feat: add noIrregularWhitespace rule (#3333)

crates/biome_diagnostics/src/display/frame.rs

@@ -230,10 +230,8 @@ pub(super) fn print_frame(fmt: &mut fmt::Formatter<'_>, location: Location<'_>)
                     match c {
                         '\t' => fmt.write_str("\t")?,
                         _ => {
-                            if let Some(width) = c.width() {
-                                for _ in 0..width {
-                                    fmt.write_str(" ")?;
-                                }
+                            for _ in 0..char_width(c) {
+                                fmt.write_str(" ")?;
                             }
                         }
                     }
@@ -338,20 +336,35 @@ pub(super) fn calculate_print_width(mut value: OneIndexed) -> NonZeroUsize {
     width
 }
 
+/// Compute the unicode display width of a string, with the width of tab
+/// characters set to [TAB_WIDTH] and the width of control characters set to 0
+pub(super) fn text_width(text: &str) -> usize {
+    text.chars().map(char_width).sum()
+}
+
 /// We need to set a value here since we have no way of knowing what the user's
 /// preferred tab display width is, so this is set to `2` to match how tab
 /// characters are printed by [print_invisibles]
 const TAB_WIDTH: usize = 2;
 
-/// Compute the unicode display width of a string, with the width of tab
-/// characters set to [TAB_WIDTH] and the width of control characters set to 0
-pub(super) fn text_width(text: &str) -> usize {
-    text.chars()
-        .map(|char| match char {
-            '\t' => TAB_WIDTH,
-            _ => char.width().unwrap_or(0),
-        })
-        .sum()
+/// Some esoteric space characters don't return a width using `char.width()`, so
+/// we need to assume a fixed length for them
+const ESOTERIC_SPACE_WIDTH: usize = 1;
+
+/// Return the width of characters, treating whitespace characters in the way
+/// we need to properly display it
+pub(super) fn char_width(char: char) -> usize {
+    match char {
+        '\t' => TAB_WIDTH,
+        '\u{c}' => ESOTERIC_SPACE_WIDTH,
+        '\u{b}' => ESOTERIC_SPACE_WIDTH,
+        '\u{85}' => ESOTERIC_SPACE_WIDTH,
+        '\u{feff}' => ESOTERIC_SPACE_WIDTH,
+        '\u{180e}' => ESOTERIC_SPACE_WIDTH,
+        '\u{200b}' => ESOTERIC_SPACE_WIDTH,
+        '\u{3000}' => ESOTERIC_SPACE_WIDTH,
+        _ => char.width().unwrap_or(0),
+    }
 }
 
 pub(super) struct PrintInvisiblesOptions {
@@ -462,14 +475,31 @@ pub(super) fn print_invisibles(
 
