From 93f3b2b114877af17db219e501ae4551df61738d Mon Sep 17 00:00:00 2001
From: "Alex Lam S.L" <alexlamsl@gmail.com>
Date: Mon, 11 Dec 2017 01:15:44 +0800
Subject: escape consecutive unpaired surrogates (#2576)

fixes #2569
---
 lib/output.js | 15 ++++++++++-----
 lib/parse.js  | 12 ++++++++++++
 2 files changed, 22 insertions(+), 5 deletions(-)

(limited to 'lib')

diff --git a/lib/output.js b/lib/output.js
index 1aa63450..a4c41f11 100644
--- a/lib/output.js
+++ b/lib/output.js
@@ -121,11 +121,16 @@ function OutputStream(options) {
             }
         });
     } : function(str) {
-        return str.replace(/[\ud800-\udbff](?![\udc00-\udfff])/g, function(ch) {
-            return "\\u" + ch.charCodeAt(0).toString(16);
-        }).replace(/(^|[^\ud800-\udbff])([\udc00-\udfff])/g, function(match, prefix, ch) {
-            return prefix + "\\u" + ch.charCodeAt(0).toString(16);
-        });
+        var s = "";
+        for (var i = 0, len = str.length; i < len; i++) {
+            if (is_surrogate_pair_head(str[i]) && !is_surrogate_pair_tail(str[i + 1])
+                || is_surrogate_pair_tail(str[i]) && !is_surrogate_pair_head(str[i - 1])) {
+                s += "\\u" + str.charCodeAt(i).toString(16);
+            } else {
+                s += str[i];
+            }
+        }
+        return s;
     };
 
     function make_string(str, quote) {
diff --git a/lib/parse.js b/lib/parse.js
index 099fc49a..f0098c75 100644
--- a/lib/parse.js
+++ b/lib/parse.js
@@ -132,6 +132,18 @@ function is_letter(code) {
         || (code >= 0xaa && UNICODE.letter.test(String.fromCharCode(code)));
 };
 
+function is_surrogate_pair_head(code) {
+    if (typeof code == "string")
+        code = code.charCodeAt(0);
+    return code >= 0xd800 && code <= 0xdbff;
+}
+
+function is_surrogate_pair_tail(code) {
+    if (typeof code == "string")
+        code = code.charCodeAt(0);
+    return code >= 0xdc00 && code <= 0xdfff;
+}
+
 function is_digit(code) {
     return code >= 48 && code <= 57;
 };
-- 
cgit v1.2.3