Add support for pycodestyle E101 (#2038)

Rule described here: https://www.flake8rules.com/rules/E101.html

I tried to follow contributing guidelines closely, I've never worked with Rust before. Stumbled across Ruff a few days ago and would like to use it in our project, but we use a bunch of flake8 rules that are not yet implemented in ruff, so I decided to give it a go.

README.md

@@ -596,6 +596,7 @@ For more, see [pycodestyle](https://pypi.org/project/pycodestyle/2.9.1/) on PyPI
 #### Error (E)
 | Code | Name | Message | Fix |
 | ---- | ---- | ------- | --- |
+| E101 | mixed-spaces-and-tabs | Indentation contains mixed spaces and tabs |  |
 | E401 | multiple-imports-on-one-line | Multiple imports on one line |  |
 | E402 | module-import-not-at-top-of-file | Module level import not at top of file |  |
 | E501 | line-too-long | Line too long ({length} > {limit} characters) |  |

resources/test/fixtures/pycodestyle/E101.py

@@ -0,0 +1,19 @@
+def func_all_spaces():
+    # No error
+    print("spaces")
+
+def func_tabs():
+    # No error
+	print("tabs")
+
+def func_mixed_start_with_tab():
+    # E101
+	  print("mixed starts with tab")
+
+def func_mixed_start_with_space():
+    # E101
+ 	  	  	   print("mixed starts with space")
+
+def xyz():
+    # E101
+   	print("xyz");

ruff.schema.json

@@ -1311,6 +1311,9 @@
         "DTZ011",
         "DTZ012",
         "E",
+        "E1",
+        "E10",
+        "E101",
         "E4",
         "E40",
         "E401",

src/checkers/lines.rs

@@ -2,7 +2,7 @@
 
 use crate::registry::{Diagnostic, Rule};
 use crate::rules::pycodestyle::rules::{
-    doc_line_too_long, line_too_long, no_newline_at_end_of_file,
+    doc_line_too_long, line_too_long, mixed_spaces_and_tabs, no_newline_at_end_of_file,
 };
 use crate::rules::pygrep_hooks::rules::{blanket_noqa, blanket_type_ignore};
 use crate::rules::pyupgrade::rules::unnecessary_coding_comment;
@@ -25,6 +25,7 @@ pub fn check_lines(
     let enforce_unnecessary_coding_comment = settings
         .rules
         .enabled(&Rule::PEP3120UnnecessaryCodingComment);
+    let enforce_mixed_spaces_and_tabs = settings.rules.enabled(&Rule::MixedSpacesAndTabs);
 
     let mut commented_lines_iter = commented_lines.iter().peekable();
     let mut doc_lines_iter = doc_lines.iter().peekable();
@@ -72,6 +73,12 @@ pub fn check_lines(
             }
         }
 
+        if enforce_mixed_spaces_and_tabs {
+            if let Some(diagnostic) = mixed_spaces_and_tabs(index, line) {
+                diagnostics.push(diagnostic);
+            }
+        }
+
         if enforce_line_too_long {
             if let Some(diagnostic) = line_too_long(index, line, settings) {
                 diagnostics.push(diagnostic);

src/registry.rs

@@ -14,6 +14,7 @@ use crate::{rules, violations};
 
 ruff_macros::define_rule_mapping!(
     // pycodestyle errors
+    E101 => violations::MixedSpacesAndTabs,
     E401 => violations::MultipleImportsOnOneLine,
     E402 => violations::ModuleImportNotAtTopOfFile,
     E501 => violations::LineTooLong,
@@ -556,7 +557,8 @@ impl Rule {
             | Rule::DocLineTooLong
             | Rule::PEP3120UnnecessaryCodingComment
             | Rule::BlanketTypeIgnore
-            | Rule::BlanketNOQA => &LintSource::Lines,
+            | Rule::BlanketNOQA
+            | Rule::MixedSpacesAndTabs => &LintSource::Lines,
             Rule::AmbiguousUnicodeCharacterComment
             | Rule::AmbiguousUnicodeCharacterDocstring
             | Rule::AmbiguousUnicodeCharacterString

src/rules/pycodestyle/mod.rs

@@ -36,6 +36,7 @@ mod tests {
     #[test_case(Rule::NoNewLineAtEndOfFile, Path::new("W292_4.py"))]
     #[test_case(Rule::InvalidEscapeSequence, Path::new("W605_0.py"))]
     #[test_case(Rule::InvalidEscapeSequence, Path::new("W605_1.py"))]
+    #[test_case(Rule::MixedSpacesAndTabs, Path::new("E101.py"))]
     fn rules(rule_code: Rule, path: &Path) -> Result<()> {
         let snapshot = format!("{}_{}", rule_code.code(), path.to_string_lossy());
         let diagnostics = test_path(

src/rules/pycodestyle/rules.rs

@@ -78,6 +78,23 @@ pub fn line_too_long(lineno: usize, line: &str, settings: &Settings) -> Option<D
     }
 }
 
+/// E101
+pub fn mixed_spaces_and_tabs(lineno: usize, line: &str) -> Option<Diagnostic> {
+    let indent = leading_space(line);
+
+    if indent.contains(' ') && indent.contains('\t') {
+        Some(Diagnostic::new(
+            violations::MixedSpacesAndTabs,
+            Range::new(
+                Location::new(lineno + 1, 0),
+                Location::new(lineno + 1, indent.chars().count()),
+            ),
+        ))
+    } else {
+        None
+    }
+}
+
 /// W505
 pub fn doc_line_too_long(lineno: usize, line: &str, settings: &Settings) -> Option<Diagnostic> {
     let Some(limit) = settings.pycodestyle.max_doc_length else {

src/rules/pycodestyle/snapshots/ruff__rules__pycodestyle__tests__E101_E101.py.snap

@@ -0,0 +1,35 @@
+---
+source: src/rules/pycodestyle/mod.rs
+expression: diagnostics
+---
+- kind:
+    MixedSpacesAndTabs: ~
+  location:
+    row: 11
+    column: 0
+  end_location:
+    row: 11
+    column: 3
+  fix: ~
+  parent: ~
+- kind:
+    MixedSpacesAndTabs: ~
+  location:
+    row: 15
+    column: 0
+  end_location:
+    row: 15
+    column: 11
+  fix: ~
+  parent: ~
+- kind:
+    MixedSpacesAndTabs: ~
+  location:
+    row: 19
+    column: 0
+  end_location:
+    row: 19
+    column: 4
+  fix: ~
+  parent: ~
+

src/violations.rs

@@ -48,6 +48,16 @@ impl Violation for LineTooLong {
     }
 }
 
+define_violation!(
+    pub struct MixedSpacesAndTabs;
+);
+impl Violation for MixedSpacesAndTabs {
+    #[derive_message_formats]
+    fn message(&self) -> String {
+        format!("Indentation contains mixed spaces and tabs")
+    }
+}
+
 #[derive(Debug, PartialEq, Eq, Serialize, Deserialize)]
 pub enum EqCmpop {
     Eq,