|
Re: LiteralSpecification [message #477675 is a reply to message #477665] |
Mon, 11 August 2008 22:59 |
Eclipse User |
|
|
|
Originally posted by: cdamus.zeligsoft.com
Hi, Timothy,
The LiteralSpecification metaclass inherits a stringValue() operation
from the ValueSpecification metaclass. It converts the value to a string.
Does that help?
The problem with defining a value attribute that would be redefined by
the subclasses is that UML doesn't have a primitive type that it could
use in defining a LiteralSpecification::value. There is no primitive
type that UML's four standard types conform to. Moreover, because
primitive types are value-based, not identity-based, you can't provide a
useful general type, in contrast with identity (reference) types where
you could define Object as the most general type.
Cheers,
Christian
Timothy Marc wrote:
> Hi all,
>
> i have a question concerning the LiteralSpecification abstract class. I
> think it should have a value-Attribute, that has a general type and the
> subclasses should redefine that value attribute to the specific type. I
> don't know, whether it is possible to provide such a general type for the
> value-Attribute, but IMHO, that would provide a better programatically
> access to the values of a LiteralSpecification subclass.
>
> What are your opinions for that topic?
>
> Thx
> Timothy
>
>
|
|
|
Re: LiteralSpecification [message #626867 is a reply to message #477665] |
Mon, 11 August 2008 22:59 |
Eclipse User |
|
|
|
Originally posted by: cdamus.zeligsoft.com
Hi, Timothy,
The LiteralSpecification metaclass inherits a stringValue() operation
from the ValueSpecification metaclass. It converts the value to a string.
Does that help?
The problem with defining a value attribute that would be redefined by
the subclasses is that UML doesn't have a primitive type that it could
use in defining a LiteralSpecification::value. There is no primitive
type that UML's four standard types conform to. Moreover, because
primitive types are value-based, not identity-based, you can't provide a
useful general type, in contrast with identity (reference) types where
you could define Object as the most general type.
Cheers,
Christian
Timothy Marc wrote:
> Hi all,
>
> i have a question concerning the LiteralSpecification abstract class. I
> think it should have a value-Attribute, that has a general type and the
> subclasses should redefine that value attribute to the specific type. I
> don't know, whether it is possible to provide such a general type for the
> value-Attribute, but IMHO, that would provide a better programatically
> access to the values of a LiteralSpecification subclass.
>
> What are your opinions for that topic?
>
> Thx
> Timothy
>
>
|
|
|
Powered by
FUDForum. Page generated in 0.03779 seconds