Sounds to me like you're accidentally declaring b.myC as a static, perhaps.

Actually, double check your examples (perhaps using the link report option
on MXMLC). It might be that you're not actually referencing B and it's not
becoming part of your SWF, in which case you're getting a "best guess"
object where you think you're getting an instance of B (you can't check for
"is B" without including it in your swf, but you can check for "!(is A)"
without including the B class.

In which case, it's the unmarshaller that's killing your data when trying to
convert it to a B (becuase now it knows to do it, once you've referenced the
class and it's in your SWF). Double check how you're defining C in your Flex
*and in your Java*.

Also, if it's the unmarshaller that's the problem, I'd definitely try a
compile against Flex 3.01 and see if the problem goes away. If you've got a
smallest-failiing case project for this bug it should only take you a few
minutes.

-Josh

On Tue, Sep 9, 2008 at 9:53 AM, dmiramontesval
<[EMAIL PROTECTED]>wrote:

> Ummmmm actually my example is wrong, i was missing something, let me
> explain again:
>
> I have 3 classes as a matter of fact, classes A and B are correct but
> i forgot to mention that inside B there is an object C:
>
> [RemoteClass(alias="com.test.B")]
> public class B {
> public var id:Number;
> public var name:String;
> public var myC:C
> }
>
> [RemoteClass(alias="com.test.C")]
> public class C {
> public var testVar1:Number;
> public var testVar2:String;
> }
>
>
> A objectA = new A();
> objectA.myMap = new HashMap();
> B objectB1 = new B();
> C objectC1 = new C();
> objectB1.id = 1;
> objectB1.name = "Test";
> objectC1.testVar1 = 2;
> objectC1.testVar2 = "Hello";
> objectB1.myC = objectC1;
> myMap.put("key1", objectB1);
>
>
> Now here is the deal, the only property being reset is the C object,
> the id and name properties are not reset
>
>
> for each(var i:* in objectA.myMap) {
>   if (i is B) {
>     trace(i.id + " " + i.name);
>      trace(i.myC.testVar1 + " " + i.myC.testVar2);
>   }
> }
>
> The first trace works fine, but the second trace shows NaN and empty
> string.
> If i change the code to:
>
> for each(var i:* in objectA.myMap) {
>   if (!(i is A)) {
>     trace(i.id + " " + i.name);
>      trace(i.myC.testVar1 + " " + i.myC.testVar2);
>   }
> }
>
> Both traces work fine, prints 1 Test and 2 Hello.
>
> Like you said i have some logic inside the B constructor:
>
> public function B() {
>   myC = new C();
> }
>
> I commented the constructor and instead of getting the values reset i
> got an error saying i tried to access a null reference, since C is
> null (if i don't use an instance of B then it works fine).
>
> So i am kinda back where i started, when i use the B class the C
> object inside each B object is reset.
>
>
> --- In flexcoders@yahoogroups.com, "Josh McDonald" <[EMAIL PROTECTED]> wrote:
> >
> > Right, I see. I suppose you may be doing something funky in the
> constructor
> > for your actual impl of A or B. Try commenting out the constructor
> and see
> > if you have the same problem.
> >
> > -Josh
> >
> > On Tue, Sep 9, 2008 at 9:17 AM, dmiramontesval
> > <[EMAIL PROTECTED]>wrote:
> >
> > > They keys in the Java map are Strings and the values are B objects,
> > > you could say it is HashMap<String, B>.
> > > So suppose in Java i do this:
> > >
> > > A objectA = new A();
> > > objectA.myMap = new HashMap();
> > > B objectB1 = new B();
> > > objectB1.id = 1;
> > > objectB1.name = "Test";
> > > myMap.put("key1", objectB1);
> > >
> > > In Flex if i iterate over the myMap property:
> > >
> > > for each(var i:* in objectA.myMap) {
> > >   if (i is B) {
> > >      trace(i.id + " " + i.name);
> > >   }
> > > }
> > >
> > > Doing this resets the B object inside myMap, so the trace prints NaN
> > > and empty string.
> > >
> > > If i dont use an instance of B:
> > >
> > > for each(var i:* in objectA.myMap) {
> > >   if (!(i is A)) {
> > >      trace(i.id + " " + i.name);
> > >   }
> > > }
> > >
> > > The trace prints 1 Test.
> > >
> > > The main issue is that if i create an instance of B anywhere in my
> > > code, the B object inside myMap resets.
> > >
> > > The second issue is that inside the myMap property there are two
> > > objects instead of just one, this additional object is a reference to
> > > A and its key is "owner", so the myMap property looks like:
> > >
> > > var myMap:Object = {owner:the A object, key1: the B object};
> > >
> > >
> > > I have no idea where this reference to the A object comes from, but i
> > > can live with that, the main issue is that i can't use the B class
> > > because if i do the B objects inside the map are reset.
> > >
> > >
> > > --- In flexcoders@yahoogroups.com, "Josh McDonald" <dznuts@> wrote:
> > > >
> > > > The problem I think is you're using the wrong data structures.
> > > >
> > > > If your Java map looks like this:
> > > >
> > > > HashMap<int, String> myMap;
> > > >
> > > > What you'll get in Flex isn't what you've got below, it's an
> object like
> > > > this:
> > > >
> > > > var mymap : Object =
> > > >   {
> > > >     1: "String 1",
> > > >     2: "String 2",
