Allow non-latin characters in slugs (#32232)

* fixes #24710

* Additional test

* typo

* fix failing tests

* update test description

* update docs

packages/editor/src/utils/test/url.js

@@ -8,6 +8,10 @@ describe( 'cleanForSlug()', () => {
 		expect( cleanForSlug( '/Is th@t Déjà_vu? ' ) ).toBe( 'is-tht-deja_vu' );
 	} );
 
+	it( 'Should allow non-latin characters', () => {
+		expect( cleanForSlug( 'Καλημέρα Κόσμε' ) ).toBe( 'καλημέρα-κόσμε' );
+	} );
+
 	it( 'Should return an empty string for missing argument', () => {
 		expect( cleanForSlug() ).toBe( '' );
 	} );

packages/editor/src/utils/url.js

@@ -28,11 +28,11 @@ export function getWPAdminURL( page, query ) {
  * This replicates some of what sanitize_title() does in WordPress core, but
  * is only designed to approximate what the slug will be.
  *
- * Converts Latin-1 Supplement and Latin Extended-A letters to basic Latin
- * letters. Removes combining diacritical marks. Converts whitespace, periods,
+ * Converts Latin-1 Supplement and Latin Extended-A letters to basic Latin letters.
+ * Removes combining diacritical marks. Converts whitespace, periods,
  * and forward slashes to hyphens. Removes any remaining non-word characters
- * except hyphens. Converts remaining string to lowercase. It does not account
- * for octets, HTML entities, or other encoded characters.
+ * except hyphens and underscores. Converts remaining string to lowercase.
+ * It does not account for octets, HTML entities, or other encoded characters.
  *
  * @param {string} string Title or slug to be processed
  *
@@ -45,7 +45,7 @@ export function cleanForSlug( string ) {
 	return trim(
 		deburr( string )
 			.replace( /[\s\./]+/g, '-' )
-			.replace( /[^\w-]+/g, '' )
+			.replace( /[^\p{L}\p{N}_-]+/gu, '' )
 			.toLowerCase(),
 		'-'
 	);