevel = tt = lexer.get_token()
+   98 |+    toplevel = lexer.get_token()
 99 99 | 
-100 100 | def f():
+100 100 | 
+101 101 | def f():
 
-F841_3.py:101:14: F841 [*] Local variable `toplevel` is assigned to but never used
+F841_3.py:102:5: F841 [*] Local variable `toplevel` is assigned to but never used
     |
-101 | def f():
-102 |     (a, b) = toplevel = lexer.get_token()
+102 | def f():
+103 |     toplevel = (a, b) = lexer.get_token()
     |     ^^^^^^^^ F841
     |
     = help: Remove assignment to unused variable `toplevel`
 
 ℹ Suggested fix
-98  98  | 
 99  99  | 
-100 100 | def f():
-101     |-    (a, b) = toplevel = lexer.get_token()
-    101 |+    (a, b) = lexer.get_token()
-102 102 | 
+100 100 | 
+101 101 | def f():
+102     |-    toplevel = (a, b) = lexer.get_token()
+    102 |+    (a, b) = lexer.get_token()
 103 103 | 
-104 104 | def f():
+104 104 | 
+105 105 | def f():
 
-F841_3.py:105:5: F841 [*] Local variable `toplevel` is assigned to but never used
+F841_3.py:106:14: F841 [*] Local variable `toplevel` is assigned to but never used
     |
-105 | def f():
-106 |     toplevel = tt = 1
+106 | def f():
+107 |     (a, b) = toplevel = lexer.get_token()
     |              ^^^^^^^^ F841
     |
     = help: Remove assignment to unused variable `toplevel`
 
 ℹ Suggested fix
-102 102 | 
 103 103 | 
-104 104 | def f():
-105     |-    toplevel = tt = 1
-    105 |+    tt = 1
+104 104 | 
+105 105 | def f():
+106     |-    (a, b) = toplevel = lexer.get_token()
+    106 |+    (a, b) = lexer.get_token()
+107 107 | 
+108 108 | 
+109 109 | def f():
 
-F841_3.py:105:16: F841 [*] Local variable `tt` is assigned to but never used
+F841_3.py:110:5: F841 [*] Local variable `toplevel` is assigned to but never used
     |
-105 | def f():
-106 |     toplevel = tt = 1
+110 | def f():
+111 |     toplevel = tt = 1
+    |     ^^^^^^^^ F841
+    |
+    = help: Remove assignment to unused variable `toplevel`
+
+ℹ Suggested fix
+107 107 | 
+108 108 | 
+109 109 | def f():
+110     |-    toplevel = tt = 1
+    110 |+    tt = 1
+
+F841_3.py:110:16: F841 [*] Local variable `tt` is assigned to but never used
+    |
+110 | def f():
+111 |     toplevel = tt = 1
     |                ^^ F841
     |
     = help: Remove assignment to unused variable `tt`
 
 ℹ Suggested fix
-102 102 | 
-103 103 | 
-104 104 | def f():
-105     |-    toplevel = tt = 1
-    105 |+    toplevel = 1
+107 107 | 
+108 108 | 
+109 109 | def f():
+110     |-    toplevel = tt = 1
+    110 |+    toplevel = 1
 
 

crates/ruff/src/rules/pylint/rules/iteration_over_set.rs

@@ -1,4 +1,4 @@
-use rustpython_parser::ast::{Expr, ExprName, Ranged};
+use rustpython_parser::ast::{Expr, Ranged};
 
 use ruff_diagnostics::{Diagnostic, Violation};
 use ruff_macros::{derive_message_formats, violation};
@@ -6,7 +6,7 @@ use ruff_macros::{derive_message_formats, violation};
 use crate::checkers::ast::Checker;
 
 /// ## What it does
-/// Checks for iterations over `set` literals and comprehensions.
+/// Checks for iterations over `set` literals.
 ///
 /// ## Why is this bad?
 /// Iterating over a `set` is less efficient than iterating over a sequence
@@ -38,23 +38,7 @@ impl Violation for IterationOverSet {
 
 /// PLC0208
 pub(crate) fn iteration_over_set(checker: &mut Checker, expr: &Expr) {
-    let is_set = match expr {
-        // Ex) `for i in {1, 2, 3}`
-        Expr::Set(_) => true,
-        // Ex)` for i in {n for n in range(1, 4)}`
-        Expr::SetComp(_) => true,
-        // Ex) `for i in set(1, 2, 3)`
-        Expr::Call(call) => {
-            if let Expr::Name(ExprName { id, .. }) = call.func.as_ref() {
-                id.as_str() == "set" && checker.semantic_model().is_builtin("set")
-            } else {
-                false
-            }
-        }
-        _ => false,
-    };
-
-    if is_set {
+    if expr.is_set_expr() {
         checker
             .diagnostics
             .push(Diagnostic::new(IterationOverSet, expr.range()));

crates/ruff/src/rules/pylint/snapshots/ruff__rules__pylint__tests__PLC0208_iteration_over_set.py.snap

@@ -10,62 +10,44 @@ iteration_over_set.py:3:13: PLC0208 Use a sequence type instead of a `set` when
 6 |     print(f"I like {item}.")
   |
 
-iteration_over_set.py:6:13: PLC0208 Use a sequence type instead of a `set` when iterating over values
+iteration_over_set.py:6:28: PLC0208 Use a sequence type instead of a `set` when iterating over values
    |
-6 |     print(f"I like {item}.")
-7 | 
-8 | for item in set(("apples", "lemons", "water")):  # flags set() calls
-  |             ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ PLC0208
-9 |     print(f"I like {item}.")
+ 6 |     print(f"I like {item}.")
+ 7 | 
+ 8 | numbers_list = [i for i in {1, 2, 3}]  # flags sets in list comprehensions
+   |                            ^^^^^^^^^ PLC0208
+ 9 | 
+10 | numbers_set = {i for i in {1, 2, 3}}  # flags sets in set comprehensions
    |
 
-iteration_over_set.py:9:15: PLC0208 Use a sequence type instead of a `set` when iterating over values
+iteration_over_set.py:8:27: PLC0208 Use a sequence type instead of a `set` when iterating over values
    |
- 9 |     print(f"I like {item}.")
-10 | 
-11 | for number in {i for i in range(10)}:  # flags set comprehensions
-   |               ^^^^^^^^^^^^^^^^^^^^^^ PLC0208
-12 |     print(number)
+ 8 | numbers_list = [i for i in {1, 2, 3}]  # flags sets in list comprehensions
+ 9 | 
+10 | numbers_set = {i for i in {1, 2, 3}}  # flags sets in set comprehensions
+   |                           ^^^^^^^^^ PLC0208
+11 | 
+12 | numbers_dict = {str(i): i for i in {1, 2, 3}}  # flags sets in dict comprehensions
    |
 
-iteration_over_set.py:12:28: PLC0208 Use a sequence type instead of a `set` when iterating over values
+iteration_over_set.py:10:36: PLC0208 Use a sequence type instead of a `set` when iterating over values
    |
-12 |     print(number)
+10 | numbers_set = {i for i in {1, 2, 3}}  # flags sets in set comprehensions
+11 | 
+12 | numbers_dict = {str(i): i for i in {1, 2, 3}}  # flags sets in dict comprehensions
+   |                                    ^^^^^^^^^ PLC0208
 13 | 
-14 | numbers_list = [i for i in {1, 2, 3}]  # flags sets in list comprehensions
+14 | numbers_gen = (i for i in {1, 2, 3})  # flags sets in generator expressions
+   |
+
+iteration_over_set.py:12:27: PLC0208 Use a sequence type instead of a `set` when iterating over values
+   |
+12 | numbers_dict = {str(i): i for i in {1, 2, 3}}  # flags sets in dict comprehensions
+13 | 
+14 | numbers_gen = (i for i in {1, 2, 3})  # flags sets in generator expressions
    |                           ^^^^^^^^^ PLC0208
 15 | 
-16 | numbers_set = {i for i in {1, 2, 3}}  # flags sets in set comprehensions
-   |
-
-iteration_over_set.py:14:27: PLC0208 Use a sequence type instead of a `set` when iterating over values
-   |
-14 | numbers_list = [i for i in {1, 2, 3}]  # flags sets in list comprehensions
-15 | 
-16 | numbers_set = {i for i in {1, 2, 3}}  # flags sets in set comprehensions
-   |                           ^^^^^^^^^ PLC0208
-17 | 
-18 | numbers_dict = {str(i): i for i in {1, 2, 3}}  # flags sets in dict comprehensions
-   |
-
-iteration_over_set.py:16:36: PLC0208 Use a sequence type instead of a `set` when iterating over values
-   |
-16 | numbers_set = {i for i in {1, 2, 3}}  # flags sets in set comprehensions
-17 | 
-18 | numbers_dict = {str(i): i for i in {1, 2, 3}}  # flags sets in dict comprehensions
-   |                                    ^^^^^^^^^ PLC0208
-19 | 
-20 | numbers_gen = (i for i in {1, 2, 3})  # flags sets in generator expressions
-   |
-
-iteration_over_set.py:18:27: PLC0208 Use a sequence type instead of a `set` when iterating over values
-   |
-18 | numbers_dict = {str(i): i for i in {1, 2, 3}}  # flags sets in dict comprehensions
-19 | 
-20 | numbers_gen = (i for i in {1, 2, 3})  # flags sets in generator expressions
-   |                           ^^^^^^^^^ PLC0208
-21 | 
-22 | # Non-errors
+16 | # Non-errors
    |
 
 

crates/ruff_cli/Cargo.toml

@@ -1,6 +1,6 @@
 [package]
 name = "ruff_cli"
-version = "0.0.270"
+version = "0.0.271"
 authors = ["Charlie Marsh <charlie.r.marsh@gmail.com>"]
 edition = { workspace = true }
 rust-version = { workspace = true }

crates/ruff_python_ast/src/function.rs

@@ -0,0 +1,114 @@
+use crate::node::AnyNodeRef;
+use ruff_text_size::TextRange;
+use rustpython_ast::{
+    Arguments, Expr, Identifier, Ranged, StmtAsyncFunctionDef, StmtFunctionDef, Suite,
+};
+
+/// Enum that represents any python function definition.
+#[derive(Copy, Clone, PartialEq, Debug)]
+pub enum AnyFunctionDefinition<'a> {
+    FunctionDefinition(&'a StmtFunctionDef),
+    AsyncFunctionDefinition(&'a StmtAsyncFunctionDef),
+}
+
+impl<'a> AnyFunctionDefinition<'a> {
+    pub const fn cast_ref(reference: AnyNodeRef<'a>) -> Option<Self> {
+        match reference {
+            AnyNodeRef::StmtAsyncFunctionDef(definition) => {
+                Some(Self::AsyncFunctionDefinition(definition))
+            }
+            AnyNodeRef::StmtFunctionDef(definition) => Some(Self::FunctionDefinition(definition)),
+            _ => None,
+        }
+    }
+
+    /// Returns `Some` if this is a [`StmtFunctionDef`] and `None` otherwise.
+    pub const fn as_function_definition(self) -> Option<&'a StmtFunctionDef> {
+        if let Self::FunctionDefinition(definition) = self {
+            Some(definition)
+        } else {
+            None
+        }
+    }
+
+    /// Returns `Some` if this is a [`StmtAsyncFunctionDef`] and `None` otherwise.
+    pub const fn as_async_function_definition(self) -> Option<&'a StmtAsyncFunctionDef> {
+        if let Self::AsyncFunctionDefinition(definition) = self {
+            Some(definition)
+        } else {
+            None
+        }
+    }
+
+    /// Returns the function's name
+    pub const fn name(self) -> &'a Identifier {
+        match self {
+            Self::FunctionDefinition(definition) => &definition.name,
+            Self::AsyncFunctionDefinition(definition) => &definition.name,
+        }
+    }
+
+    /// Returns the function arguments (parameters).
+    pub fn arguments(self) -> &'a Arguments {
+        match self {
+            Self::FunctionDefinition(definition) => definition.args.as_ref(),
+            Self::AsyncFunctionDefinition(definition) => definition.args.as_ref(),
+        }
+    }
+
+    /// Returns the function's body
+    pub const fn body(self) -> &'a Suite {
+        match self {
+            Self::FunctionDefinition(definition) => &definition.body,
+            Self::AsyncFunctionDefinition(definition) => &definition.body,
+        }
+    }
+
+    /// Returns the decorators attributing the function.
+    pub fn decorators(self) -> &'a [Expr] {
+        match self {
+            Self::FunctionDefinition(definition) => &definition.decorator_list,
+            Self::AsyncFunctionDefinition(definition) => &definition.decorator_list,
+        }
+    }
+
+    pub fn returns(self) -> Option<&'a Expr> {
+        match self {
+            Self::FunctionDefinition(definition) => definition.returns.as_deref(),
+            Self::AsyncFunctionDefinition(definition) => definition.returns.as_deref(),
+        }
+    }
+
+    pub fn type_comments(self) -> Option<&'a str> {
+        match self {
+            Self::FunctionDefinition(definition) => definition.type_comment.as_deref(),
+            Self::AsyncFunctionDefinition(definition) => definition.type_comment.as_deref(),
+        }
+    }
+
+    /// Returns `true` if this is [`Self::AsyncFunctionDefinition`]
+    pub const fn is_async(self) -> bool {
+        matches!(self, Self::AsyncFunctionDefinition(_))
+    }
+}
+
+impl Ranged for AnyFunctionDefinition<'_> {
+    fn range(&self) -> TextRange {
+        match self {
+            AnyFunctionDefinition::FunctionDefinition(definition) => definition.range(),
+            AnyFunctionDefinition::AsyncFunctionDefinition(definition) => definition.range(),
+        }
+    }
+}
+
+impl<'a> From<&'a StmtFunctionDef> for AnyFunctionDefinition<'a> {
+    fn from(value: &'a StmtFunctionDef) -> Self {
+        Self::FunctionDefinition(value)
+    }
+}
+
+impl<'a> From<&'a StmtAsyncFunctionDef> for AnyFunctionDefinition<'a> {
+    fn from(value: &'a StmtAsyncFunctionDef) -> Self {
+        Self::AsyncFunctionDefinition(value)
+    }
+}

crates/ruff_python_ast/src/lib.rs

@@ -2,6 +2,7 @@ pub mod all;
 pub mod call_path;
 pub mod cast;
 pub mod comparable;
+pub mod function;
 pub mod hashable;
 pub mod helpers;
 pub mod imports;

crates/ruff_python_formatter/resources/test/fixtures/ruff/expression/binary_expression.py

