[ruff] Indented form feeds (RUF054) (#16049)

## Summary

Resolves #12321.

The physical-line-based `RUF054` checks for form feed characters that
are preceded by only tabs and spaces, but not any other characters,
including form feeds.

## Test Plan

`cargo nextest run` and `cargo insta test`.

crates/ruff_linter/resources/test/fixtures/ruff/RUF054.py

@@ -0,0 +1,32 @@
+############# Warning ############
+# This file contains form feeds. #
+############# Warning ############
+
+
+# Errors
+
+
+
+
+
+def _():
+		pass
+
+if False:
+    print('F')
+    print('T')
+
+
+# No errors
+
+
+
+
+
+
+
+def _():
+    pass
+
+def f():
+	pass 

crates/ruff_linter/src/checkers/physical_lines.rs

@@ -13,6 +13,7 @@ use crate::rules::pycodestyle::rules::{
     trailing_whitespace,
 };
 use crate::rules::pylint;
+use crate::rules::ruff::rules::indented_form_feed;
 use crate::settings::LinterSettings;
 use crate::Locator;
 
@@ -71,6 +72,12 @@ pub(crate) fn check_physical_lines(
                 diagnostics.push(diagnostic);
             }
         }
+
+        if settings.rules.enabled(Rule::IndentedFormFeed) {
+            if let Some(diagnostic) = indented_form_feed(&line) {
+                diagnostics.push(diagnostic);
+            }
+        }
     }
 
     if enforce_no_newline_at_end_of_file {

crates/ruff_linter/src/codes.rs

@@ -1006,6 +1006,7 @@ pub fn code_to_rule(linter: Linter, code: &str) -> Option<(RuleGroup, Rule)> {
         (Ruff, "051") => (RuleGroup::Preview, rules::ruff::rules::IfKeyInDictDel),
         (Ruff, "052") => (RuleGroup::Preview, rules::ruff::rules::UsedDummyVariable),
         (Ruff, "053") => (RuleGroup::Preview, rules::ruff::rules::ClassWithMixedTypeVars),
+        (Ruff, "054") => (RuleGroup::Preview, rules::ruff::rules::IndentedFormFeed),
         (Ruff, "055") => (RuleGroup::Preview, rules::ruff::rules::UnnecessaryRegularExpression),
         (Ruff, "056") => (RuleGroup::Preview, rules::ruff::rules::FalsyDictGetFallback),
         (Ruff, "057") => (RuleGroup::Preview, rules::ruff::rules::UnnecessaryRound),

crates/ruff_linter/src/registry.rs

@@ -253,6 +253,7 @@ impl Rule {
             Rule::BidirectionalUnicode
             | Rule::BlankLineWithWhitespace
             | Rule::DocLineTooLong
+            | Rule::IndentedFormFeed
             | Rule::LineTooLong
             | Rule::MissingCopyrightNotice
             | Rule::MissingNewlineAtEndOfFile

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

@@ -11,11 +11,10 @@ mod tests {
 
     use anyhow::Result;
     use regex::Regex;
+    use ruff_source_file::SourceFileBuilder;
     use rustc_hash::FxHashSet;
     use test_case::test_case;
 
-    use ruff_source_file::SourceFileBuilder;
-
     use crate::pyproject_toml::lint_pyproject_toml;
     use crate::registry::Rule;
     use crate::settings::types::{
@@ -436,6 +435,7 @@ mod tests {
     #[test_case(Rule::StarmapZip, Path::new("RUF058_0.py"))]
     #[test_case(Rule::StarmapZip, Path::new("RUF058_1.py"))]
     #[test_case(Rule::ClassWithMixedTypeVars, Path::new("RUF053.py"))]
+    #[test_case(Rule::IndentedFormFeed, Path::new("RUF054.py"))]
     fn preview_rules(rule_code: Rule, path: &Path) -> Result<()> {
         let snapshot = format!(
             "preview__{}_{}",

