[pycodestyle] Implement blank-line-at-end-of-file (W391) (#10243)

## Summary

Implements the [blank line at end of
file](https://pycodestyle.pycqa.org/en/latest/intro.html#error-codes)
rule (W391) from pycodestyle. Renamed to TooManyNewlinesAtEndOfFile for
clarity.

## Test Plan

New fixtures have been added

Part of #2402

.gitattributes

@@ -2,6 +2,8 @@
 
 crates/ruff_linter/resources/test/fixtures/isort/line_ending_crlf.py text eol=crlf
 crates/ruff_linter/resources/test/fixtures/pycodestyle/W605_1.py text eol=crlf
+crates/ruff_linter/resources/test/fixtures/pycodestyle/W391_2.py text eol=crlf
+crates/ruff_linter/resources/test/fixtures/pycodestyle/W391_3.py text eol=crlf
 
 crates/ruff_python_formatter/resources/test/fixtures/ruff/docstring_code_examples_crlf.py text eol=crlf
 crates/ruff_python_formatter/tests/snapshots/format@docstring_code_examples_crlf.py.snap text eol=crlf

crates/ruff_linter/resources/test/fixtures/pycodestyle/W391_0.py

@@ -0,0 +1,14 @@
+# Unix style
+def foo() -> None:
+    pass
+
+
+def bar() -> None:
+    pass
+
+
+
+if __name__ == '__main__':
+    foo()
+    bar()
+

crates/ruff_linter/resources/test/fixtures/pycodestyle/W391_1.py

@@ -0,0 +1,13 @@
+# Unix style
+def foo() -> None:
+    pass
+
+
+def bar() -> None:
+    pass
+
+
+
+if __name__ == '__main__':
+    foo()
+    bar()

crates/ruff_linter/resources/test/fixtures/pycodestyle/W391_2.py

@@ -0,0 +1,17 @@
+# Windows style
+def foo() -> None:
+    pass
+
+
+def bar() -> None:
+    pass
+
+
+
+if __name__ == '__main__':
+    foo()
+    bar()
+
+
+
+

crates/ruff_linter/resources/test/fixtures/pycodestyle/W391_3.py

@@ -0,0 +1,13 @@
+# Windows style
+def foo() -> None:
+    pass
+
+
+def bar() -> None:
+    pass
+
+
+
+if __name__ == '__main__':
+    foo()
+    bar()

crates/ruff_linter/resources/test/fixtures/pycodestyle/W391_4.py

@@ -0,0 +1,5 @@
+# This is fine
+def foo():
+    pass
+
+    # Some comment

crates/ruff_linter/src/checkers/tokens.rs

@@ -203,6 +203,10 @@ pub(crate) fn check_tokens(
         flake8_fixme::rules::todos(&mut diagnostics, &todo_comments);
     }
 
+    if settings.rules.enabled(Rule::TooManyNewlinesAtEndOfFile) {
+        pycodestyle::rules::too_many_newlines_at_end_of_file(&mut diagnostics, tokens);
+    }
+
     diagnostics.retain(|diagnostic| settings.rules.enabled(diagnostic.kind.rule()));
 
     diagnostics

crates/ruff_linter/src/codes.rs

@@ -168,6 +168,7 @@ pub fn code_to_rule(linter: Linter, code: &str) -> Option<(RuleGroup, Rule)> {
         (Pycodestyle, "W291") => (RuleGroup::Stable, rules::pycodestyle::rules::TrailingWhitespace),
         (Pycodestyle, "W292") => (RuleGroup::Stable, rules::pycodestyle::rules::MissingNewlineAtEndOfFile),
         (Pycodestyle, "W293") => (RuleGroup::Stable, rules::pycodestyle::rules::BlankLineWithWhitespace),
+        (Pycodestyle, "W391") => (RuleGroup::Preview, rules::pycodestyle::rules::TooManyNewlinesAtEndOfFile),
         (Pycodestyle, "W505") => (RuleGroup::Stable, rules::pycodestyle::rules::DocLineTooLong),
         (Pycodestyle, "W605") => (RuleGroup::Stable, rules::pycodestyle::rules::InvalidEscapeSequence),
 