@@ -0,0 +1,43 @@
+(aaaaaaaa
+    + # trailing operator comment
+    b # trailing right comment
+)
+
+
+(aaaaaaaa # trailing left comment
+    +  # trailing operator comment
+    # leading right comment
+    b
+)
+
+
+# Black breaks the right side first for the following expressions:
+aaaaaaaaaaaaaa + caaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaal(argument1, argument2, argument3)
+aaaaaaaaaaaaaa + [bbbbbbbbbbbbbbbbbbbbbb, ccccccccccccccccccccc, dddddddddddddddd, eeeeeee]
+aaaaaaaaaaaaaa + (bbbbbbbbbbbbbbbbbbbbbb, ccccccccccccccccccccc, dddddddddddddddd, eeeeeee)
+aaaaaaaaaaaaaa + { key1:bbbbbbbbbbbbbbbbbbbbbb, key2: ccccccccccccccccccccc, key3: dddddddddddddddd, key4: eeeeeee }
+aaaaaaaaaaaaaa + { bbbbbbbbbbbbbbbbbbbbbb, ccccccccccccccccccccc, dddddddddddddddd, eeeeeee }
+aaaaaaaaaaaaaa + [a for x in bbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbb ]
+aaaaaaaaaaaaaa + (a for x in bbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbb )
+aaaaaaaaaaaaaa + {a for x in bbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbb}
+
+# Wraps it in parentheses if it needs to break both left and right
+aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa + [
+    bbbbbbbbbbbbbbbbbbbbbb,
+    ccccccccccccccccccccc,
+    dddddddddddddddd,
+    eee
+] # comment
+
+
+
+# But only for expressions that have a statement parent.
+not (aaaaaaaaaaaaaa + {a for x in bbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbb})
+[a + [bbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbb] in c ]
+
+
+# leading comment
+(
+    # comment
+    content + b
+)

crates/ruff_python_formatter/src/builders.rs

@@ -12,7 +12,7 @@ pub(crate) trait PyFormatterExtensions<'ast, 'buf> {
     ///
     /// * [`NodeLevel::Module`]: Up to two empty lines
     /// * [`NodeLevel::CompoundStatement`]: Up to one empty line
-    /// * [`NodeLevel::Parenthesized`]: No empty lines
+    /// * [`NodeLevel::Expression`]: No empty lines
     fn join_nodes<'fmt>(&'fmt mut self, level: NodeLevel) -> JoinNodesBuilder<'fmt, 'ast, 'buf>;
 }
 
@@ -48,18 +48,18 @@ impl<'fmt, 'ast, 'buf> JoinNodesBuilder<'fmt, 'ast, 'buf> {
     {
         let node_level = self.node_level;
         let separator = format_with(|f: &mut PyFormatter| match node_level {
-            NodeLevel::TopLevel => match lines_before(f.context().contents(), node.start()) {
+            NodeLevel::TopLevel => match lines_before(node.start(), f.context().contents()) {
                 0 | 1 => hard_line_break().fmt(f),
                 2 => empty_line().fmt(f),
                 _ => write!(f, [empty_line(), empty_line()]),
             },
             NodeLevel::CompoundStatement => {
-                match lines_before(f.context().contents(), node.start()) {
+                match lines_before(node.start(), f.context().contents()) {
                     0 | 1 => hard_line_break().fmt(f),
                     _ => empty_line().fmt(f),
                 }
             }
-            NodeLevel::Parenthesized => hard_line_break().fmt(f),
+            NodeLevel::Expression => hard_line_break().fmt(f),
         });
 
         self.entry_with_separator(&separator, content);
@@ -200,7 +200,7 @@ no_leading_newline = 30"#
     // Removes all empty lines
     #[test]
     fn ranged_builder_parenthesized_level() {
-        let printed = format_ranged(NodeLevel::Parenthesized);
+        let printed = format_ranged(NodeLevel::Expression);
 
         assert_eq!(
             &printed,

crates/ruff_python_formatter/src/comments/format.rs

@@ -39,7 +39,7 @@ impl Format<PyFormatContext<'_>> for FormatLeadingComments<'_> {
         for comment in leading_comments {
             let slice = comment.slice();
 
-            let lines_after_comment = lines_after(f.context().contents(), slice.end());
+            let lines_after_comment = lines_after(slice.end(), f.context().contents());
             write!(
                 f,
                 [format_comment(comment), empty_lines(lines_after_comment)]
@@ -80,7 +80,7 @@ impl Format<PyFormatContext<'_>> for FormatLeadingAlternateBranchComments<'_> {
         if let Some(first_leading) = self.comments.first() {
             // Leading comments only preserves the lines after the comment but not before.
             // Insert the necessary lines.
-            if lines_before(f.context().contents(), first_leading.slice().start()) > 1 {
+            if lines_before(first_leading.slice().start(), f.context().contents()) > 1 {
                 write!(f, [empty_line()])?;
             }
 
@@ -88,7 +88,7 @@ impl Format<PyFormatContext<'_>> for FormatLeadingAlternateBranchComments<'_> {
         } else if let Some(last_preceding) = self.last_node {
             // The leading comments formatting ensures that it preserves the right amount of lines after
             // We need to take care of this ourselves, if there's no leading `else` comment.
-            if lines_after(f.context().contents(), last_preceding.end()) > 1 {
+            if lines_after(last_preceding.end(), f.context().contents()) > 1 {
                 write!(f, [empty_line()])?;
             }
         }
@@ -132,7 +132,7 @@ impl Format<PyFormatContext<'_>> for FormatTrailingComments<'_> {
             has_trailing_own_line_comment |= trailing.position().is_own_line();
 
             if has_trailing_own_line_comment {
-                let lines_before_comment = lines_before(f.context().contents(), slice.start());
+                let lines_before_comment = lines_before(slice.start(), f.context().contents());
 
                 // A trailing comment at the end of a body or list
                 // ```python
@@ -175,20 +175,26 @@ pub(crate) fn dangling_node_comments<T>(node: &T) -> FormatDanglingComments
 where
     T: AstNode,
 {
-    FormatDanglingComments {
-        node: node.as_any_node_ref(),
-    }
+    FormatDanglingComments::Node(node.as_any_node_ref())
 }
 
-pub(crate) struct FormatDanglingComments<'a> {
-    node: AnyNodeRef<'a>,
+pub(crate) fn dangling_comments(comments: &[SourceComment]) -> FormatDanglingComments {
+    FormatDanglingComments::Comments(comments)
+}
+
+pub(crate) enum FormatDanglingComments<'a> {
+    Node(AnyNodeRef<'a>),
+    Comments(&'a [SourceComment]),
 }
 
 impl Format<PyFormatContext<'_>> for FormatDanglingComments<'_> {
     fn fmt(&self, f: &mut Formatter<PyFormatContext>) -> FormatResult<()> {
         let comments = f.context().comments().clone();
 
-        let dangling_comments = comments.dangling_comments(self.node);
+        let dangling_comments = match self {
+            Self::Comments(comments) => comments,
+            Self::Node(node) => comments.dangling_comments(*node),
+        };
 
         let mut first = true;
         for comment in dangling_comments {
@@ -200,7 +206,7 @@ impl Format<PyFormatContext<'_>> for FormatDanglingComments<'_> {
                 f,
                 [
                     format_comment(comment),
-                    empty_lines(lines_after(f.context().contents(), comment.slice().end()))
+                    empty_lines(lines_after(comment.slice().end(), f.context().contents()))
                 ]
             )?;
 
@@ -301,7 +307,7 @@ impl Format<PyFormatContext<'_>> for FormatEmptyLines {
             },
 
             // Remove all whitespace in parenthesized expressions
-            NodeLevel::Parenthesized => write!(f, [hard_line_break()]),
+            NodeLevel::Expression => write!(f, [hard_line_break()]),
         }
     }
 }

crates/ruff_python_formatter/src/comments/mod.rs

@@ -103,8 +103,8 @@ use crate::comments::map::MultiMap;
 use crate::comments::node_key::NodeRefEqualityKey;
 use crate::comments::visitor::CommentsVisitor;
 pub(crate) use format::{
-    dangling_node_comments, leading_alternate_branch_comments, leading_node_comments,
-    trailing_comments, trailing_node_comments,
+    dangling_comments, dangling_node_comments, leading_alternate_branch_comments,
+    leading_node_comments, trailing_comments, trailing_node_comments,
 };
 use ruff_formatter::{SourceCode, SourceCodeSlice};
 use ruff_python_ast::node::AnyNodeRef;

crates/ruff_python_formatter/src/comments/placement.rs

@@ -520,8 +520,8 @@ fn handle_trailing_end_of_line_condition_comment<'a>(
     if preceding.ptr_eq(last_before_colon) {
         let mut start = preceding.end();
         while let Some((offset, c)) = find_first_non_trivia_character_in_range(
-            locator.contents(),
             TextRange::new(start, following.start()),
+            locator.contents(),
         ) {
             match c {
                 ':' => {
@@ -655,7 +655,7 @@ fn handle_trailing_binary_expression_left_or_operator_comment<'a>(
     );
 
     let operator_offset = loop {
-        match find_first_non_trivia_character_in_range(locator.contents(), between_operands_range) {
+        match find_first_non_trivia_character_in_range(between_operands_range, locator.contents()) {
             // Skip over closing parens
             Some((offset, ')')) => {
                 between_operands_range =
@@ -733,17 +733,17 @@ fn find_pos_only_slash_offset(
     locator: &Locator,
 ) -> Option<TextSize> {
     // First find the comma separating the two arguments
-    find_first_non_trivia_character_in_range(locator.contents(), between_arguments_range).and_then(
+    find_first_non_trivia_character_in_range(between_arguments_range, locator.contents()).and_then(
         |(comma_offset, comma)| {
             debug_assert_eq!(comma, ',');
 