crates/ruff_linter/src/rules/ruff/rules/indented_form_feed.rs

@@ -0,0 +1,71 @@
+use memchr::memchr;
+
+use ruff_diagnostics::{Diagnostic, Violation};
+use ruff_macros::{derive_message_formats, ViolationMetadata};
+use ruff_source_file::Line;
+use ruff_text_size::{TextRange, TextSize};
+
+/// ## What it does
+/// Checks for form feed characters preceded by either a space or a tab.
+///
+/// ## Why is this bad?
+/// [The language reference][lexical-analysis-indentation] states:
+///
+/// > A formfeed character may be present at the start of the line;
+/// > it will be ignored for the indentation calculations above.
+/// > Formfeed characters occurring elsewhere in the leading whitespace
+/// > have an undefined effect (for instance, they may reset the space count to zero).
+///
+/// ## Example
+///
+/// ```python
+/// if foo():\n    \fbar()
+/// ```
+///
+/// Use instead:
+///
+/// ```python
+/// if foo():\n    bar()
+/// ```
+///
+/// [lexical-analysis-indentation]: https://docs.python.org/3/reference/lexical_analysis.html#indentation
+#[derive(ViolationMetadata)]
+pub(crate) struct IndentedFormFeed;
+
+impl Violation for IndentedFormFeed {
+    #[derive_message_formats]
+    fn message(&self) -> String {
+        "Indented form feed".to_string()
+    }
+
+    fn fix_title(&self) -> Option<String> {
+        Some("Remove form feed".to_string())
+    }
+}
+
+const FORM_FEED: u8 = b'\x0c';
+const SPACE: u8 = b' ';
+const TAB: u8 = b'\t';
+
+/// RUF054
+pub(crate) fn indented_form_feed(line: &Line) -> Option<Diagnostic> {
+    let index_relative_to_line = memchr(FORM_FEED, line.as_bytes())?;
+
+    if index_relative_to_line == 0 {
+        return None;
+    }
+
+    if line[..index_relative_to_line]
+        .as_bytes()
+        .iter()
+        .any(|byte| *byte != SPACE && *byte != TAB)
+    {
+        return None;
+    }
+
+    let relative_index = u32::try_from(index_relative_to_line).ok()?;
+    let absolute_index = line.start() + TextSize::new(relative_index);
+    let range = TextRange::at(absolute_index, 1.into());
+
+    Some(Diagnostic::new(IndentedFormFeed, range))
+}

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

@@ -13,6 +13,7 @@ pub(crate) use function_call_in_dataclass_default::*;
 pub(crate) use if_key_in_dict_del::*;
 pub(crate) use implicit_optional::*;
 pub(crate) use incorrectly_parenthesized_tuple_in_subscript::*;
+pub(crate) use indented_form_feed::*;
 pub(crate) use invalid_assert_message_literal_argument::*;
 pub(crate) use invalid_formatter_suppression_comment::*;
 pub(crate) use invalid_index_type::*;
@@ -69,6 +70,7 @@ mod helpers;
 mod if_key_in_dict_del;
 mod implicit_optional;
 mod incorrectly_parenthesized_tuple_in_subscript;
+mod indented_form_feed;
 mod invalid_assert_message_literal_argument;
 mod invalid_formatter_suppression_comment;
 mod invalid_index_type;

crates/ruff_linter/src/rules/ruff/snapshots/ruff_linter__rules__ruff__tests__preview__RUF054_RUF054.py.snap

@@ -0,0 +1,39 @@
+---
+source: crates/ruff_linter/src/rules/ruff/mod.rs
+---
+RUF054.py:8:2: RUF054 Indented form feed
+  |
+6 | # Errors
+7 |
+8 |  
+  |  ^ RUF054
+  |
+  = help: Remove form feed
+
+RUF054.py:10:3: RUF054 Indented form feed
+   |
+10 |         
+   |         ^ RUF054
+11 |
+12 | def _():
+   |
+   = help: Remove form feed
+
+RUF054.py:13:2: RUF054 Indented form feed
+   |
+12 | def _():
+13 |        pass
+   |     ^ RUF054
+14 |
+15 | if False:
+   |
+   = help: Remove form feed
+
+RUF054.py:17:5: RUF054 Indented form feed
+   |
+15 | if False:
+16 |     print('F')
+17 |     print('T')
+   |     ^ RUF054
+   |
+   = help: Remove form feed

scripts/check_docs_formatted.py

@@ -97,6 +97,7 @@
 KNOWN_PARSE_ERRORS = [
     "blank-line-with-whitespace",
     "indentation-with-invalid-multiple-comment",
+    "indented-form-feed",
     "missing-newline-at-end-of-file",
     "mixed-spaces-and-tabs",
     "no-indented-block",