From ff1734f7c4882fb350cb0e1f366f04ce63675643 Mon Sep 17 00:00:00 2001
From: Erik Demaine <edemaine@mit.edu>
Date: Sat, 28 Aug 2021 18:55:05 -0400
Subject: [PATCH] fix: \char support for >16-bit Unicode characters (#3006)

* fix: \char support for >16-bit Unicode characters

`String.fromCharCode` only supports Unicode characters up to 16-bit.
`String.fromCodePoint` from ECMAScript 2015 supports all Unicode code points.
Unfortunately, IE doesn't support the latter, so use former as fallback.

Fixes #3004

Co-authored-by: ylemkimon <y@ylem.kim>
---
 src/functions/char.js | 15 +++++++++++++--
 test/katex-spec.js    |  6 ++++++
 2 files changed, 19 insertions(+), 2 deletions(-)

diff --git a/src/functions/char.js b/src/functions/char.js
index ed1e413b..1600ffbf 100644
--- a/src/functions/char.js
+++ b/src/functions/char.js
@@ -21,14 +21,25 @@ defineFunction({
             const node = assertNodeType(group[i], "textord");
             number += node.text;
         }
-        const code = parseInt(number);
+        let code = parseInt(number);
+        let text;
         if (isNaN(code)) {
             throw new ParseError(`\\@char has non-numeric argument ${number}`);
+        // If we drop IE support, the following code could be replaced with
+        // text = String.fromCodePoint(code)
+        } else if (code < 0 || code >= 0x10ffff) {
+            throw new ParseError(`\\@char with invalid code point ${number}`);
+        } else if (code <= 0xffff) {
+            text = String.fromCharCode(code);
+        } else { // Astral code point; split into surrogate halves
+            code -= 0x10000;
+            text = String.fromCharCode((code >> 10) + 0xd800,
+                                       (code & 0x3ff) + 0xdc00);
         }
         return {
             type: "textord",
             mode: parser.mode,
-            text: String.fromCharCode(code),
+            text: text,
         };
     },
 });
diff --git a/test/katex-spec.js b/test/katex-spec.js
index c926229b..ed521d77 100644
--- a/test/katex-spec.js
+++ b/test/katex-spec.js
@@ -3313,6 +3313,12 @@ describe("A macro expander", function() {
         expect(parsedChar[0].type).toEqual("textord");
     });
 
+    it("\\char handles >16-bit characters", () => {
+        const parsed = getParsed('\\char"1d7d9');
+        expect(parsed[0].type).toEqual("textord");
+        expect(parsed[0].text).toEqual("𝟙");
+    });
+
     it("should build Unicode private area characters", function() {
         expect`\gvertneqq\lvertneqq\ngeqq\ngeqslant\nleqq`.toBuild();
         expect`\nleqslant\nshortmid\nshortparallel\varsubsetneq`.toBuild();