             // Then find the position of the `/` operator
             find_first_non_trivia_character_in_range(
-                locator.contents(),
                 TextRange::new(
                     comma_offset + TextSize::new(1),
                     between_arguments_range.end(),
                 ),
+                locator.contents(),
             )
             .map(|(offset, c)| {
                 debug_assert_eq!(c, '/');

crates/ruff_python_formatter/src/context.rs

@@ -82,5 +82,5 @@ pub(crate) enum NodeLevel {
     CompoundStatement,
 
     /// Formatting nodes that are enclosed in a parenthesized expression.
-    Parenthesized,
+    Expression,
 }

crates/ruff_python_formatter/src/expression/expr_attribute.rs

@@ -1,3 +1,6 @@
+use crate::expression::parentheses::{
+    default_expression_needs_parentheses, NeedsParentheses, Parentheses, Parenthesize,
+};
 use crate::{verbatim_text, FormatNodeRule, PyFormatter};
 use ruff_formatter::{write, Buffer, FormatResult};
 use rustpython_parser::ast::ExprAttribute;
@@ -10,3 +13,9 @@ impl FormatNodeRule<ExprAttribute> for FormatExprAttribute {
         write!(f, [verbatim_text(item.range)])
     }
 }
+
+impl NeedsParentheses for ExprAttribute {
+    fn needs_parentheses(&self, parenthesize: Parenthesize, source: &str) -> Parentheses {
+        default_expression_needs_parentheses(self.into(), parenthesize, source)
+    }
+}

crates/ruff_python_formatter/src/expression/expr_await.rs

@@ -1,3 +1,6 @@
+use crate::expression::parentheses::{
+    default_expression_needs_parentheses, NeedsParentheses, Parentheses, Parenthesize,
+};
 use crate::{verbatim_text, FormatNodeRule, PyFormatter};
 use ruff_formatter::{write, Buffer, FormatResult};
 use rustpython_parser::ast::ExprAwait;
@@ -10,3 +13,9 @@ impl FormatNodeRule<ExprAwait> for FormatExprAwait {
         write!(f, [verbatim_text(item.range)])
     }
 }
+
+impl NeedsParentheses for ExprAwait {
+    fn needs_parentheses(&self, parenthesize: Parenthesize, source: &str) -> Parentheses {
+        default_expression_needs_parentheses(self.into(), parenthesize, source)
+    }
+}

crates/ruff_python_formatter/src/expression/expr_bin_op.rs

@@ -1,12 +1,215 @@
-use crate::{verbatim_text, FormatNodeRule, PyFormatter};
-use ruff_formatter::{write, Buffer, FormatResult};
-use rustpython_parser::ast::ExprBinOp;
+use crate::comments::trailing_comments;
+use crate::expression::parentheses::{
+    default_expression_needs_parentheses, NeedsParentheses, Parenthesize,
+};
+use crate::expression::Parentheses;
+use crate::prelude::*;
+use crate::FormatNodeRule;
+use ruff_formatter::{
+    format_args, write, FormatOwnedWithRule, FormatRefWithRule, FormatRuleWithOptions,
+};
+use ruff_python_ast::node::AstNode;
+use rustpython_parser::ast::{
+    Constant, Expr, ExprAttribute, ExprBinOp, ExprConstant, ExprUnaryOp, Operator, Unaryop,
+};
 
 #[derive(Default)]
-pub struct FormatExprBinOp;
+pub struct FormatExprBinOp {
+    parentheses: Option<Parentheses>,
+}
+
+impl FormatRuleWithOptions<ExprBinOp, PyFormatContext<'_>> for FormatExprBinOp {
+    type Options = Option<Parentheses>;
+
+    fn with_options(mut self, options: Self::Options) -> Self {
+        self.parentheses = options;
+        self
+    }
+}
 
 impl FormatNodeRule<ExprBinOp> for FormatExprBinOp {
     fn fmt_fields(&self, item: &ExprBinOp, f: &mut PyFormatter) -> FormatResult<()> {
-        write!(f, [verbatim_text(item.range)])
+        let ExprBinOp {
+            left,
+            right,
+            op,
+            range: _,
+        } = item;
+
+        let should_break_right = self.parentheses == Some(Parentheses::Custom);
+
+        if should_break_right {
+            let left = left.format().memoized();
+            let right = right.format().memoized();
+
+            write!(
+                f,
+                [best_fitting![
+                    // The whole expression on a single line
+                    format_args![left, space(), op.format(), space(), right],
+                    // Break the right, but keep the left flat
+                    format_args![
+                        left,
+                        space(),
+                        op.format(),
+                        space(),
+                        group(&right).should_expand(true),
+                    ],
+                    // Break after the operator, try to keep the right flat, otherwise expand it
+                    format_args![
+                        text("("),
+                        block_indent(&format_args![
+                            left,
+                            hard_line_break(),
+                            op.format(),
+                            space(),
+                            group(&right),
+                        ]),
+                        text(")")
+                    ],
+                ]]
+            )
+        } else {
+            let comments = f.context().comments().clone();
+            let operator_comments = comments.dangling_comments(item.as_any_node_ref());
+            let needs_space = !is_simple_power_expression(item);
+
+            let before_operator_space = if needs_space {
+                soft_line_break_or_space()
+            } else {
+                soft_line_break()
+            };
+
+            write!(
+                f,
+                [
+                    left.format(),
+                    before_operator_space,
+                    op.format(),
+                    trailing_comments(operator_comments),
+                ]
+            )?;
+
+            // Format the operator on its own line if the operator has trailing comments and the right side has leading comments.
+            if !operator_comments.is_empty() && comments.has_leading_comments(right.as_ref().into())
+            {
+                write!(f, [hard_line_break()])?;
+            } else if needs_space {
+                write!(f, [space()])?;
+            }
+
+            write!(f, [group(&right.format())])
+        }
+    }
+
+    fn fmt_dangling_comments(&self, _node: &ExprBinOp, _f: &mut PyFormatter) -> FormatResult<()> {
+        // Handled inside of `fmt_fields`
+        Ok(())
+    }
+}
+
+const fn is_simple_power_expression(expr: &ExprBinOp) -> bool {
+    expr.op.is_pow() && is_simple_power_operand(&expr.left) && is_simple_power_operand(&expr.right)
+}
+
+/// Return `true` if an [`Expr`] adheres to [Black's definition](https://black.readthedocs.io/en/stable/the_black_code_style/current_style.html#line-breaks-binary-operators)
+/// of a non-complex expression, in the context of a power operation.
+const fn is_simple_power_operand(expr: &Expr) -> bool {
+    match expr {
+        Expr::UnaryOp(ExprUnaryOp {
+            op: Unaryop::Not, ..
+        }) => false,
+        Expr::Constant(ExprConstant {
+            value: Constant::Complex { .. } | Constant::Float(_) | Constant::Int(_),
+            ..
+        }) => true,
+        Expr::Name(_) => true,
+        Expr::UnaryOp(ExprUnaryOp { operand, .. }) => is_simple_power_operand(operand),
+        Expr::Attribute(ExprAttribute { value, .. }) => is_simple_power_operand(value),
+        _ => false,
+    }
+}
+
+#[derive(Copy, Clone)]
+pub struct FormatOperator;
+
+impl<'ast> AsFormat<PyFormatContext<'ast>> for Operator {
+    type Format<'a> = FormatRefWithRule<'a, Operator, FormatOperator, PyFormatContext<'ast>>;
+
+    fn format(&self) -> Self::Format<'_> {
+        FormatRefWithRule::new(self, FormatOperator)
+    }
+}
+
+impl<'ast> IntoFormat<PyFormatContext<'ast>> for Operator {
+    type Format = FormatOwnedWithRule<Operator, FormatOperator, PyFormatContext<'ast>>;
+    fn into_format(self) -> Self::Format {
+        FormatOwnedWithRule::new(self, FormatOperator)
+    }
+}
+
+impl FormatRule<Operator, PyFormatContext<'_>> for FormatOperator {
+    fn fmt(&self, item: &Operator, f: &mut Formatter<PyFormatContext<'_>>) -> FormatResult<()> {
+        let operator = match item {
+            Operator::Add => "+",
+            Operator::Sub => "-",
+            Operator::Mult => "*",
+            Operator::MatMult => "@",
+            Operator::Div => "/",
+            Operator::Mod => "%",
+            Operator::Pow => "**",
+            Operator::LShift => "<<",
+            Operator::RShift => ">>",
+            Operator::BitOr => "|",
+            Operator::BitXor => "^",
+            Operator::BitAnd => "&",
+            Operator::FloorDiv => "//",
+        };
+
+        text(operator).fmt(f)
+    }
+}
+
+impl NeedsParentheses for ExprBinOp {
+    fn needs_parentheses(&self, parenthesize: Parenthesize, source: &str) -> Parentheses {
+        match default_expression_needs_parentheses(self.into(), parenthesize, source) {
+            Parentheses::Optional => {
+                if should_binary_break_right_side_first(self) {
+                    Parentheses::Custom
+                } else {
+                    Parentheses::Optional
+                }
+            }
+            parentheses => parentheses,
+        }
+    }
+}
+
+pub(super) fn should_binary_break_right_side_first(expr: &ExprBinOp) -> bool {
+    use ruff_python_ast::prelude::*;
+
+    if expr.left.is_bin_op_expr() {
+        false
+    } else {
+        match expr.right.as_ref() {
+            Expr::Tuple(ExprTuple {
+                elts: expressions, ..
+            })
+            | Expr::List(ExprList {
+                elts: expressions, ..
+            })
+            | Expr::Set(ExprSet {
+                elts: expressions, ..
+            })
+            | Expr::Dict(ExprDict {
+                values: expressions,
+                ..
+            }) => !expressions.is_empty(),
+            Expr::Call(ExprCall { args, keywords, .. }) => !args.is_empty() && !keywords.is_empty(),
+            Expr::ListComp(_) | Expr::SetComp(_) | Expr::DictComp(_) | Expr::GeneratorExp(_) => {
+                true
+            }
+            _ => false,
+        }
     }
 }

crates/ruff_python_formatter/src/expression/expr_bool_op.rs

@@ -1,3 +1,6 @@
+use crate::expression::parentheses::{
+    default_expression_needs_parentheses, NeedsParentheses, Parentheses, Parenthesize,
+};
 use crate::{verbatim_text, FormatNodeRule, PyFormatter};
 use ruff_formatter::{write, Buffer, FormatResult};
 use rustpython_parser::ast::ExprBoolOp;
@@ -10,3 +13,9 @@ impl FormatNodeRule<ExprBoolOp> for FormatExprBoolOp {
         write!(f, [verbatim_text(item.range)])
     }
 }
+
+impl NeedsParentheses for ExprBoolOp {
+    fn needs_parentheses(&self, parenthesize: Parenthesize, source: &str) -> Parentheses {
+        default_expression_needs_parentheses(self.into(), parenthesize, source)
+    }
+}

crates/ruff_python_formatter/src/expression/expr_call.rs

@@ -1,3 +1,6 @@
+use crate::expression::parentheses::{
+    default_expression_needs_parentheses, NeedsParentheses, Parentheses, Parenthesize,
+};
 use crate::{verbatim_text, FormatNodeRule, PyFormatter};
 use ruff_formatter::{write, Buffer, FormatResult};
 use rustpython_parser::ast::ExprCall;
@@ -10,3 +13,12 @@ impl FormatNodeRule<ExprCall> for FormatExprCall {
         write!(f, [verbatim_text(item.range)])
     }
 }
+
+impl NeedsParentheses for ExprCall {
+    fn needs_parentheses(&self, parenthesize: Parenthesize, source: &str) -> Parentheses {
+        match default_expression_needs_parentheses(self.into(), parenthesize, source) {
+            Parentheses::Optional => Parentheses::Never,
+            parentheses => parentheses,
+        }
+    }
+}

crates/ruff_python_formatter/src/expression/expr_compare.rs

@@ -1,3 +1,6 @@
+use crate::expression::parentheses::{
+    default_expression_needs_parentheses, NeedsParentheses, Parentheses, Parenthesize,
+};
 use crate::{verbatim_text, FormatNodeRule, PyFormatter};
 use ruff_formatter::{write, Buffer, FormatResult};
 use rustpython_parser::ast::ExprCompare;
@@ -10,3 +13,9 @@ impl FormatNodeRule<ExprCompare> for FormatExprCompare {
         write!(f, [verbatim_text(item.range)])
     }
 }
+
+impl NeedsParentheses for ExprCompare {
+    fn needs_parentheses(&self, parenthesize: Parenthesize, source: &str) -> Parentheses {
+        default_expression_needs_parentheses(self.into(), parenthesize, source)
+    }
+}

crates/ruff_python_formatter/src/expression/expr_constant.rs

@@ -1,3 +1,6 @@
+use crate::expression::parentheses::{
+    default_expression_needs_parentheses, NeedsParentheses, Parentheses, Parenthesize,
+};
 use crate::{verbatim_text, FormatNodeRule, PyFormatter};
 use ruff_formatter::{write, Buffer, FormatResult};
 use rustpython_parser::ast::ExprConstant;
@@ -10,3 +13,9 @@ impl FormatNodeRule<ExprConstant> for FormatExprConstant {
         write!(f, [verbatim_text(item.range)])
     }
 }
+
+impl NeedsParentheses for ExprConstant {
+    fn needs_parentheses(&self, parenthesize: Parenthesize, source: &str) -> Parentheses {
+        default_expression_needs_parentheses(self.into(), parenthesize, source)
+    }
+}

crates/ruff_python_formatter/src/expression/expr_dict.rs

@@ -1,3 +1,6 @@
+use crate::expression::parentheses::{
+    default_expression_needs_parentheses, NeedsParentheses, Parentheses, Parenthesize,
+};
 use crate::{verbatim_text, FormatNodeRule, PyFormatter};
 use ruff_formatter::{write, Buffer, FormatResult};
 use rustpython_parser::ast::ExprDict;
@@ -10,3 +13,12 @@ impl FormatNodeRule<ExprDict> for FormatExprDict {
         write!(f, [verbatim_text(item.range)])
     }
 }
+
+impl NeedsParentheses for ExprDict {
+    fn needs_parentheses(&self, parenthesize: Parenthesize, source: &str) -> Parentheses {
+        match default_expression_needs_parentheses(self.into(), parenthesize, source) {
+            Parentheses::Optional => Parentheses::Never,
+            parentheses => parentheses,
+        }
+    }
+}

