feat: update Unicode letter detection in capitalized-comments rule (#…

…18375) * feat: simplify `LETTER_PATTERN` regexp * check characters beyond U+FFFF
eslint · Apr 25, 2024 · a498f35 · a498f35
1 parent 1579ce0
commit a498f35
Show file tree

Hide file tree

Showing 3 changed files with 48 additions and 43 deletions.
diff --git a/lib/rules/capitalized-comments.js b/lib/rules/capitalized-comments.js
@@ -8,7 +8,6 @@
 // Requirements
 //------------------------------------------------------------------------------
 
-const LETTER_PATTERN = require("./utils/patterns/letters");
 const astUtils = require("./utils/ast-utils");
 
 //------------------------------------------------------------------------------
@@ -17,7 +16,8 @@ const astUtils = require("./utils/ast-utils");
 
 const DEFAULT_IGNORE_PATTERN = astUtils.COMMENTS_IGNORE_PATTERN,
     WHITESPACE = /\s/gu,
-    MAYBE_URL = /^\s*[^:/?#\s]+:\/\/[^?#]/u; // TODO: Combine w/ max-len pattern?
+    MAYBE_URL = /^\s*[^:/?#\s]+:\/\/[^?#]/u, // TODO: Combine w/ max-len pattern?
+    LETTER_PATTERN = /\p{L}/u;
 
 /*
  * Base schema body for defining the basic capitalization rule, ignorePattern,
@@ -233,7 +233,8 @@ module.exports = {
                 return true;
             }
 
-            const firstWordChar = commentWordCharsOnly[0];
+            // Get the first Unicode character (1 or 2 code units).
+            const [firstWordChar] = commentWordCharsOnly;
 
             if (!LETTER_PATTERN.test(firstWordChar)) {
                 return true;
@@ -273,12 +274,14 @@ module.exports = {
                     messageId,
                     fix(fixer) {
                         const match = comment.value.match(LETTER_PATTERN);
+                        const char = match[0];
 
-                        return fixer.replaceTextRange(
+                        // Offset match.index by 2 to account for the first 2 characters that start the comment (// or /*)
+                        const charIndex = comment.range[0] + match.index + 2;
 
-                            // Offset match.index by 2 to account for the first 2 characters that start the comment (// or /*)
-                            [comment.range[0] + match.index + 2, comment.range[0] + match.index + 3],
-                            capitalize === "always" ? match[0].toLocaleUpperCase() : match[0].toLocaleLowerCase()
+                        return fixer.replaceTextRange(
+                            [charIndex, charIndex + char.length],
+                            capitalize === "always" ? char.toLocaleUpperCase() : char.toLocaleLowerCase()
                         );
                     }
                 });

diff --git a/lib/rules/utils/patterns/letters.js b/lib/rules/utils/patterns/letters.js
diff --git a/tests/lib/rules/capitalized-comments.js b/tests/lib/rules/capitalized-comments.js
@@ -367,6 +367,24 @@ ruleTester.run("capitalized-comments", rule, {
                 column: 1
             }]
         },
+        {
+            code: "// ꮳꮃꭹ",
+            output: "// Ꮳꮃꭹ",
+            errors: [{
+                messageId: "unexpectedLowercaseComment",
+                line: 1,
+                column: 1
+            }]
+        },
+        {
+            code: "/* 𐳡𐳡𐳡 */", // right-to-left-text
+            output: "/* 𐲡𐳡𐳡 */",
+            errors: [{
+                messageId: "unexpectedLowercaseComment",
+                line: 1,
+                column: 1
+            }]
+        },
 
         // Using "always" string option
         {
@@ -541,6 +559,26 @@ ruleTester.run("capitalized-comments", rule, {
                 column: 1
             }]
         },
+        {
+            code: "// Გ", // Georgian Mtavruli Capital Letter Gan (U+1C92)
+            output: "// გ", // Georgian Letter Gan (U+10D2)
+            options: ["never"],
+            errors: [{
+                messageId: "unexpectedUppercaseComment",
+                line: 1,
+                column: 1
+            }]
+        },
+        {
+            code: "// 𑢢", // Warang Citi Capital Letter Wi (U+118A2)
+            output: "// 𑣂", // Warang Citi Small Letter Wi (U+118C2)
+            options: ["never"],
+            errors: [{
+                messageId: "unexpectedUppercaseComment",
+                line: 1,
+                column: 1
+            }]
+        },
 
         // Default ignore words should be warned if there are non-whitespace characters in the way
         {