Updates:
Status: Accepted
Labels: Type-Bug
Comment #1 on issue 771 by [email protected]: Additional Unicode chars
missing per ECMA-262 3rd ed in Identifiers
http://code.google.com/p/v8/issues/detail?id=771
The Nl category is missing from the Letter predicate. This will be fixed.
The two symbols reported as Lm (U+02EC and U+0374) appears to have category
Sk in Unicode 5.0.
I cannot reproduce the Lo category errors. The following code reports no
errors - every character is correctly recognized as an IdentifierStart:
function testVar(varstring) {
return eval("var " + varstring + " = 42; " +
eval("'"+varstring+"'") + " == 42;");
}
var LoFails = [
"\\uA800",
// ... all the Lo code points reported above ...
"\\uA873",
];
for (var i = 0; i < LoFails.length; i++) {
if (!testVar(LoFails[i])) alert("Not valid: " + LoFails[i]);
}
Ditto for U+1EEA, which is recognized correctly.
How are you testing this: Is the source UTF-8 input or V8 strings?
--
v8-dev mailing list
[email protected]
http://groups.google.com/group/v8-dev