crates/ruff_python_formatter/src/expression/expr_dict_comp.rs

@@ -1,3 +1,6 @@
+use crate::expression::parentheses::{
+    default_expression_needs_parentheses, NeedsParentheses, Parentheses, Parenthesize,
+};
 use crate::{verbatim_text, FormatNodeRule, PyFormatter};
 use ruff_formatter::{write, Buffer, FormatResult};
 use rustpython_parser::ast::ExprDictComp;
@@ -10,3 +13,12 @@ impl FormatNodeRule<ExprDictComp> for FormatExprDictComp {
         write!(f, [verbatim_text(item.range)])
     }
 }
+
+impl NeedsParentheses for ExprDictComp {
+    fn needs_parentheses(&self, parenthesize: Parenthesize, source: &str) -> Parentheses {
+        match default_expression_needs_parentheses(self.into(), parenthesize, source) {
+            Parentheses::Optional => Parentheses::Never,
+            parentheses => parentheses,
+        }
+    }
+}

crates/ruff_python_formatter/src/expression/expr_formatted_value.rs

@@ -1,3 +1,6 @@
+use crate::expression::parentheses::{
+    default_expression_needs_parentheses, NeedsParentheses, Parentheses, Parenthesize,
+};
 use crate::{verbatim_text, FormatNodeRule, PyFormatter};
 use ruff_formatter::{write, Buffer, FormatResult};
 use rustpython_parser::ast::ExprFormattedValue;
@@ -10,3 +13,9 @@ impl FormatNodeRule<ExprFormattedValue> for FormatExprFormattedValue {
         write!(f, [verbatim_text(item.range)])
     }
 }
+
+impl NeedsParentheses for ExprFormattedValue {
+    fn needs_parentheses(&self, parenthesize: Parenthesize, source: &str) -> Parentheses {
+        default_expression_needs_parentheses(self.into(), parenthesize, source)
+    }
+}

crates/ruff_python_formatter/src/expression/expr_generator_exp.rs

@@ -1,3 +1,6 @@
+use crate::expression::parentheses::{
+    default_expression_needs_parentheses, NeedsParentheses, Parentheses, Parenthesize,
+};
 use crate::{verbatim_text, FormatNodeRule, PyFormatter};
 use ruff_formatter::{write, Buffer, FormatResult};
 use rustpython_parser::ast::ExprGeneratorExp;
@@ -10,3 +13,12 @@ impl FormatNodeRule<ExprGeneratorExp> for FormatExprGeneratorExp {
         write!(f, [verbatim_text(item.range)])
     }
 }
+
+impl NeedsParentheses for ExprGeneratorExp {
+    fn needs_parentheses(&self, parenthesize: Parenthesize, source: &str) -> Parentheses {
+        match default_expression_needs_parentheses(self.into(), parenthesize, source) {
+            Parentheses::Optional => Parentheses::Never,
+            parentheses => parentheses,
+        }
+    }
+}

crates/ruff_python_formatter/src/expression/expr_if_exp.rs

@@ -1,3 +1,6 @@
+use crate::expression::parentheses::{
+    default_expression_needs_parentheses, NeedsParentheses, Parentheses, Parenthesize,
+};
 use crate::{verbatim_text, FormatNodeRule, PyFormatter};
 use ruff_formatter::{write, Buffer, FormatResult};
 use rustpython_parser::ast::ExprIfExp;
@@ -10,3 +13,9 @@ impl FormatNodeRule<ExprIfExp> for FormatExprIfExp {
         write!(f, [verbatim_text(item.range)])
     }
 }
+
+impl NeedsParentheses for ExprIfExp {
+    fn needs_parentheses(&self, parenthesize: Parenthesize, source: &str) -> Parentheses {
+        default_expression_needs_parentheses(self.into(), parenthesize, source)
+    }
+}

crates/ruff_python_formatter/src/expression/expr_joined_str.rs

@@ -1,3 +1,6 @@
+use crate::expression::parentheses::{
+    default_expression_needs_parentheses, NeedsParentheses, Parentheses, Parenthesize,
+};
 use crate::{verbatim_text, FormatNodeRule, PyFormatter};
 use ruff_formatter::{write, Buffer, FormatResult};
 use rustpython_parser::ast::ExprJoinedStr;
@@ -10,3 +13,9 @@ impl FormatNodeRule<ExprJoinedStr> for FormatExprJoinedStr {
         write!(f, [verbatim_text(item.range)])
     }
 }
+
+impl NeedsParentheses for ExprJoinedStr {
+    fn needs_parentheses(&self, parenthesize: Parenthesize, source: &str) -> Parentheses {
+        default_expression_needs_parentheses(self.into(), parenthesize, source)
+    }
+}

crates/ruff_python_formatter/src/expression/expr_lambda.rs

@@ -1,3 +1,6 @@
+use crate::expression::parentheses::{
+    default_expression_needs_parentheses, NeedsParentheses, Parentheses, Parenthesize,
+};
 use crate::{verbatim_text, FormatNodeRule, PyFormatter};
 use ruff_formatter::{write, Buffer, FormatResult};
 use rustpython_parser::ast::ExprLambda;
@@ -10,3 +13,9 @@ impl FormatNodeRule<ExprLambda> for FormatExprLambda {
         write!(f, [verbatim_text(item.range)])
     }
 }
+
+impl NeedsParentheses for ExprLambda {
+    fn needs_parentheses(&self, parenthesize: Parenthesize, source: &str) -> Parentheses {
+        default_expression_needs_parentheses(self.into(), parenthesize, source)
+    }
+}

crates/ruff_python_formatter/src/expression/expr_list.rs

@@ -1,5 +1,10 @@
-use crate::{verbatim_text, FormatNodeRule, PyFormatter};
-use ruff_formatter::{write, Buffer, FormatResult};
+use crate::comments::dangling_comments;
+use crate::expression::parentheses::{
+    default_expression_needs_parentheses, NeedsParentheses, Parentheses, Parenthesize,
+};
+use crate::prelude::*;
+use crate::FormatNodeRule;
+use ruff_formatter::{format_args, write};
 use rustpython_parser::ast::ExprList;
 
 #[derive(Default)]
@@ -7,6 +12,55 @@ pub struct FormatExprList;
 
 impl FormatNodeRule<ExprList> for FormatExprList {
     fn fmt_fields(&self, item: &ExprList, f: &mut PyFormatter) -> FormatResult<()> {
-        write!(f, [verbatim_text(item.range)])
+        let ExprList {
+            range: _,
+            elts,
+            ctx: _,
+        } = item;
+
+        let items = format_with(|f| {
+            let mut iter = elts.iter();
+
+            if let Some(first) = iter.next() {
+                write!(f, [first.format()])?;
+            }
+
+            for item in iter {
+                write!(f, [text(","), soft_line_break_or_space(), item.format()])?;
+            }
+
+            if !elts.is_empty() {
+                write!(f, [if_group_breaks(&text(","))])?;
+            }
+
+            Ok(())
+        });
+
+        let comments = f.context().comments().clone();
+        let dangling = comments.dangling_comments(item.into());
+
+        write!(
+            f,
+            [group(&format_args![
+                text("["),
+                dangling_comments(dangling),
+                soft_block_indent(&items),
+                text("]")
+            ])]
+        )
+    }
+
+    fn fmt_dangling_comments(&self, _node: &ExprList, _f: &mut PyFormatter) -> FormatResult<()> {
+        // Handled as part of `fmt_fields`
+        Ok(())
+    }
+}
+
+impl NeedsParentheses for ExprList {
+    fn needs_parentheses(&self, parenthesize: Parenthesize, source: &str) -> Parentheses {
+        match default_expression_needs_parentheses(self.into(), parenthesize, source) {
+            Parentheses::Optional => Parentheses::Never,
+            parentheses => parentheses,
+        }
     }
 }

crates/ruff_python_formatter/src/expression/expr_list_comp.rs

@@ -1,3 +1,6 @@
+use crate::expression::parentheses::{
+    default_expression_needs_parentheses, NeedsParentheses, Parentheses, Parenthesize,
+};
 use crate::{verbatim_text, FormatNodeRule, PyFormatter};
 use ruff_formatter::{write, Buffer, FormatResult};
 use rustpython_parser::ast::ExprListComp;
@@ -10,3 +13,12 @@ impl FormatNodeRule<ExprListComp> for FormatExprListComp {
         write!(f, [verbatim_text(item.range)])
     }
 }
+
+impl NeedsParentheses for ExprListComp {
+    fn needs_parentheses(&self, parenthesize: Parenthesize, source: &str) -> Parentheses {
+        match default_expression_needs_parentheses(self.into(), parenthesize, source) {
+            Parentheses::Optional => Parentheses::Never,
+            parentheses => parentheses,
+        }
+    }
+}

crates/ruff_python_formatter/src/expression/expr_name.rs

@@ -1,3 +1,6 @@
+use crate::expression::parentheses::{
+    default_expression_needs_parentheses, NeedsParentheses, Parentheses, Parenthesize,
+};
 use crate::prelude::*;
 use crate::FormatNodeRule;
 use ruff_formatter::{write, FormatContext};
@@ -22,6 +25,12 @@ impl FormatNodeRule<ExprName> for FormatExprName {
     }
 }
 
+impl NeedsParentheses for ExprName {
+    fn needs_parentheses(&self, parenthesize: Parenthesize, source: &str) -> Parentheses {
+        default_expression_needs_parentheses(self.into(), parenthesize, source)
+    }
+}
+
 #[cfg(test)]
 mod tests {
     use ruff_text_size::{TextRange, TextSize};

crates/ruff_python_formatter/src/expression/expr_named_expr.rs

@@ -1,3 +1,6 @@
+use crate::expression::parentheses::{
+    default_expression_needs_parentheses, NeedsParentheses, Parentheses, Parenthesize,
+};
 use crate::{verbatim_text, FormatNodeRule, PyFormatter};
 use ruff_formatter::{write, Buffer, FormatResult};
 use rustpython_parser::ast::ExprNamedExpr;
@@ -10,3 +13,9 @@ impl FormatNodeRule<ExprNamedExpr> for FormatExprNamedExpr {
         write!(f, [verbatim_text(item.range)])
     }
 }
+
+impl NeedsParentheses for ExprNamedExpr {
+    fn needs_parentheses(&self, parenthesize: Parenthesize, source: &str) -> Parentheses {
+        default_expression_needs_parentheses(self.into(), parenthesize, source)
+    }
+}

crates/ruff_python_formatter/src/expression/expr_set.rs

@@ -1,3 +1,6 @@
+use crate::expression::parentheses::{
+    default_expression_needs_parentheses, NeedsParentheses, Parentheses, Parenthesize,
+};
 use crate::{verbatim_text, FormatNodeRule, PyFormatter};
 use ruff_formatter::{write, Buffer, FormatResult};
 use rustpython_parser::ast::ExprSet;
@@ -10,3 +13,12 @@ impl FormatNodeRule<ExprSet> for FormatExprSet {
         write!(f, [verbatim_text(item.range)])
     }
 }
+
+impl NeedsParentheses for ExprSet {
+    fn needs_parentheses(&self, parenthesize: Parenthesize, source: &str) -> Parentheses {
+        match default_expression_needs_parentheses(self.into(), parenthesize, source) {
+            Parentheses::Optional => Parentheses::Never,
+            parentheses => parentheses,
+        }
+    }
+}

crates/ruff_python_formatter/src/expression/expr_set_comp.rs

@@ -1,3 +1,6 @@
+use crate::expression::parentheses::{
+    default_expression_needs_parentheses, NeedsParentheses, Parentheses, Parenthesize,
+};
 use crate::{verbatim_text, FormatNodeRule, PyFormatter};
 use ruff_formatter::{write, Buffer, FormatResult};
 use rustpython_parser::ast::ExprSetComp;
@@ -10,3 +13,12 @@ impl FormatNodeRule<ExprSetComp> for FormatExprSetComp {
         write!(f, [verbatim_text(item.range)])
     }
 }
+
+impl NeedsParentheses for ExprSetComp {
+    fn needs_parentheses(&self, parenthesize: Parenthesize, source: &str) -> Parentheses {
+        match default_expression_needs_parentheses(self.into(), parenthesize, source) {
+            Parentheses::Optional => Parentheses::Never,
+            parentheses => parentheses,
+        }
+    }
+}

