Developing VDM ++ operations from UML diagrams

Mit, E

Abstract

The Unified Modelling Language (UML) is widely accepted as the standard notation
for Object-Oriented methods. However, one of its major problem is its
lack of formal semantics. The development of the Object Constraints Language
(OCL) in the last few years seems to be the answer to this problem, but some
argue that it is difficult to use. In addition, inadequate information and lack
of precise definitions of UML models [31] are the two main reasons why current
existing OO supporting tools do not emphasis on the standard input syntax or
format used with UML models.
On the other hand, Formal Methods (FM) are seen as a rigourous way to
develop software specification, but they are difficult and expensive to use. The
integration of OO and FM attempts to gain the ease of use of OO methods and
the rigour of FM. However, current work on the integration of OO methods and
FM and the available supporting tools are still ineffective.
In this thesis, the author attempts to provide answers to some of these questions
by developing a set of transformation rules from UML to VDM++ and
then develops an architecture for a CASE tool to support the transformation
process. The author first increments the UML models by the missing information
required to produce full VDM++ specifications, and uses an intermediate
Extensible Markup Language (XML) representation to save the UML models.
Later, a set of translation rules are defined and based on these rules, VDM++
specifications are generated by mapping the tag values of the XML representation
to the corresponding elements of VDM+-f- specifications.
The prototype tool (FOTool] that supports this transformation process has
been developed. It emphasises the syntax and the format of the input to the UML
models and the mandatory fields, which are essential for generating complete
VDM++ specifications.

• Thesis Type: Thesis
• Deposit Date: Aug 19, 2021
• Award Date: Jan 1, 2008