> > > >     79: "Blue"
> > > >   }
> > > >
> > > > Not the structure that you're trying to coerce it to based on the
> > > code you
> > > > posted. But since this is clearly just "example" Flex code and I
> have no
> > > > idea about what your Java looks like, I'm just guessing.
> > > >
> > > > -Josh
> > > >
> > > > On Tue, Sep 9, 2008 at 8:46 AM, dmiramontesval
> > > > <dmiramontesval@>wrote:
> > > >
> > > > > I have this weird issue and i can't figure it out, let me explain:
> > > > >
> > > > > I have a java object A which has a HashMap variable myMap,
> this map
> > > > > will hold objects of class B. In ActionScript i have the
> corresponding
> > > > > class A which has a variable myMap but since there is no
> HashMap in
> > > > > ActionScript its type is Object and i also have the corresponding
> > > class B.
> > > > >
> > > > > [RemoteClass(alias="com.test.A")]
> > > > > public class A {
> > > > >  public var myMap:Object; // this is the HashMap in Java
> > > > > }
> > > > >
> > > > > [RemoteClass(alias="com.test.B")]
> > > > > public class B {
> > > > >  public var id:Number;
> > > > >  public var name:String;
> > > > > }
> > > > >
> > > > > I need to put the objects B contained in the map inside a
> DataGrid so
> > > > > i am obtaining the object of class A using a RemoteObject and
> then i
> > > > > iterate over the myMap property using a for each in loop, like
> this:
> > > > >
> > > > > for each(var i:* in objectA.myMap) {
> > > > > someArrayCollection.add(i);
> > > > > }
> > > > >
> > > > > The first issue appears here, suppose the java HashMap contains 3
> > > > > elements, in the ActionScript loop i get 4 elements, the
> fourth one
> > > > > being a reference to the A object (its key appears as "owner"). In
> > > > > order to avoid this i use the "is" operator like this:
> > > > >
> > > > > for each(var i:* in objectA.myMap) {
> > > > >   if (i is B) {
> > > > >     someArrayCollection.add(i);
> > > > >   }
> > > > > }
> > > > >
> > > > > Now here is the real issue, when i do this the B objects
> contained in
> > > > > the myMap property are reset to their initial values (NaN and
> empty
> > > > > string according to the B class). I changed the if condition
> to this
> > > > > and worked:
> > > > >
> > > > > for each(var i:* in objectA.myMap) {
> > > > >   if (!(i is A)) {
> > > > >    someArrayCollection.add(i);
> > > > >   }
> > > > > }
> > > > > When i did this the DataGrid shows the elements in the
> > > > > someArrayCollection perfectly fine. I did a little more
> testing and i
> > > > > "figured" out the issue, if i use an instance of class B
> anywhere in
> > > > > my application, the values of the B objects inside the myMap
> property
> > > > > are reset, for example:
> > > > >
> > > > > for each(var i:* in objectA.myMap) {
> > > > >   if (i is B) {
> > > > >      someArrayCollection.add(i);
> > > > >   }
> > > > > }
> > > > > var test:B = new B();
> > > > >
> > > > > This resets the B objects inside the myMap property of the A
> object
> > > > > which makes no absolute sense, this is why when i used the B
> class in
> > > > > the if condition the objects reset their values.
> > > > >
> > > > > So the thing is whenever i use the B class, which is the class
> of the
> > > > > objects inside myMap, the B objects are reset!!!! Does anyone
> know how
> > > > > to fix this?????
> > > > >
> > > > > I am using Flex 2.01 hotfix 3, i wonder if this is some kind
> of bug
> > > > > that was solved in a later release.
> > > > >
> > > > > Any help would be appreciated.
> > > > >
> > > > >
> > > > > ------------------------------------
> > > > >
> > > > > --
> > > > > Flexcoders Mailing List
> > > > > FAQ:
> http://groups.yahoo.com/group/flexcoders/files/flexcodersFAQ.txt
> > > > > Search Archives:
> > > > > http://www.mail-archive.com/flexcoders%40yahoogroups.comYahoo!
> Groups
> > > > > Links
> > > > >
> > > > >
> > > > >
> > > > >
> > > >
> > > >
> > > > --
> > > > "Therefore, send not to know For whom the bell tolls. It tolls for
> > > thee."
> > > >
> > > > http://flex.joshmcdonald.info/
> > > >
> > > > :: Josh 'G-Funk' McDonald
> > > > :: 0437 221 380 :: josh@
> > > >
> > >
> > >
> > >
> > > ------------------------------------
> > >
> > > --
> > > Flexcoders Mailing List
> > > FAQ: http://groups.yahoo.com/group/flexcoders/files/flexcodersFAQ.txt
> > > Search Archives:
> > > http://www.mail-archive.com/flexcoders%40yahoogroups.comYahoo! Groups
> > > Links
> > >
> > >
> > >
> > >
> >
> >
> > --
> > "Therefore, send not to know For whom the bell tolls. It tolls for
> thee."
> >
> > http://flex.joshmcdonald.info/
> >
> > :: Josh 'G-Funk' McDonald
> > :: 0437 221 380 :: [EMAIL PROTECTED]
> >
>
>
>
> ------------------------------------
>
> --
> Flexcoders Mailing List
> FAQ: http://groups.yahoo.com/group/flexcoders/files/flexcodersFAQ.txt
> Search Archives:
> http://www.mail-archive.com/flexcoders%40yahoogroups.comYahoo! Groups
> Links
>
>
>
>


-- 
"Therefore, send not to know For whom the bell tolls. It tolls for thee."

http://flex.joshmcdonald.info/

:: Josh 'G-Funk' McDonald
:: 0437 221 380 :: [EMAIL PROTECTED]

Reply via email to