crates/ruff_python_formatter/src/expression/expr_slice.rs

@@ -1,3 +1,6 @@
+use crate::expression::parentheses::{
+    default_expression_needs_parentheses, NeedsParentheses, Parentheses, Parenthesize,
+};
 use crate::{verbatim_text, FormatNodeRule, PyFormatter};
 use ruff_formatter::{write, Buffer, FormatResult};
 use rustpython_parser::ast::ExprSlice;
@@ -10,3 +13,9 @@ impl FormatNodeRule<ExprSlice> for FormatExprSlice {
         write!(f, [verbatim_text(item.range)])
     }
 }
+
+impl NeedsParentheses for ExprSlice {
+    fn needs_parentheses(&self, parenthesize: Parenthesize, source: &str) -> Parentheses {
+        default_expression_needs_parentheses(self.into(), parenthesize, source)
+    }
+}

crates/ruff_python_formatter/src/expression/expr_starred.rs

@@ -1,3 +1,6 @@
+use crate::expression::parentheses::{
+    default_expression_needs_parentheses, NeedsParentheses, Parentheses, Parenthesize,
+};
 use crate::{verbatim_text, FormatNodeRule, PyFormatter};
 use ruff_formatter::{write, Buffer, FormatResult};
 use rustpython_parser::ast::ExprStarred;
@@ -10,3 +13,9 @@ impl FormatNodeRule<ExprStarred> for FormatExprStarred {
         write!(f, [verbatim_text(item.range)])
     }
 }
+
+impl NeedsParentheses for ExprStarred {
+    fn needs_parentheses(&self, parenthesize: Parenthesize, source: &str) -> Parentheses {
+        default_expression_needs_parentheses(self.into(), parenthesize, source)
+    }
+}

crates/ruff_python_formatter/src/expression/expr_subscript.rs

@@ -1,3 +1,6 @@
+use crate::expression::parentheses::{
+    default_expression_needs_parentheses, NeedsParentheses, Parentheses, Parenthesize,
+};
 use crate::{verbatim_text, FormatNodeRule, PyFormatter};
 use ruff_formatter::{write, Buffer, FormatResult};
 use rustpython_parser::ast::ExprSubscript;
@@ -10,3 +13,9 @@ impl FormatNodeRule<ExprSubscript> for FormatExprSubscript {
         write!(f, [verbatim_text(item.range)])
     }
 }
+
+impl NeedsParentheses for ExprSubscript {
+    fn needs_parentheses(&self, parenthesize: Parenthesize, source: &str) -> Parentheses {
+        default_expression_needs_parentheses(self.into(), parenthesize, source)
+    }
+}

crates/ruff_python_formatter/src/expression/expr_tuple.rs

@@ -1,3 +1,6 @@
+use crate::expression::parentheses::{
+    default_expression_needs_parentheses, NeedsParentheses, Parentheses, Parenthesize,
+};
 use crate::{verbatim_text, FormatNodeRule, PyFormatter};
 use ruff_formatter::{write, Buffer, FormatResult};
 use rustpython_parser::ast::ExprTuple;
@@ -10,3 +13,12 @@ impl FormatNodeRule<ExprTuple> for FormatExprTuple {
         write!(f, [verbatim_text(item.range)])
     }
 }
+
+impl NeedsParentheses for ExprTuple {
+    fn needs_parentheses(&self, parenthesize: Parenthesize, source: &str) -> Parentheses {
+        match default_expression_needs_parentheses(self.into(), parenthesize, source) {
+            Parentheses::Optional => Parentheses::Never,
+            parentheses => parentheses,
+        }
+    }
+}

crates/ruff_python_formatter/src/expression/expr_unary_op.rs

@@ -1,3 +1,6 @@
+use crate::expression::parentheses::{
+    default_expression_needs_parentheses, NeedsParentheses, Parentheses, Parenthesize,
+};
 use crate::{verbatim_text, FormatNodeRule, PyFormatter};
 use ruff_formatter::{write, Buffer, FormatResult};
 use rustpython_parser::ast::ExprUnaryOp;
@@ -10,3 +13,9 @@ impl FormatNodeRule<ExprUnaryOp> for FormatExprUnaryOp {
         write!(f, [verbatim_text(item.range)])
     }
 }
+
+impl NeedsParentheses for ExprUnaryOp {
+    fn needs_parentheses(&self, parenthesize: Parenthesize, source: &str) -> Parentheses {
+        default_expression_needs_parentheses(self.into(), parenthesize, source)
+    }
+}

crates/ruff_python_formatter/src/expression/expr_yield.rs

@@ -1,3 +1,6 @@
+use crate::expression::parentheses::{
+    default_expression_needs_parentheses, NeedsParentheses, Parentheses, Parenthesize,
+};
 use crate::{verbatim_text, FormatNodeRule, PyFormatter};
 use ruff_formatter::{write, Buffer, FormatResult};
 use rustpython_parser::ast::ExprYield;
@@ -10,3 +13,9 @@ impl FormatNodeRule<ExprYield> for FormatExprYield {
         write!(f, [verbatim_text(item.range)])
     }
 }
+
+impl NeedsParentheses for ExprYield {
+    fn needs_parentheses(&self, parenthesize: Parenthesize, source: &str) -> Parentheses {
+        default_expression_needs_parentheses(self.into(), parenthesize, source)
+    }
+}

crates/ruff_python_formatter/src/expression/expr_yield_from.rs

@@ -1,3 +1,6 @@
+use crate::expression::parentheses::{
+    default_expression_needs_parentheses, NeedsParentheses, Parentheses, Parenthesize,
+};
 use crate::{verbatim_text, FormatNodeRule, PyFormatter};
 use ruff_formatter::{write, Buffer, FormatResult};
 use rustpython_parser::ast::ExprYieldFrom;
@@ -10,3 +13,9 @@ impl FormatNodeRule<ExprYieldFrom> for FormatExprYieldFrom {
         write!(f, [verbatim_text(item.range)])
     }
 }
+
+impl NeedsParentheses for ExprYieldFrom {
+    fn needs_parentheses(&self, parenthesize: Parenthesize, source: &str) -> Parentheses {
+        default_expression_needs_parentheses(self.into(), parenthesize, source)
+    }
+}

crates/ruff_python_formatter/src/expression/maybe_parenthesize.rs

@@ -1,53 +0,0 @@
-use crate::context::NodeLevel;
-use crate::prelude::*;
-use ruff_formatter::{format_args, write};
-use rustpython_parser::ast::Expr;
-
-/// Formats the passed expression. Adds parentheses if the expression doesn't fit on a line.
-pub(crate) const fn maybe_parenthesize(expression: &Expr) -> MaybeParenthesize {
-    MaybeParenthesize { expression }
-}
-
-pub(crate) struct MaybeParenthesize<'a> {
-    expression: &'a Expr,
-}
-
-impl Format<PyFormatContext<'_>> for MaybeParenthesize<'_> {
-    fn fmt(&self, f: &mut Formatter<PyFormatContext<'_>>) -> FormatResult<()> {
-        let saved_level = f.context().node_level();
-        f.context_mut().set_node_level(NodeLevel::Parenthesized);
-
-        let result = if needs_parentheses(self.expression) {
-            write!(
-                f,
-                [group(&format_args![
-                    if_group_breaks(&text("(")),
-                    soft_block_indent(&self.expression.format()),
-                    if_group_breaks(&text(")"))
-                ])]
-            )
-        } else {
-            // Don't add parentheses around expressions that  have parentheses on their own (e.g. list, dict, tuple, call expression)
-            self.expression.format().fmt(f)
-        };
-
-        f.context_mut().set_node_level(saved_level);
-
-        result
-    }
-}
-
-const fn needs_parentheses(expr: &Expr) -> bool {
-    !matches!(
-        expr,
-        Expr::Tuple(_)
-            | Expr::List(_)
-            | Expr::Set(_)
-            | Expr::Dict(_)
-            | Expr::ListComp(_)
-            | Expr::SetComp(_)
-            | Expr::DictComp(_)
-            | Expr::GeneratorExp(_)
-            | Expr::Call(_)
-    )
-}

crates/ruff_python_formatter/src/expression/mod.rs

@@ -1,5 +1,9 @@
+use crate::context::NodeLevel;
+use crate::expression::parentheses::{NeedsParentheses, Parentheses, Parenthesize};
 use crate::prelude::*;
-use ruff_formatter::{FormatOwnedWithRule, FormatRefWithRule, FormatRule};
+use ruff_formatter::{
+    format_args, write, FormatOwnedWithRule, FormatRefWithRule, FormatRule, FormatRuleWithOptions,
+};
 use rustpython_parser::ast::Expr;
 
 pub(crate) mod expr_attribute;
@@ -29,17 +33,30 @@ pub(crate) mod expr_tuple;
 pub(crate) mod expr_unary_op;
 pub(crate) mod expr_yield;
 pub(crate) mod expr_yield_from;
-pub(crate) mod maybe_parenthesize;
+pub(crate) mod parentheses;
 
 #[derive(Default)]
-pub struct FormatExpr;
+pub struct FormatExpr {
+    parenthesize: Parenthesize,
+}
+
+impl FormatRuleWithOptions<Expr, PyFormatContext<'_>> for FormatExpr {
+    type Options = Parenthesize;
+
+    fn with_options(mut self, options: Self::Options) -> Self {
+        self.parenthesize = options;
+        self
+    }
+}
 
 impl FormatRule<Expr, PyFormatContext<'_>> for FormatExpr {
     fn fmt(&self, item: &Expr, f: &mut PyFormatter) -> FormatResult<()> {
-        match item {
+        let parentheses = item.needs_parentheses(self.parenthesize, f.context().contents());
+
+        let format_expr = format_with(|f| match item {
             Expr::BoolOp(expr) => expr.format().fmt(f),
             Expr::NamedExpr(expr) => expr.format().fmt(f),
-            Expr::BinOp(expr) => expr.format().fmt(f),
+            Expr::BinOp(expr) => expr.format().with_options(Some(parentheses)).fmt(f),
             Expr::UnaryOp(expr) => expr.format().fmt(f),
             Expr::Lambda(expr) => expr.format().fmt(f),
             Expr::IfExp(expr) => expr.format().fmt(f),
@@ -64,6 +81,72 @@ impl FormatRule<Expr, PyFormatContext<'_>> for FormatExpr {
             Expr::List(expr) => expr.format().fmt(f),
             Expr::Tuple(expr) => expr.format().fmt(f),
             Expr::Slice(expr) => expr.format().fmt(f),
+        });
+
+        let saved_level = f.context().node_level();
+        f.context_mut().set_node_level(NodeLevel::Expression);
+
+        let result = match parentheses {
+            Parentheses::Always => {
+                write!(
+                    f,
+                    [group(&format_args![
+                        text("("),
+                        soft_block_indent(&format_expr),
+                        text(")")
+                    ])]
+                )
+            }
+            // Add optional parentheses. Ignore if the item renders parentheses itself.
+            Parentheses::Optional => {
+                write!(
+                    f,
+                    [group(&format_args![
+                        if_group_breaks(&text("(")),
+                        soft_block_indent(&format_expr),
+                        if_group_breaks(&text(")"))
+                    ])]
+                )
+            }
+            Parentheses::Custom | Parentheses::Never => Format::fmt(&format_expr, f),
+        };
+
+        f.context_mut().set_node_level(saved_level);
+
+        result
+    }
+}
+
+impl NeedsParentheses for Expr {
+    fn needs_parentheses(&self, parenthesize: Parenthesize, source: &str) -> Parentheses {
+        match self {
+            Expr::BoolOp(expr) => expr.needs_parentheses(parenthesize, source),
+            Expr::NamedExpr(expr) => expr.needs_parentheses(parenthesize, source),
+            Expr::BinOp(expr) => expr.needs_parentheses(parenthesize, source),
+            Expr::UnaryOp(expr) => expr.needs_parentheses(parenthesize, source),
+            Expr::Lambda(expr) => expr.needs_parentheses(parenthesize, source),
+            Expr::IfExp(expr) => expr.needs_parentheses(parenthesize, source),
+            Expr::Dict(expr) => expr.needs_parentheses(parenthesize, source),
+            Expr::Set(expr) => expr.needs_parentheses(parenthesize, source),
+            Expr::ListComp(expr) => expr.needs_parentheses(parenthesize, source),
+            Expr::SetComp(expr) => expr.needs_parentheses(parenthesize, source),
+            Expr::DictComp(expr) => expr.needs_parentheses(parenthesize, source),
+            Expr::GeneratorExp(expr) => expr.needs_parentheses(parenthesize, source),
+            Expr::Await(expr) => expr.needs_parentheses(parenthesize, source),
+            Expr::Yield(expr) => expr.needs_parentheses(parenthesize, source),
+            Expr::YieldFrom(expr) => expr.needs_parentheses(parenthesize, source),
+            Expr::Compare(expr) => expr.needs_parentheses(parenthesize, source),
+            Expr::Call(expr) => expr.needs_parentheses(parenthesize, source),
+            Expr::FormattedValue(expr) => expr.needs_parentheses(parenthesize, source),
+            Expr::JoinedStr(expr) => expr.needs_parentheses(parenthesize, source),
+            Expr::Constant(expr) => expr.needs_parentheses(parenthesize, source),
+            Expr::Attribute(expr) => expr.needs_parentheses(parenthesize, source),
+            Expr::Subscript(expr) => expr.needs_parentheses(parenthesize, source),
+            Expr::Starred(expr) => expr.needs_parentheses(parenthesize, source),
+            Expr::Name(expr) => expr.needs_parentheses(parenthesize, source),
+            Expr::List(expr) => expr.needs_parentheses(parenthesize, source),
+            Expr::Tuple(expr) => expr.needs_parentheses(parenthesize, source),
+            Expr::Slice(expr) => expr.needs_parentheses(parenthesize, source),
         }
     }
 }