 fn show_invisible_char(char: char) -> Option<&'static str> {
     match char {
-        ' ' => Some("\u{b7}"),      // Middle Dot
-        '\r' => Some("\u{240d}"),   // Carriage Return Symbol
-        '\n' => Some("\u{23ce}"),   // Return Symbol
-        '\t' => Some("\u{2192} "),  // Rightwards Arrow
-        '\0' => Some("\u{2400}"),   // Null Symbol
-        '\x0b' => Some("\u{240b}"), // Vertical Tabulation Symbol
-        '\x08' => Some("\u{232b}"), // Backspace Symbol
-        '\x0c' => Some("\u{21a1}"), // Downards Two Headed Arrow
+        ' ' => Some("\u{b7}"),          // Middle Dot
+        '\r' => Some("\u{240d}"),       // Carriage Return Symbol
+        '\n' => Some("\u{23ce}"),       // Return Symbol
+        '\t' => Some("\u{2192} "),      // Rightwards Arrow
+        '\0' => Some("\u{2400}"),       // Null Symbol
+        '\x0b' => Some("\u{240b}"),     // Vertical Tabulation Symbol
+        '\x08' => Some("\u{232b}"),     // Backspace Symbol
+        '\x0c' => Some("\u{21a1}"),     // Downwards Two Headed Arrow
+        '\u{85}' => Some("\u{2420}"),   // Space Symbol
+        '\u{a0}' => Some("\u{2420}"),   // Space Symbol
+        '\u{1680}' => Some("\u{2420}"), // Space Symbol
+        '\u{2000}' => Some("\u{2420}"), // Space Symbol
+        '\u{2001}' => Some("\u{2420}"), // Space Symbol
+        '\u{2002}' => Some("\u{2420}"), // Space Symbol
+        '\u{2003}' => Some("\u{2420}"), // Space Symbol
+        '\u{2004}' => Some("\u{2420}"), // Space Symbol
+        '\u{2005}' => Some("\u{2420}"), // Space Symbol
+        '\u{2006}' => Some("\u{2420}"), // Space Symbol
+        '\u{2007}' => Some("\u{2420}"), // Space Symbol
+        '\u{2008}' => Some("\u{2420}"), // Space Symbol
+        '\u{2009}' => Some("\u{2420}"), // Space Symbol
+        '\u{200a}' => Some("\u{2420}"), // Space Symbol
+        '\u{202f}' => Some("\u{2420}"), // Space Symbol
+        '\u{205f}' => Some("\u{2420}"), // Space Symbol
+        '\u{3000}' => Some("\u{2420}"), // Space Symbol
         _ => None,
     }
 }

crates/biome_diagnostics_categories/src/categories.rs

@@ -130,6 +130,7 @@ define_categories! {
     "lint/nursery/noImportantInKeyframe": "https://biomejs.dev/linter/rules/no-important-in-keyframe",
     "lint/nursery/noInvalidDirectionInLinearGradient": "https://biomejs.dev/linter/rules/no-invalid-direction-in-linear-gradient",
     "lint/nursery/noInvalidPositionAtImportRule": "https://biomejs.dev/linter/rules/no-invalid-position-at-import-rule",
+    "lint/nursery/noIrregularWhitespace": "https://biomejs.dev/linter/rules/no-irregular-whitespace",
     "lint/nursery/noLabelWithoutControl": "https://biomejs.dev/linter/rules/no-label-without-control",
     "lint/nursery/noMisplacedAssertion": "https://biomejs.dev/linter/rules/no-misplaced-assertion",
     "lint/nursery/noMissingGenericFamilyKeyword": "https://biomejs.dev/linter/rules/no-missing-generic-family-keyword",

