Mia-Generation is used to build code generators complying with specific architectures and frameworks.Generation rules are specified with a full IDE dedicated to generator building (template-based generation engine).

It provides a dynamic bridge to developers between analysis and design, and supports iterative cycles by preserving handcoded code fragments between two generations.

Features:

§        Full support of UML, but open to any other meta-model

§                              Open to any file format for input models : can be connected to major UML1 modeling tools (Rose™, Rhapsody™, Together™, Poseidon™, and any XMI compliant tool) and UML2 tools (IBM RSM/RSA™, No Magic MagicDraw™, and any tool based on UML2 Eclipse).

§        Template based generation engine

§        Java as scripts langage

§        Built-in support for UML Profiles

§        Dedicated development environment

§        Facilities for generators deployment

§        Integration as plugin in Eclipse, MagicDraw and Maven

§        Modular architecture

§        Support for iterative development cycles (can preserve hand-written code)

Mia-Generation provides the developers with very usefull information about what has been generated, indicates the code fragments to be completed manually and triggers alerts when anomalies ared detected during the generation. MiaGeneration allows to get more precise information about the status of the generated files and the fragments that constitute them. The generated file status is explicit thanks to various icons. The user can detect easily the new files, the modified files those containing some manual code. The information is relayed on a finer level, on the code fragments, which allows to know real impacts of the modifications provided by the last generation.

Mia-Generation allows to show the different code fragments constituting a generated file, notably thanks to different background colors in the text editor. Each selected fragment in the navigator tree appears surrounded in the text editor to facilitate reading and search. When a generated file is different compared to the previous generation, modified code fragments are showed in a comparator.

By default, Mia-Generation respects UML 1.4. If specific constraints are defined, a report generator can be built to double check the Model. This report generator can be run along with the code generation and if an anomaly is detected, a warning is issued and the generation can be sfaq_topped.

Assessing to improve management:

For each executed scenario, a table is displayed with different metrics for each generated file :

-         file size (in bytes)

-         user-code size (in bytes)

-         percentage of generated code (based on sizes)

-         number of user-code parts

-         number of user-code parts written by the user

-         generation time

-         generation time percentage

Benefits:

Mia-Generation is a user-friendly environment to create your own specific generator, fully compliant with your architecture. It means Mia-Generation is multi-target and works for any code. The limit is only the existence of transformation rules from model to code.

Therefore, Mia-Generation is already put in place by our clients to produce various targets such as :

-         Java/JEE : Java, EJB, XML, HTML, JSP, STRUTS, JSF, HIBERNATE, SPRING.

-         Microsoft : C#, ASP .NET, XML, SOAP, HTML, VB 5, VB 6, ASP.

-         Database : DDL, SQL.

-         Others technologies : PHP, C, C++, DELPHI, ADA 95, VHDL, COBOL.