crates/ruff_python_formatter/src/expression/parentheses.rs

@@ -0,0 +1,93 @@
+use crate::trivia::{
+    find_first_non_trivia_character_after, find_first_non_trivia_character_before,
+};
+use ruff_python_ast::node::AnyNodeRef;
+
+pub(crate) trait NeedsParentheses {
+    fn needs_parentheses(&self, parenthesize: Parenthesize, source: &str) -> Parentheses;
+}
+
+pub(super) fn default_expression_needs_parentheses(
+    node: AnyNodeRef,
+    parenthesize: Parenthesize,
+    source: &str,
+) -> Parentheses {
+    debug_assert!(
+        node.is_expression(),
+        "Should only be called for expressions"
+    );
+
+    // `Optional` or `Preserve` and expression has parentheses in source code.
+    if !parenthesize.is_if_breaks() && is_expression_parenthesized(node, source) {
+        Parentheses::Always
+    }
+    // `Optional` or `IfBreaks`: Add parentheses if the expression doesn't fit on a line
+    else if !parenthesize.is_preserve() {
+        Parentheses::Optional
+    } else {
+        //`Preserve` and expression has no parentheses in the source code
+        Parentheses::Never
+    }
+}
+
+/// Configures if the expression should be parenthesized.
+#[derive(Copy, Clone, Debug, Default)]
+pub enum Parenthesize {
+    /// Parenthesize the expression if it has parenthesis in the source.
+    #[default]
+    Preserve,
+
+    /// Parenthesizes the expression if it doesn't fit on a line OR if the expression is parenthesized in the source code.
+    Optional,
+
+    /// Parenthesizes the expression only if it doesn't fit on a line.
+    IfBreaks,
+}
+
+impl Parenthesize {
+    const fn is_if_breaks(self) -> bool {
+        matches!(self, Parenthesize::IfBreaks)
+    }
+
+    const fn is_preserve(self) -> bool {
+        matches!(self, Parenthesize::Preserve)
+    }
+}
+
+/// Whether it is necessary to add parentheses around an expression.
+/// This is different from [`Parenthesize`] in that it is the resolved representation: It takes into account
+/// whether there are parentheses in the source code or not.
+#[derive(Copy, Clone, Debug, Eq, PartialEq)]
+pub enum Parentheses {
+    /// Always create parentheses
+    Always,
+
+    /// Only add parentheses when necessary because the expression breaks over multiple lines.
+    Optional,
+
+    /// Custom handling by the node's formatter implementation
+    Custom,
+
+    /// Never add parentheses
+    Never,
+}
+
+fn is_expression_parenthesized(expr: AnyNodeRef, contents: &str) -> bool {
+    use rustpython_parser::ast::Ranged;
+
+    debug_assert!(
+        expr.is_expression(),
+        "Should only be called for expressions"
+    );
+
+    // Search backwards to avoid ambiguity with `(a, )` and because it's faster
+    matches!(
+        find_first_non_trivia_character_after(expr.end(), contents),
+        Some((_, ')'))
+    )
+        // Search forwards to confirm that this is not a nested expression `(5 + d * 3)`
+        && matches!(
+        find_first_non_trivia_character_before(expr.start(), contents),
+        Some((_, '('))
+    )
+}

crates/ruff_python_formatter/src/lib.rs

@@ -51,7 +51,7 @@ where
         self.fmt_node(node, f)?;
         self.fmt_dangling_comments(node, f)?;
         self.fmt_trailing_comments(node, f)?;
-        write!(f, [source_position(node.start())])
+        write!(f, [source_position(node.end())])
     }
 
     /// Formats the node without comments. Ignores any suppression comments.
@@ -225,8 +225,11 @@ if    True:
 
         let formatted_code = printed.as_code();
 
-        let reformatted =
-            format_module(formatted_code).expect("Expected formatted code to be valid syntax");
+        let reformatted = format_module(formatted_code).unwrap_or_else(|err| {
+            panic!(
+                "Formatted code resulted introduced a syntax error {err:#?}. Code:\n{formatted_code}"
+            )
+        });
 
         if reformatted.as_code() != formatted_code {
             let diff = TextDiff::from_lines(formatted_code, reformatted.as_code())
@@ -314,7 +317,7 @@ Formatted twice:
         let formatted_code = printed.as_code();
 
         let reformatted =
-            format_module(formatted_code).expect("Expected formatted code to be valid syntax");
+            format_module(formatted_code).unwrap_or_else(|err| panic!("Expected formatted code to be valid syntax but it contains syntax errors: {err}\n{formatted_code}"));
 
         if reformatted.as_code() != formatted_code {
             let diff = TextDiff::from_lines(formatted_code, reformatted.as_code())
@@ -378,7 +381,9 @@ other
         // Uncomment the `dbg` to print the IR.
         // Use `dbg_write!(f, []) instead of `write!(f, [])` in your formatting code to print some IR
         // inside of a `Format` implementation
-        // dbg!(formatted.document());
+        // dbg!(formatted
+        //     .document()
+        //     .display(formatted.context().source_code()));
 
         let printed = formatted.print().unwrap();
 

crates/ruff_python_formatter/src/snapshots/ruff_python_formatter__tests__black_test__collections_py.snap

@@ -168,7 +168,7 @@ if True:
 -    2,
 -    3,
 -]
-+[1, 2, 3,]
++[1, 2, 3]
  
 -division_result_tuple = (6 / 2,)
 +division_result_tuple = (6/2,)
@@ -250,7 +250,7 @@ for x in (1,):
 for (x,) in (1,), (2,), (3,):
     pass
 
-[1, 2, 3,]
+[1, 2, 3]
 
 division_result_tuple = (6/2,)
 print("foo %r", (foo.bar,))

crates/ruff_python_formatter/src/snapshots/ruff_python_formatter__tests__black_test__comments_py.snap

@@ -109,7 +109,21 @@ async def wat():
 ```diff
 --- Black
 +++ Ruff
-@@ -19,14 +19,9 @@
+@@ -4,10 +4,12 @@
+ #
+ # Has many lines. Many, many lines.
+ # Many, many, many lines.
+-"""Module docstring.
++(
++    """Module docstring.
+ 
+ Possibly also many, many lines.
+ """
++)
+ 
+ import os.path
+ import sys
+@@ -19,9 +21,6 @@
      import fast
  except ImportError:
      import slow as fast
@@ -117,16 +131,9 @@ async def wat():
 -
 -# Some comment before a function.
  y = 1
--(
--    # some strings
--    y  # type: ignore
--)
-+# some strings
-+y  # type: ignore
- 
- 
- def function(default=None):
-@@ -93,4 +88,4 @@
+ (
+     # some strings
+@@ -93,4 +92,4 @@
  
  # Some closing comments.
  # Maybe Vim or Emacs directives for formatting.
@@ -144,10 +151,12 @@ async def wat():
 #
 # Has many lines. Many, many lines.
 # Many, many, many lines.
-"""Module docstring.
+(
+    """Module docstring.
 
 Possibly also many, many lines.
 """
+)
 
 import os.path
 import sys
@@ -160,8 +169,10 @@ try:
 except ImportError:
     import slow as fast
 y = 1
-# some strings
-y  # type: ignore
+(
+    # some strings
+    y  # type: ignore
+)
 
 
 def function(default=None):

crates/ruff_python_formatter/src/snapshots/ruff_python_formatter__tests__black_test__expression_py.snap

@@ -276,17 +276,7 @@ last_call()
  Name
  None
  True
-@@ -22,119 +23,83 @@
- v1 << 2
- 1 >> v2
- 1 % finished
--1 + v2 - v3 * 4 ^ 5**v6 / 7 // 8
--((1 + v2) - (v3 * 4)) ^ (((5**v6) / 7) // 8)
-+1 + v2 - v3 * 4 ^ 5 ** v6 / 7 // 8
-+((1 + v2) - (v3 * 4)) ^ (((5 ** v6) / 7) // 8)
- not great
- ~great
- +value
+@@ -30,56 +31,39 @@
  -1
  ~int and not v1 ^ 123 + v2 | True
  (~int) and (not ((v1 ^ (123 + v2)) | True))
@@ -312,14 +302,14 @@ last_call()
  (str or None) if True else (str or bytes or None)
  str or None if (1 if True else 2) else str or bytes or None
  (str or None) if (1 if True else 2) else (str or bytes or None)
--(
+ (
 -    (super_long_variable_name or None)
 -    if (1 if super_long_test_name else 2)
 -    else (str or bytes or None)
--)
++    (super_long_variable_name or None) if (1 if super_long_test_name else 2) else (str or bytes or None)
+ )
 -{"2.7": dead, "3.7": (long_live or die_hard)}
 -{"2.7": dead, "3.7": (long_live or die_hard), **{"3.6": verygood}}
-+(super_long_variable_name or None) if (1 if super_long_test_name else 2) else (str or bytes or None)
 +{'2.7': dead, '3.7': (long_live or die_hard)}
 +{'2.7': dead, '3.7': (long_live or die_hard), **{'3.6': verygood}}
  {**a, **b, **c}
@@ -338,7 +328,7 @@ last_call()
 -    2,
 -    3,
 -]
-+[1, 2, 3,]
++[1, 2, 3]
  [*a]
  [*range(10)]
 -[
@@ -351,15 +341,14 @@ last_call()
 -    *a,
 -    5,
 -]
--[
--    this_is_a_very_long_variable_which_will_force_a_delimiter_split,
--    element,
--    another,
--    *more,
--]
-+[*a, 4, 5,]
-+[4, *a, 5,]
-+[this_is_a_very_long_variable_which_will_force_a_delimiter_split, element, another, *more]
++[*a, 4, 5]
++[4, *a, 5]
+ [
+     this_is_a_very_long_variable_which_will_force_a_delimiter_split,
+     element,
+@@ -87,54 +71,44 @@
+     *more,
+ ]
  {i for i in (1, 2, 3)}
 -{(i**2) for i in (1, 2, 3)}
 -{(i**2) for i, _ in ((1, "a"), (2, "b"), (3, "c"))}
@@ -417,18 +406,21 @@ last_call()
  dict[str, int]
  tuple[str, ...]
 -tuple[str, int, float, dict[str, int]]
- tuple[
+-tuple[
 -    str,
 -    int,
 -    float,
 -    dict[str, int],
++(
++    tuple[
 +    str, int, float, dict[str, int]
  ]
++)
 +tuple[str, int, float, dict[str, int],]
  very_long_variable_name_filters: t.List[
      t.Tuple[str, t.Union[str, t.List[t.Optional[str]]]],
  ]
-@@ -144,9 +109,9 @@
+@@ -144,9 +118,9 @@
  xxxx_xxx_xxxx_xxxxx_xxxx_xxx: Callable[..., List[SomeClass]] = classmethod(  # type: ignore
      sync(async_xxxx_xxx_xxxx_xxxxx_xxxx_xxx.__func__)
  )
@@ -441,7 +433,7 @@ last_call()
  slice[0]
  slice[0:1]
  slice[0:1:2]
-@@ -174,57 +139,29 @@
+@@ -174,57 +148,29 @@
  numpy[:, ::-1]
  numpy[np.newaxis, :]
  (str or None) if (sys.version_info[0] > (3,)) else (str or bytes or None)
@@ -450,9 +442,8 @@ last_call()
 +{'2.7': dead, '3.7': long_live or die_hard}
 +{'2.7', '3.6', '3.7', '3.8', '3.9', '4.0' if gilectomy else '3.10'}
  [1, 2, 3, 4, 5, 6, 7, 8, 9, 10 or A, 11 or B, 12 or C]
--(SomeName)
+ (SomeName)
  SomeName
-+SomeName
  (Good, Bad, Ugly)
  (i for i in (1, 2, 3))
 -((i**2) for i in (1, 2, 3))
@@ -511,7 +502,7 @@ last_call()
  Ø = set()
  authors.łukasz.say_thanks()
  mapping = {
-@@ -237,134 +174,86 @@
+@@ -237,114 +183,80 @@
  
  def gen():
      yield from outside_of_generator
@@ -655,30 +646,7 @@ last_call()
 +    ~ aaaaaaaaaaaaaaaa.a + aaaaaaaaaaaaaaaa.b - aaaaaaaaaaaaaaaa.c * aaaaaaaaaaaaaaaa.d @ aaaaaaaaaaaaaaaa.e | aaaaaaaaaaaaaaaa.f & aaaaaaaaaaaaaaaa.g % aaaaaaaaaaaaaaaa.h ^ aaaaaaaaaaaaaaaa.i << aaaaaaaaaaaaaaaa.k >> aaaaaaaaaaaaaaaa.l ** aaaaaaaaaaaaaaaa.m // aaaaaaaaaaaaaaaa.n
  ):
      return True
--(
--    aaaaaaaaaaaaaaaa
--    + aaaaaaaaaaaaaaaa
--    - aaaaaaaaaaaaaaaa
--    * (aaaaaaaaaaaaaaaa + aaaaaaaaaaaaaaaa)
--    / (aaaaaaaaaaaaaaaa + aaaaaaaaaaaaaaaa + aaaaaaaaaaaaaaaa)
--)
-+aaaaaaaaaaaaaaaa + aaaaaaaaaaaaaaaa - aaaaaaaaaaaaaaaa * (aaaaaaaaaaaaaaaa + aaaaaaaaaaaaaaaa) / (aaaaaaaaaaaaaaaa + aaaaaaaaaaaaaaaa + aaaaaaaaaaaaaaaa)
- aaaaaaaaaaaaaaaa + aaaaaaaaaaaaaaaa
--(
--    aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa
--    >> aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa
--    << aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa
--)
-+aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa >> aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa << aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa
- bbbb >> bbbb * bbbb
--(
--    aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa
--    ^ bbbb.a & aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa
--    ^ aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa
--)
-+aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa ^bbbb.a & aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa^aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa
- last_call()
- # standalone comment at ENDMARKER
+ (
 ```
 