crates/biome_js_analyze/src/lint/nursery/no_irregular_whitespace.rs

@@ -0,0 +1,112 @@
+use biome_analyze::{context::RuleContext, declare_lint_rule, Ast, Rule, RuleDiagnostic};
+use biome_analyze::{RuleSource, RuleSourceKind};
+use biome_console::markup;
+use biome_js_syntax::{JsLanguage, JsModule};
+use biome_rowan::{AstNode, Direction, SyntaxTriviaPiece, TextRange};
+
+const IRREGULAR_WHITESPACES: &[char; 22] = &[
+    '\u{c}', '\u{b}', '\u{85}', '\u{feff}', '\u{a0}', '\u{1680}', '\u{180e}', '\u{2000}',
+    '\u{2001}', '\u{2002}', '\u{2003}', '\u{2004}', '\u{2005}', '\u{2006}', '\u{2007}', '\u{2008}',
+    '\u{2009}', '\u{200a}', '\u{200b}', '\u{202f}', '\u{205f}', '\u{3000}',
+];
+
+declare_lint_rule! {
+    /// Disallows the use of irregular whitespace characters.
+    ///
+    /// Invalid or irregular whitespace causes issues with ECMAScript 5 parsers and also makes code harder to debug.
+    ///
+    /// ## Examples
+    ///
+    /// ### Invalid
+    ///
+    /// ```js,expect_diagnostic
+    /// constcount=1;
+    /// ```
+    ///
+    /// ```js,expect_diagnostic
+    /// const foo = 'thing';
+    /// ```
+    ///
+    /// ### Valid
+    ///
+    /// ```js
+    /// const count = 1;
+    /// ```
+    ///
+    /// ```js
+    /// const foo = '';
+    /// ```
+    ///
+    pub NoIrregularWhitespace {
+        version: "next",
+        name: "noIrregularWhitespace",
+        language: "js",
+        recommended: false,
+        sources: &[RuleSource::Eslint("no-irregular-whitespace")],
+        source_kind: RuleSourceKind::SameLogic,
+    }
+}
+
+impl Rule for NoIrregularWhitespace {
+    type Query = Ast<JsModule>;
+    type State = TextRange;
+    type Signals = Vec<Self::State>;
+    type Options = ();
+
+    fn run(ctx: &RuleContext<Self>) -> Self::Signals {
+        let node = ctx.query();
+        get_irregular_whitespace(node)
+    }
+
+    fn diagnostic(_ctx: &RuleContext<Self>, range: &Self::State) -> Option<RuleDiagnostic> {
+        Some(
+            RuleDiagnostic::new(
+                rule_category!(),
+                range,
+                markup! {
+                    "Irregular whitespaces found."
+                },
+            )
+            .note(markup! {
+                "Replace the irregular whitespaces with normal whitespaces or tabs."
+            }),
+        )
+    }
+}
+
+fn get_irregular_whitespace(node: &JsModule) -> Vec<TextRange> {
+    let syntax = node.syntax();
+    let mut all_whitespaces_trivia: Vec<SyntaxTriviaPiece<JsLanguage>> = vec![];
+    let is_whitespace = |trivia: &SyntaxTriviaPiece<JsLanguage>| {
+        trivia.is_whitespace() && !trivia.text().replace(' ', "").is_empty()
+    };
+
+    for token in syntax.descendants_tokens(Direction::Next) {
+        let leading_trivia_pieces = token.leading_trivia().pieces();
+        let trailing_trivia_pieces = token.trailing_trivia().pieces();
+
+        for trivia in leading_trivia_pieces {
+            if is_whitespace(&trivia) {
+                all_whitespaces_trivia.push(trivia);
+            }
+        }
+
+        for trivia in trailing_trivia_pieces {
+            if is_whitespace(&trivia) {
+                all_whitespaces_trivia.push(trivia);
+            }
+        }
+    }
+
+    all_whitespaces_trivia
+        .iter()
+        .filter_map(|trivia| {
+            let has_irregular_whitespace = trivia.text().chars().any(|char| {
+                IRREGULAR_WHITESPACES
+                    .iter()
+                    .any(|irregular_whitespace| &char == irregular_whitespace)
+            });
+            has_irregular_whitespace.then(|| trivia.text_range())
+        })
+        .collect::<Vec<TextRange>>()
+}

crates/biome_js_analyze/tests/specs/nursery/noIrregularWhitespace/invalid.js

@@ -0,0 +1,20 @@
+/* \u{b}    */ constfoo='thing';
+/* \u{c}    */ constfoo='thing';
+/* \u{feff} */ constfoo='thing';
+/* \u{a0}   */ const foo = 'thing';
+/* \u{1680} */ const foo = 'thing';
+/* \u{2000} */ const foo = 'thing';
+/* \u{2001} */ const foo = 'thing';
+/* \u{2002} */ const foo = 'thing';
+/* \u{2003} */ const foo = 'thing';
+/* \u{2004} */ const foo = 'thing';
+/* \u{2005} */ const foo = 'thing';
+/* \u{2006} */ const foo = 'thing';
+/* \u{2007} */ const foo = 'thing';
+/* \u{2008} */ const foo = 'thing';
+/* \u{2009} */ const foo = 'thing';
+/* \u{200a} */ const foo = 'thing';
+/* \u{200b} */ const​foo​=​'thing';
+/* \u{202f} */ const foo = 'thing';
+/* \u{205f} */ const foo = 'thing';
+/* \u{3000} */ const　foo　=　'thing';