Fix SCSSParsingError on unicode characters in SCSS selectors (#14065)

scss-parser@1.0.6 tokenizer uses ASCII-only regex for identifiers,
rejecting valid non-ASCII CSS characters (e.g., #présentation).
Encode non-ASCII characters as ASCII codepoint placeholders before
parsing since the parser is only used for variable analysis, not
CSS generation.

Fixes #14065

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

Author: cderv

src/core/sass/analyzer/parse.ts

@@ -41,6 +41,14 @@ export const makeParserModule = (
         "$1: $2",
       );
 
+      // scss-parser's tokenizer only handles ASCII identifier characters.
+      // Encode non-ASCII characters as ASCII codepoint placeholders since the
+      // parser is only used for variable analysis, not CSS generation.
+      contents = contents.replaceAll(
+        /[^\x00-\x7F]/g,
+        (ch) => `_u${ch.codePointAt(0)!.toString(16)}_`,
+      );
+
       // This is relatively painful, because unfortunately the error message of scss-parser
       // is not helpful.
 

tests/docs/smoke-all/2026/02/20/custom-14065.scss

@@ -0,0 +1,5 @@
+/*-- scss:rules --*/
+
+#présentation p {
+  line-height: 2;
+}

tests/docs/smoke-all/2026/02/20/issue-14065.qmd

@@ -0,0 +1,17 @@
+---
+title: "Unicode SCSS Test"
+format:
+  html:
+    theme:
+      - default
+      - custom-14065.scss
+_quarto:
+  tests:
+    html:
+      ensureCssRegexMatches:
+        - ['#présentation p', '--quarto-scss-export-']
+---
+
+## Présentation
+
+Unicode characters in headings become CSS selectors.