 ## Ruff Output
@@ -709,8 +677,8 @@ Name1 or Name2 and Name3 or Name4
 v1 << 2
 1 >> v2
 1 % finished
-1 + v2 - v3 * 4 ^ 5 ** v6 / 7 // 8
-((1 + v2) - (v3 * 4)) ^ (((5 ** v6) / 7) // 8)
+1 + v2 - v3 * 4 ^ 5**v6 / 7 // 8
+((1 + v2) - (v3 * 4)) ^ (((5**v6) / 7) // 8)
 not great
 ~great
 +value
@@ -731,7 +699,9 @@ str or None if True else str or bytes or None
 (str or None) if True else (str or bytes or None)
 str or None if (1 if True else 2) else str or bytes or None
 (str or None) if (1 if True else 2) else (str or bytes or None)
-(super_long_variable_name or None) if (1 if super_long_test_name else 2) else (str or bytes or None)
+(
+    (super_long_variable_name or None) if (1 if super_long_test_name else 2) else (str or bytes or None)
+)
 {'2.7': dead, '3.7': (long_live or die_hard)}
 {'2.7': dead, '3.7': (long_live or die_hard), **{'3.6': verygood}}
 {**a, **b, **c}
@@ -743,12 +713,17 @@ str or None if (1 if True else 2) else str or bytes or None
 (1, 2, 3)
 []
 [1, 2, 3, 4, 5, 6, 7, 8, 9, (10 or A), (11 or B), (12 or C)]
-[1, 2, 3,]
+[1, 2, 3]
 [*a]
 [*range(10)]
-[*a, 4, 5,]
-[4, *a, 5,]
-[this_is_a_very_long_variable_which_will_force_a_delimiter_split, element, another, *more]
+[*a, 4, 5]
+[4, *a, 5]
+[
+    this_is_a_very_long_variable_which_will_force_a_delimiter_split,
+    element,
+    another,
+    *more,
+]
 {i for i in (1, 2, 3)}
 {(i ** 2) for i in (1, 2, 3)}
 {(i ** 2) for i, _ in ((1, 'a'), (2, 'b'), (3, 'c'))}
@@ -782,9 +757,11 @@ call.me(maybe)
 list[str]
 dict[str, int]
 tuple[str, ...]
-tuple[
+(
+    tuple[
     str, int, float, dict[str, int]
 ]
+)
 tuple[str, int, float, dict[str, int],]
 very_long_variable_name_filters: t.List[
     t.Tuple[str, t.Union[str, t.List[t.Optional[str]]]],
@@ -828,7 +805,7 @@ numpy[np.newaxis, :]
 {'2.7': dead, '3.7': long_live or die_hard}
 {'2.7', '3.6', '3.7', '3.8', '3.9', '4.0' if gilectomy else '3.10'}
 [1, 2, 3, 4, 5, 6, 7, 8, 9, 10 or A, 11 or B, 12 or C]
-SomeName
+(SomeName)
 SomeName
 (Good, Bad, Ugly)
 (i for i in (1, 2, 3))
@@ -936,11 +913,25 @@ if (
     ~ aaaaaaaaaaaaaaaa.a + aaaaaaaaaaaaaaaa.b - aaaaaaaaaaaaaaaa.c * aaaaaaaaaaaaaaaa.d @ aaaaaaaaaaaaaaaa.e | aaaaaaaaaaaaaaaa.f & aaaaaaaaaaaaaaaa.g % aaaaaaaaaaaaaaaa.h ^ aaaaaaaaaaaaaaaa.i << aaaaaaaaaaaaaaaa.k >> aaaaaaaaaaaaaaaa.l ** aaaaaaaaaaaaaaaa.m // aaaaaaaaaaaaaaaa.n
 ):
     return True
-aaaaaaaaaaaaaaaa + aaaaaaaaaaaaaaaa - aaaaaaaaaaaaaaaa * (aaaaaaaaaaaaaaaa + aaaaaaaaaaaaaaaa) / (aaaaaaaaaaaaaaaa + aaaaaaaaaaaaaaaa + aaaaaaaaaaaaaaaa)
+(
+    aaaaaaaaaaaaaaaa
+    + aaaaaaaaaaaaaaaa
+    - aaaaaaaaaaaaaaaa
+    * (aaaaaaaaaaaaaaaa + aaaaaaaaaaaaaaaa)
+    / (aaaaaaaaaaaaaaaa + aaaaaaaaaaaaaaaa + aaaaaaaaaaaaaaaa)
+)
 aaaaaaaaaaaaaaaa + aaaaaaaaaaaaaaaa
-aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa >> aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa << aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa
+(
+    aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa
+    >> aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa
+    << aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa
+)
 bbbb >> bbbb * bbbb
-aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa ^bbbb.a & aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa^aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa
+(
+    aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa
+    ^ bbbb.a & aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa
+    ^ aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa
+)
 last_call()
 # standalone comment at ENDMARKER
 ```

crates/ruff_python_formatter/src/snapshots/ruff_python_formatter__tests__black_test__slices_py.snap

@@ -76,16 +76,51 @@ x[
 ```diff
 --- Black
 +++ Ruff
-@@ -56,4 +56,8 @@
+@@ -31,14 +31,17 @@
+ ham[lower + offset : upper + offset]
+ 
+ slice[::, ::]
+-slice[
++(
++    slice[
+     # A
+     :
+     # B
+     :
+     # C
+ ]
+-slice[
++)
++(
++    slice[
+     # A
+     1:
+     # B
+@@ -46,8 +49,10 @@
+     # C
+     3
+ ]
++)
+ 
+-slice[
++(
++    slice[
+     # A
+     1
+     + 2 :
+@@ -56,4 +61,11 @@
      # C
      4
  ]
 -x[1:2:3]  # A  # B  # C
-+x[
++)
++(
++    x[
 +    1:  # A
 +    2:  # B
 +    3  # C
 +]
++)
 ```
 
 ## Ruff Output
@@ -124,14 +159,17 @@ ham[: upper_fn(x) : step_fn(x)], ham[:: step_fn(x)]
 ham[lower + offset : upper + offset]
 
 slice[::, ::]
-slice[
+(
+    slice[
     # A
     :
     # B
     :
     # C
 ]
-slice[
+)
+(
+    slice[
     # A
     1:
     # B
@@ -139,8 +177,10 @@ slice[
     # C
     3
 ]
+)
 
-slice[
+(
+    slice[
     # A
     1
     + 2 :
@@ -149,11 +189,14 @@ slice[
     # C
     4
 ]
-x[
+)
+(
+    x[
     1:  # A
     2:  # B
     3  # C
 ]
+)
 ```
 
 ## Black Output

crates/ruff_python_formatter/src/snapshots/ruff_python_formatter__tests__black_test__torture_py.snap

@@ -42,14 +42,13 @@ assert (
 ```diff
 --- Black
 +++ Ruff
-@@ -1,58 +1,33 @@
- importA
--(
--    ()
--    << 0
+@@ -2,18 +2,13 @@
+ (
+     ()
+     << 0
 -    ** 101234234242352525425252352352525234890264906820496920680926538059059209922523523525
--)  #
-+() << 0 ** 101234234242352525425252352352525234890264906820496920680926538059059209922523523525  #
++    **101234234242352525425252352352525234890264906820496920680926538059059209922523523525
+ )  #
  
  assert sort_by_dependency(
      {
@@ -65,12 +64,7 @@ assert (
      }
  ) == ["2a", "2b", "2", "3a", "3b", "3", "1"]
  
- importA
- 0
--0 ^ 0  #
-+0^0  #
- 
- 
+@@ -25,34 +20,18 @@
  class A:
      def foo(self):
          for _ in range(10):
@@ -120,7 +114,11 @@ assert (
 
 ```py
 importA
-() << 0 ** 101234234242352525425252352352525234890264906820496920680926538059059209922523523525  #
+(
+    ()
+    << 0
+    **101234234242352525425252352352525234890264906820496920680926538059059209922523523525
+)  #
 
 assert sort_by_dependency(
     {
@@ -131,7 +129,7 @@ assert sort_by_dependency(
 
 importA
 0
-0^0  #
+0 ^ 0  #
 
 
 class A:

crates/ruff_python_formatter/src/snapshots/ruff_python_formatter__tests__ruff_test__expression__binary_expression_py.snap

@@ -0,0 +1,123 @@
+---
+source: crates/ruff_python_formatter/src/lib.rs
+expression: snapshot
+---
+## Input
+```py
+(aaaaaaaa
+    + # trailing operator comment
+    b # trailing right comment
+)
+
+
+(aaaaaaaa # trailing left comment
+    +  # trailing operator comment
+    # leading right comment
+    b
+)
+
+
+# Black breaks the right side first for the following expressions:
+aaaaaaaaaaaaaa + caaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaal(argument1, argument2, argument3)
+aaaaaaaaaaaaaa + [bbbbbbbbbbbbbbbbbbbbbb, ccccccccccccccccccccc, dddddddddddddddd, eeeeeee]
+aaaaaaaaaaaaaa + (bbbbbbbbbbbbbbbbbbbbbb, ccccccccccccccccccccc, dddddddddddddddd, eeeeeee)
+aaaaaaaaaaaaaa + { key1:bbbbbbbbbbbbbbbbbbbbbb, key2: ccccccccccccccccccccc, key3: dddddddddddddddd, key4: eeeeeee }
+aaaaaaaaaaaaaa + { bbbbbbbbbbbbbbbbbbbbbb, ccccccccccccccccccccc, dddddddddddddddd, eeeeeee }
+aaaaaaaaaaaaaa + [a for x in bbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbb ]
+aaaaaaaaaaaaaa + (a for x in bbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbb )
+aaaaaaaaaaaaaa + {a for x in bbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbb}
+
+# Wraps it in parentheses if it needs to break both left and right
+aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa + [
+    bbbbbbbbbbbbbbbbbbbbbb,
+    ccccccccccccccccccccc,
+    dddddddddddddddd,
+    eee
+] # comment
+
+
+
+# But only for expressions that have a statement parent.
+not (aaaaaaaaaaaaaa + {a for x in bbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbb})
+[a + [bbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbb] in c ]
+
+
+# leading comment
+(
+    # comment
+    content + b
+)
+```
+
+
+
+## Output
+```py
+(
+    aaaaaaaa
+    + b  # trailing operator comment  # trailing right comment
+)
+
+
+(
+    aaaaaaaa  # trailing left comment
+    +  # trailing operator comment
+    # leading right comment
+    b
+)
+# Black breaks the right side first for the following expressions:
+(
+    aaaaaaaaaaaaaa
+    + caaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaal(argument1, argument2, argument3)
+)
+aaaaaaaaaaaaaa + [
+    bbbbbbbbbbbbbbbbbbbbbb,
+    ccccccccccccccccccccc,
+    dddddddddddddddd,
+    eeeeeee,
+]
+(
+    aaaaaaaaaaaaaa
+    + (bbbbbbbbbbbbbbbbbbbbbb, ccccccccccccccccccccc, dddddddddddddddd, eeeeeee)
+)
+(
+    aaaaaaaaaaaaaa
+    + { key1:bbbbbbbbbbbbbbbbbbbbbb, key2: ccccccccccccccccccccc, key3: dddddddddddddddd, key4: eeeeeee }
+)
+(
+    aaaaaaaaaaaaaa
+    + { bbbbbbbbbbbbbbbbbbbbbb, ccccccccccccccccccccc, dddddddddddddddd, eeeeeee }
+)
+(
+    aaaaaaaaaaaaaa
+    + [a for x in bbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbb ]
+)
+(
+    aaaaaaaaaaaaaa
+    + (a for x in bbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbb )
+)
+(
+    aaaaaaaaaaaaaa
+    + {a for x in bbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbb}
+)
+# Wraps it in parentheses if it needs to break both left and right
+(
+    aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa
+    + [bbbbbbbbbbbbbbbbbbbbbb, ccccccccccccccccccccc, dddddddddddddddd, eee]
+)  # comment
+# But only for expressions that have a statement parent.
+(
+    not (aaaaaaaaaaaaaa + {a for x in bbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbb})
+)
+[
+    a + [bbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbb] in c,
+]
+# leading comment
+(
+    # comment
+    content
+    + b
+)
+```
+
+

crates/ruff_python_formatter/src/statement/stmt_expr.rs

@@ -1,3 +1,4 @@
+use crate::expression::parentheses::Parenthesize;
 use crate::prelude::*;
 use crate::FormatNodeRule;
 use rustpython_parser::ast::StmtExpr;
@@ -9,6 +10,6 @@ impl FormatNodeRule<StmtExpr> for FormatStmtExpr {
     fn fmt_fields(&self, item: &StmtExpr, f: &mut PyFormatter) -> FormatResult<()> {
         let StmtExpr { value, .. } = item;
 
