improve unicode handling (#3648)

2 files changed, 12 insertions, 11 deletions

diff --git a/lib/output.js b/lib/output.js
index cc39c2df..38b4f8bc 100644
--- a/lib/output.js
+++ b/lib/output.js
@@ -119,15 +119,20 @@ function OutputStream(options) {
         });
     } : function(str) {
         var s = "";
-        for (var i = 0; i < str.length; 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];
+        for (var i = 0, j = 0; i < str.length; i++) {
+            var code = str.charCodeAt(i);
+            if (is_surrogate_pair_head(code)) {
+                if (is_surrogate_pair_tail(str.charCodeAt(i + 1))) {
+                    i++;
+                    continue;
+                }
+            } else if (!is_surrogate_pair_tail(code)) {
+                continue;
             }
+            s += str.slice(j, i) + "\\u" + code.toString(16);
+            j = i + 1;
         }
-        return s;
+        return j == 0 ? str : s + str.slice(j);
     };
 
     function make_string(str, quote) {
diff --git a/lib/parse.js b/lib/parse.js
index 740ef5f2..270af9b4 100644
--- a/lib/parse.js
+++ b/lib/parse.js
@@ -133,14 +133,10 @@ function is_letter(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;
 }