crates/ruff_linter/src/registry.rs

@@ -300,6 +300,7 @@ impl Rule {
             | Rule::SingleLineImplicitStringConcatenation
             | Rule::TabIndentation
             | Rule::TooManyBlankLines
+            | Rule::TooManyNewlinesAtEndOfFile
             | Rule::TrailingCommaOnBareTuple
             | Rule::TypeCommentInStub
             | Rule::UselessSemicolon

crates/ruff_linter/src/rules/pycodestyle/mod.rs

@@ -72,6 +72,11 @@ mod tests {
     #[test_case(Rule::TypeComparison, Path::new("E721.py"))]
     #[test_case(Rule::ModuleImportNotAtTopOfFile, Path::new("E402_2.py"))]
     #[test_case(Rule::RedundantBackslash, Path::new("E502.py"))]
+    #[test_case(Rule::TooManyNewlinesAtEndOfFile, Path::new("W391_0.py"))]
+    #[test_case(Rule::TooManyNewlinesAtEndOfFile, Path::new("W391_1.py"))]
+    #[test_case(Rule::TooManyNewlinesAtEndOfFile, Path::new("W391_2.py"))]
+    #[test_case(Rule::TooManyNewlinesAtEndOfFile, Path::new("W391_3.py"))]
+    #[test_case(Rule::TooManyNewlinesAtEndOfFile, Path::new("W391_4.py"))]
     fn preview_rules(rule_code: Rule, path: &Path) -> Result<()> {
         let snapshot = format!(
             "preview__{}_{}",

crates/ruff_linter/src/rules/pycodestyle/rules/missing_newline_at_end_of_file.rs

@@ -42,7 +42,7 @@ pub(crate) fn no_newline_at_end_of_file(
 ) -> Option<Diagnostic> {
     let source = locator.contents();
 
-    // Ignore empty and BOM only files
+    // Ignore empty and BOM only files.
     if source.is_empty() || source == "\u{feff}" {
         return None;
     }

crates/ruff_linter/src/rules/pycodestyle/rules/mod.rs

@@ -17,6 +17,7 @@ pub(crate) use module_import_not_at_top_of_file::*;
 pub(crate) use multiple_imports_on_one_line::*;
 pub(crate) use not_tests::*;
 pub(crate) use tab_indentation::*;
+pub(crate) use too_many_newlines_at_end_of_file::*;
 pub(crate) use trailing_whitespace::*;
 pub(crate) use type_comparison::*;
 
@@ -39,5 +40,6 @@ mod module_import_not_at_top_of_file;
 mod multiple_imports_on_one_line;
 mod not_tests;
 mod tab_indentation;
+mod too_many_newlines_at_end_of_file;
 mod trailing_whitespace;
 mod type_comparison;

crates/ruff_linter/src/rules/pycodestyle/rules/too_many_newlines_at_end_of_file.rs

@@ -0,0 +1,99 @@
+use ruff_diagnostics::{AlwaysFixableViolation, Diagnostic, Edit, Fix};
+use ruff_macros::{derive_message_formats, violation};
+use ruff_python_parser::lexer::LexResult;
+use ruff_python_parser::Tok;
+use ruff_text_size::{TextRange, TextSize};
+
+/// ## What it does
+/// Checks for files with multiple trailing blank lines.
+///
+/// ## Why is this bad?
+/// Trailing blank lines in a file are superfluous.
+///
+/// However, the last line of the file should end with a newline.
+///
+/// ## Example
+/// ```python
+/// spam(1)\n\n\n
+/// ```
+///
+/// Use instead:
+/// ```python
+/// spam(1)\n
+/// ```
+#[violation]
+pub struct TooManyNewlinesAtEndOfFile {
+    num_trailing_newlines: u32,
+}
+
+impl AlwaysFixableViolation for TooManyNewlinesAtEndOfFile {
+    #[derive_message_formats]
+    fn message(&self) -> String {
+        let TooManyNewlinesAtEndOfFile {
+            num_trailing_newlines,
+        } = self;
+
+        // We expect a single trailing newline; so two trailing newlines is one too many, three
+        // trailing newlines is two too many, etc.
+        if *num_trailing_newlines > 2 {
+            format!("Too many newlines at end of file")
+        } else {
+            format!("Extra newline at end of file")
+        }
+    }
+
+    fn fix_title(&self) -> String {
+        let TooManyNewlinesAtEndOfFile {
+            num_trailing_newlines,
+        } = self;
+        if *num_trailing_newlines > 2 {
+            "Remove trailing newlines".to_string()
+        } else {
+            "Remove trailing newline".to_string()
+        }
+    }
+}
+
+/// W391
+pub(crate) fn too_many_newlines_at_end_of_file(
+    diagnostics: &mut Vec<Diagnostic>,
+    lxr: &[LexResult],
+) {
+    let mut num_trailing_newlines = 0u32;
+    let mut start: Option<TextSize> = None;
+    let mut end: Option<TextSize> = None;
+
+    // Count the number of trailing newlines.
+    for (tok, range) in lxr.iter().rev().flatten() {
+        match tok {
+            Tok::NonLogicalNewline | Tok::Newline => {
+                if num_trailing_newlines == 0 {
+                    end = Some(range.end());
+                }
+                start = Some(range.end());
+                num_trailing_newlines += 1;
+            }
+            Tok::Dedent => continue,
+            _ => {
+                break;
+            }
+        }
+    }
+
+    if num_trailing_newlines == 0 || num_trailing_newlines == 1 {
+        return;
+    }
+
+    let range = match (start, end) {
+        (Some(start), Some(end)) => TextRange::new(start, end),
+        _ => return,
+    };
+    let mut diagnostic = Diagnostic::new(
+        TooManyNewlinesAtEndOfFile {
+            num_trailing_newlines,
+        },
+        range,
+    );
+    diagnostic.set_fix(Fix::safe_edit(Edit::range_deletion(range)));
+    diagnostics.push(diagnostic);
+}

