WriteMultiByte and "unicode"

[Rob_Knopf]

I think I may be misunderstanding the use, but I'm using WriteMultiByte as follows for a network packet:


var ba:ByteArray = new ByteArray();
ba.writeMultiByte("Bob", "unicode");
ba.writeMultiByte("Jones", "unicode");

// view the buffer for debugging
for (var i:int = 0; i < ba.length; i++)
{
trace(ba );
}



I expected the buffer length to be 16 bytes (BobJones = 8 letters, 2 bytes per), but instead it's 8 bytes (1 byte per letter, UTF-8). I also expected to see zeros in between each letter when I trace the buffer.

What am I doing wrong?

[ShuShenghong]

I also see this problem today in cs3!!!

var src:String = String.fromCharCode(0XCCEC);
trace(src.charCodeAt(0).toString(16));//ccec
var tmpbytearr:ByteArray = new ByteArray();
tmpbytearr.writeMultiByte(src, "unicode");
trace(tmpbytearr.length);//1
trace(tmpbytearr.readUnsignedByte().toString(16));//3f

[ntbdy]

quote:

---

Originally posted by: Alkahest
ba.writeMultiByte("Bob", "unicode");

---

yes, indeed. And the compiler will not complain if you do:

ba.writeMultiByte("Bob", "anything whatsoever");

it just happily puts out UTF-8 regardless and ignores the nonsense argument. I don't know the web address to report bugs. I guess the authors just let that one slip by, or didn't have time to implement it.

[Rob_Knopf]

Hrm.. so you are saying that it ignores the second argument, or are you saing that "unicode" not the correct value I should be passing? I've tried just about every combination from the docs...

[ntbdy]

Alkahest, I tried a nonsense argument thinking that if no error is thrown, then the argument is likely being ignored. That was instead of trying them all. I previously only tried the alias to 'unicode' (which was 'utf-16'), and that hadn't worked either.

I don't think we can see the source anywhere. That would determine for sure that the argument is ignored, at least for now until that gets implemented in an update.