Re: svn commit: r956086 - in /click/trunk/click/framework/src/org/apache/click/util: MessagesMap.java SessionMap.java

[Bob Schellink <sa...@gmail.com>]

On 19/06/2010 04:37, bckfnn@apache.org wrote:

> also fix a wrong values in SessionMap.entrySet()


Seems SessionMap#entrySet should have been Map#values? Guess we can move that code under #values
instead of throwing exception.

I think the nested Entry class will give issues inside Velocity since it's not public. Can you confirm?

regards

Bob

Re: svn commit: r956086 - in /click/trunk/click/framework/src/org/apache/click/util: MessagesMap.java SessionMap.java

[Bob Schellink <sa...@gmail.com>]

On 21/06/2010 06:52, Finn Bock wrote:
> No, it is ok. HashMap.Entry is defined the same way. As long as the
> object is only used as a Map.Entry, not as a SessionMap.Entry
> instance, then all is fine. I have added a test for this.


Appreciate the tests you've been adding. I'm starting to feel guilty for not contributing more in
this area.

Kind regards

Bob

Re: svn commit: r956086 - in /click/trunk/click/framework/src/org/apache/click/util: MessagesMap.java SessionMap.java

[Finn Bock <bc...@gmail.com>]

2010/6/19 Bob Schellink
> On 19/06/2010 04:37, bckfnn@apache.org wrote:
>
> Seems SessionMap#entrySet should have been Map#values? Guess we can move that code under #values
> instead of throwing exception.

Done.

> I think the nested Entry class will give issues inside Velocity since it's not public. Can you confirm?

No, it is ok. HashMap.Entry is defined the same way. As long as the
object is only used as a Map.Entry, not as a SessionMap.Entry
instance, then all is fine. I have added a test for this.

regards,
Finn