[Uml-devel] UML standard deviation (was curiosity)

tomas.walch at se.transport.bombardier.com tomas.walch at se.transport.bombardier.com
Thu Sep 19 07:37:02 UTC 2002


Hi!

One deviation I have found from the standard UML is the format of attributes
which are displayed in a C++ kind of way ( [visibility] [type]  name ) where the
correct UML notation is (quoting entire definition, even though everything isn't
supported in our yet-to-be-properly-named application) :

[visibility] name [multiplicity] [: type] [= initial-value] [{property-string}]

Operations deviate in that "void" return types show up in parsed code, it isn't
customary UML, and a related beaty flaw is that a colon is appended to
operations without any return type, like constructors.

Attached are updated versions of attribute.cpp and operation.cpp where I have
applied the standard notation to the toString()-methods. (I don't have
cvs-access and I wouldn't be able to use it anyway 'cause of a far too rigid
firewall at my office).

The classparser also adds qualyfiers like "virtual" to the return type, do we
wan't this? I guess it helps code generation, but it's more C++ than UML.

regards,

Tomas Walch.


(See attached file: operation.cpp)(See attached file: attribute.cpp)

-------------- next part --------------
A non-text attachment was scrubbed...
Name: operation.cpp
Type: application/octet-stream
Size: 6825 bytes
Desc: not available
URL: <http://mail.kde.org/pipermail/umbrello-devel/attachments/20020919/7bc3ea7f/attachment.obj>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: attribute.cpp
Type: application/octet-stream
Size: 4443 bytes
Desc: not available
URL: <http://mail.kde.org/pipermail/umbrello-devel/attachments/20020919/7bc3ea7f/attachment-0001.obj>


More information about the umbrello-devel mailing list