Add some support for the code-point code-unit distinction.

sdk/lib/_internal/compiler/implementation/lib/js_helper.dart

 // Copyright (c) 2012, the Dart project authors.  Please see the AUTHORS file
 // for details. All rights reserved. Use of this source code is governed by a
 // BSD-style license that can be found in the LICENSE file.
 
 #library('dart:_js_helper');
 
 #import('dart:coreimpl');
 #import('dart:collection');
 
 #source('constant_map.dart');
@@ skipping 516 matching lines @@
     if ((length is !int) || (length < 0)) {
       throw new ArgumentError(length);
     }
     var result = JS('Object', r'new Array(#)', length);
     JS('void', r'#.fixed$length = #', result, true);
     return result;
   }
 
   static num dateNow() => JS('num', r'Date.now()');
 
+  static const MAX_CODE_UNIT = 0xffff;
+  static const MAX_CODE_POINT = 0x10ffff;
+  static const SURROGATE_BASE = 0xd800;
+  static const SURROGATE_END = 0xdfff;
+  static const LEAD_BASE = 0xd800;
+  static const TRAIL_BASE = 0xdc00;
+  static const SURROGATE_MASK = 0x3ff;
+  static const SURROGATE_ENCODING_BASE = 0x10000;
+
   static String stringFromCharCodes(charCodes) {

[floitsch, 2012/11/08 15:28:21]

I would move this code to the patch-class and redirect to fromCodeUnits there.
Maybe ask the others first what they think.

[ngeoffray, 2012/11/15 13:24:20]

I don't think the patch classes have access to 'JS', so I would keep the method
where it is. The less libraries that touch JS, the better.

+    bool surrogatePairs = false;
     for (var i in charCodes) {
       if (i is !int) throw new ArgumentError(i);
+      if (i > MAX_CODE_POINT) throw new ArgumentError(i);
+      if (i > MAX_CODE_UNIT) surrogatePairs = true;
+      if (i >= SURROGATE_BASE && i <= SURROGATE_END) throw new ArgumentError(i);
     }
-    return JS('String', r'String.fromCharCode.apply(#, #)', null, charCodes);
+    if (!surrogatePairs) {
+      return JS('String', r'String.fromCharCode.apply(#, #)', null, charCodes);
+    }
+    var expanded = [];
+    for (var i in charCodes) {
+      if (i <= MAX_CODE_UNIT) {
+        expanded.add(i);
+      } else {
+        expanded.add(LEAD_BASE +
+                     (((i - SURROGATE_ENCODING_BASE) >> 10) & SURROGATE_MASK));
+        expanded.add(TRAIL_BASE + (i & SURROGATE_MASK));
+      }
+    }
+    return JS('String', r'String.fromCharCode.apply(#, #)', null, expanded);
+  }
+
+  static String stringFromCodeUnits(codeUnits) {
+    for (var i in codeUnits) {
+      if (i is !int) throw new ArgumentError(i);
+      if (i > MAX_CODE_UNIT) throw new ArgumentError(i);
+    }
+    return JS('String', r'String.fromCharCode.apply(#, #)', null, codeUnits);
   }
 
   static String getTimeZoneName(receiver) {
     // When calling toString on a Date it will emit the timezone in parenthesis.
     // Example: "Wed May 16 2012 21:13:00 GMT+0200 (CEST)".
     // We extract this name using a regexp.
     var d = lazyAsJsDate(receiver);
     return JS('String', r'/\((.*)\)/.exec(#.toString())[1]', d);
   }
 
@@ skipping 925 matching lines @@
     result = new TypeImpl(key);
     JS('var', r'#[#] = #', runtimeTypeCache, key, result);
   }
   return result;
 }
 
 String getRuntimeTypeString(var object) {
   var typeInfo = JS('Object', r'#.builtin$typeInfo', object);
   return JS('String', r'#.runtimeType', typeInfo);
 }