-        value.format().fmt(f)
+        value.format().with_options(Parenthesize::Optional).fmt(f)
     }
 }

crates/ruff_python_formatter/src/statement/stmt_while.rs

@@ -1,5 +1,5 @@
 use crate::comments::{leading_alternate_branch_comments, trailing_comments};
-use crate::expression::maybe_parenthesize::maybe_parenthesize;
+use crate::expression::parentheses::Parenthesize;
 use crate::prelude::*;
 use crate::FormatNodeRule;
 use ruff_formatter::write;
@@ -33,7 +33,7 @@ impl FormatNodeRule<StmtWhile> for FormatStmtWhile {
             [
                 text("while"),
                 space(),
-                maybe_parenthesize(test),
+                test.format().with_options(Parenthesize::IfBreaks),
                 text(":"),
                 trailing_comments(trailing_condition_comments),
                 block_indent(&body.format())

crates/ruff_python_formatter/src/trivia.rs

@@ -10,8 +10,8 @@ use ruff_text_size::{TextLen, TextRange, TextSize};
 ///
 /// Returns `None` if the range is empty or only contains trivia (whitespace or comments).
 pub(crate) fn find_first_non_trivia_character_in_range(
-    code: &str,
     range: TextRange,
+    code: &str,
 ) -> Option<(TextSize, char)> {
     let rest = &code[range];
     let mut char_iter = rest.chars();
@@ -40,8 +40,63 @@ pub(crate) fn find_first_non_trivia_character_in_range(
     None
 }
 
+pub(crate) fn find_first_non_trivia_character_after(
+    offset: TextSize,
+    code: &str,
+) -> Option<(TextSize, char)> {
+    find_first_non_trivia_character_in_range(TextRange::new(offset, code.text_len()), code)
+}
+
+pub(crate) fn find_first_non_trivia_character_before(
+    offset: TextSize,
+    code: &str,
+) -> Option<(TextSize, char)> {
+    let head = &code[TextRange::up_to(offset)];
+    let mut char_iter = head.chars();
+
+    while let Some(c) = char_iter.next_back() {
+        match c {
+            c if is_python_whitespace(c) => {
+                continue;
+            }
+
+            // Empty comment
+            '#' => continue,
+
+            non_trivia_character => {
+                // Non trivia character but we don't know if it is a comment or not. Consume all characters
+                // until the start of the line and track if the last non-whitespace character was a `#`.
+                let mut is_comment = false;
+
+                let first_non_trivia_offset = char_iter.as_str().text_len();
+
+                while let Some(c) = char_iter.next_back() {
+                    match c {
+                        '#' => {
+                            is_comment = true;
+                        }
+                        '\n' | '\r' => {
+                            if !is_comment {
+                                return Some((first_non_trivia_offset, non_trivia_character));
+                            }
+                        }
+
+                        c => {
+                            if !is_python_whitespace(c) {
+                                is_comment = false;
+                            }
+                        }
+                    }
+                }
+            }
+        }
+    }
+
+    None
+}
+
 /// Returns the number of newlines between `offset` and the first non whitespace character in the source code.
-pub(crate) fn lines_before(code: &str, offset: TextSize) -> u32 {
+pub(crate) fn lines_before(offset: TextSize, code: &str) -> u32 {
     let head = &code[TextRange::up_to(offset)];
     let mut newlines = 0u32;
 
@@ -68,7 +123,7 @@ pub(crate) fn lines_before(code: &str, offset: TextSize) -> u32 {
 }
 
 /// Counts the empty lines between `offset` and the first non-whitespace character.
-pub(crate) fn lines_after(code: &str, offset: TextSize) -> u32 {
+pub(crate) fn lines_after(offset: TextSize, code: &str) -> u32 {
     let rest = &code[usize::from(offset)..];
     let mut newlines = 0;
 
@@ -98,33 +153,33 @@ mod tests {
 
     #[test]
     fn lines_before_empty_string() {
-        assert_eq!(lines_before("", TextSize::new(0)), 0);
+        assert_eq!(lines_before(TextSize::new(0), ""), 0);
     }
 
     #[test]
     fn lines_before_in_the_middle_of_a_line() {
-        assert_eq!(lines_before("a = 20", TextSize::new(4)), 0);
+        assert_eq!(lines_before(TextSize::new(4), "a = 20"), 0);
     }
 
     #[test]
     fn lines_before_on_a_new_line() {
-        assert_eq!(lines_before("a = 20\nb = 10", TextSize::new(7)), 1);
+        assert_eq!(lines_before(TextSize::new(7), "a = 20\nb = 10"), 1);
     }
 
     #[test]
     fn lines_before_multiple_leading_newlines() {
-        assert_eq!(lines_before("a = 20\n\r\nb = 10", TextSize::new(9)), 2);
+        assert_eq!(lines_before(TextSize::new(9), "a = 20\n\r\nb = 10"), 2);
     }
 
     #[test]
     fn lines_before_with_comment_offset() {
-        assert_eq!(lines_before("a = 20\n# a comment", TextSize::new(8)), 0);
+        assert_eq!(lines_before(TextSize::new(8), "a = 20\n# a comment"), 0);
     }
 
     #[test]
     fn lines_before_with_trailing_comment() {
         assert_eq!(
-            lines_before("a = 20 # some comment\nb = 10", TextSize::new(22)),
+            lines_before(TextSize::new(22), "a = 20 # some comment\nb = 10"),
             1
         );
     }
@@ -132,40 +187,40 @@ mod tests {
     #[test]
     fn lines_before_with_comment_only_line() {
         assert_eq!(
-            lines_before("a = 20\n# some comment\nb = 10", TextSize::new(22)),
+            lines_before(TextSize::new(22), "a = 20\n# some comment\nb = 10"),
             1
         );
     }
 
     #[test]
     fn lines_after_empty_string() {
-        assert_eq!(lines_after("", TextSize::new(0)), 0);
+        assert_eq!(lines_after(TextSize::new(0), ""), 0);
     }
 
     #[test]
     fn lines_after_in_the_middle_of_a_line() {
-        assert_eq!(lines_after("a = 20", TextSize::new(4)), 0);
+        assert_eq!(lines_after(TextSize::new(4), "a = 20"), 0);
     }
 
     #[test]
     fn lines_after_before_a_new_line() {
-        assert_eq!(lines_after("a = 20\nb = 10", TextSize::new(6)), 1);
+        assert_eq!(lines_after(TextSize::new(6), "a = 20\nb = 10"), 1);
     }
 
     #[test]
     fn lines_after_multiple_newlines() {
-        assert_eq!(lines_after("a = 20\n\r\nb = 10", TextSize::new(6)), 2);
+        assert_eq!(lines_after(TextSize::new(6), "a = 20\n\r\nb = 10"), 2);
     }
 
     #[test]
     fn lines_after_before_comment_offset() {
-        assert_eq!(lines_after("a = 20 # a comment\n", TextSize::new(7)), 0);
+        assert_eq!(lines_after(TextSize::new(7), "a = 20 # a comment\n"), 0);
     }
 
     #[test]
     fn lines_after_with_comment_only_line() {
         assert_eq!(
-            lines_after("a = 20\n# some comment\nb = 10", TextSize::new(6)),
+            lines_after(TextSize::new(6), "a = 20\n# some comment\nb = 10"),
             1
         );
     }

crates/ruff_python_semantic/src/scope.rs

@@ -23,6 +23,8 @@ pub struct Scope<'a> {
     bindings: FxHashMap<&'a str, BindingId>,
     /// A map from binding ID to binding ID that it shadows.
     shadowed_bindings: HashMap<BindingId, BindingId, BuildNoHashHasher<BindingId>>,
+    /// A list of all names that have been deleted in this scope.
+    deleted_symbols: Vec<&'a str>,
     /// Index into the globals arena, if the scope contains any globally-declared symbols.
     globals_id: Option<GlobalsId>,
 }
@@ -36,6 +38,7 @@ impl<'a> Scope<'a> {
             star_imports: Vec::default(),
             bindings: FxHashMap::default(),
             shadowed_bindings: IntMap::default(),
+            deleted_symbols: Vec::default(),
             globals_id: None,
         }
     }
@@ -48,6 +51,7 @@ impl<'a> Scope<'a> {
             star_imports: Vec::default(),
             bindings: FxHashMap::default(),
             shadowed_bindings: IntMap::default(),
+            deleted_symbols: Vec::default(),
             globals_id: None,
         }
     }
@@ -67,14 +71,23 @@ impl<'a> Scope<'a> {
         }
     }
 
-    /// Returns `true` if this scope defines a binding with the given name.
-    pub fn defines(&self, name: &str) -> bool {
+    /// Removes the binding with the given name.
+    pub fn delete(&mut self, name: &'a str) -> Option<BindingId> {
+        self.deleted_symbols.push(name);
+        self.bindings.remove(name)
+    }
+
+    /// Returns `true` if this scope has a binding with the given name.
+    pub fn has(&self, name: &str) -> bool {
         self.bindings.contains_key(name)
     }
 
-    /// Removes the binding with the given name
-    pub fn remove(&mut self, name: &str) -> Option<BindingId> {
-        self.bindings.remove(name)
+    /// Returns `true` if the scope declares a symbol with the given name.
+    ///
+    /// Unlike [`Scope::has`], the name may no longer be bound to a value (e.g., it could be
+    /// deleted).
+    pub fn declares(&self, name: &str) -> bool {
+        self.has(name) || self.deleted_symbols.contains(&name)
     }
 
     /// Returns the ids of all bindings defined in this scope.

docs/tutorial.md

@@ -242,7 +242,7 @@ This tutorial has focused on Ruff's command-line interface, but Ruff can also be
 ```yaml
 - repo: https://github.com/astral-sh/ruff-pre-commit
   # Ruff version.
-  rev: v0.0.270
+  rev: v0.0.271
   hooks:
     - id: ruff
 ```

docs/usage.md

@@ -22,7 +22,7 @@ Ruff can also be used as a [pre-commit](https://pre-commit.com) hook:
 ```yaml
 - repo: https://github.com/astral-sh/ruff-pre-commit
   # Ruff version.
-  rev: v0.0.270
+  rev: v0.0.271
   hooks:
     - id: ruff
 ```
@@ -32,7 +32,7 @@ Or, to enable autofix:
 ```yaml
 - repo: https://github.com/astral-sh/ruff-pre-commit
   # Ruff version.
-  rev: v0.0.270
+  rev: v0.0.271
   hooks:
     - id: ruff
       args: [ --fix, --exit-non-zero-on-fix ]

pyproject.toml

@@ -5,7 +5,7 @@ build-backend = "maturin"
 
 [project]
 name = "ruff"
-version = "0.0.270"
+version = "0.0.271"
 description = "An extremely fast Python linter, written in Rust."
 authors = [{ name = "Charlie Marsh", email = "charlie.r.marsh@gmail.com" }]
 maintainers = [{ name = "Charlie Marsh", email = "charlie.r.marsh@gmail.com" }]