crates/ruff_linter/src/rules/pycodestyle/snapshots/ruff_linter__rules__pycodestyle__tests__preview__W391_W391_0.py.snap

@@ -0,0 +1,17 @@
+---
+source: crates/ruff_linter/src/rules/pycodestyle/mod.rs
+---
+W391_0.py:14:1: W391 [*] Extra newline at end of file
+   |
+12 |     foo()
+13 |     bar()
+14 | 
+   | ^ W391
+   |
+   = help: Remove trailing newline
+
+ℹ Safe fix
+11 11 | if __name__ == '__main__':
+12 12 |     foo()
+13 13 |     bar()
+14    |-

crates/ruff_linter/src/rules/pycodestyle/snapshots/ruff_linter__rules__pycodestyle__tests__preview__W391_W391_1.py.snap

@@ -0,0 +1,4 @@
+---
+source: crates/ruff_linter/src/rules/pycodestyle/mod.rs
+---
+

crates/ruff_linter/src/rules/pycodestyle/snapshots/ruff_linter__rules__pycodestyle__tests__preview__W391_W391_2.py.snap

@@ -0,0 +1,22 @@
+---
+source: crates/ruff_linter/src/rules/pycodestyle/mod.rs
+---
+W391_2.py:14:1: W391 [*] Too many newlines at end of file
+   |
+12 |       foo()
+13 |       bar()
+14 | / 
+15 | | 
+16 | | 
+17 | | 
+   |
+   = help: Remove trailing newlines
+
+ℹ Safe fix
+11 11 | if __name__ == '__main__':
+12 12 |     foo()
+13 13 |     bar()
+14    |-
+15    |-
+16    |-
+17    |-

crates/ruff_linter/src/rules/pycodestyle/snapshots/ruff_linter__rules__pycodestyle__tests__preview__W391_W391_3.py.snap

@@ -0,0 +1,4 @@
+---
+source: crates/ruff_linter/src/rules/pycodestyle/mod.rs
+---
+

crates/ruff_linter/src/rules/pycodestyle/snapshots/ruff_linter__rules__pycodestyle__tests__preview__W391_W391_4.py.snap

@@ -0,0 +1,4 @@
+---
+source: crates/ruff_linter/src/rules/pycodestyle/mod.rs
+---
+

ruff.schema.json

@@ -3816,6 +3816,9 @@
         "W291",
         "W292",
         "W293",
+        "W3",
+        "W39",
+        "W391",
         "W5",
         "W50",
         "W505",

scripts/check_docs_formatted.py

@@ -105,6 +105,7 @@ KNOWN_PARSE_ERRORS = [
     "tab-after-operator",
     "tab-before-keyword",
     "tab-before-operator",
+    "too-many-newlines-at-end-of-file",
     "trailing-whitespace",
     "unexpected-indentation",
 ]