SDMetrics V2.1 Javadocs

SDMetrics - the design measurement tool for the UML™.

See:
Description

Packages

sdmetrics	The top-level package only contains classes to launch the GUI, start the command line processing, or just display the help window.
sdmetrics.app	This package contains classes to calculate a set of metrics for an XMI file, check design rules, compare two designs, calculate relation matrices, and access and export the data and graphs to external files.
sdmetrics.model	This package implements the UML model of SDMetrics, and provides the XMI import.

SDMetrics - the design measurement tool for the UML™.

Where to get started:

To integrate SDMetrics's measurement capabilities in your application:

• Read Chapters 4 and 5 of the SDMetrics User Manual to become familiar with basic SDMetrics concepts (project files, filter settings, metric data tables, descriptive statistics, design rule checking, relation matrices, design comparisons, command line options).
• The classes you need to get the work done are all in package sdmetrics.app. The description of that package contains a tutorial showing you how to fill a data structure ProjectSettings, passing it to a Controller object which takes care of all the work, and how to obtain the measurement data, design rule violations, and so on from that controller object.
• All long running calculation processes report progress and errors through the SDMessageHandler interface. You can supply your own message handler to the controller to integrate these progress and error messages in your application.
• The controller obtains application parameters (default project files, percentiles) through the AppParameters interface. Class MyParameters is an implementation of this interface that provides "factory settings" and stores the application parameters in a properties file. You can provide your own implementation of this interface, with factory settings and persistence tailored to your application.
• From the controller object, you can obtain the measurement data, descriptive statistics, rule checking, and relation matrices. All this data is presented in the form of tables, and can be accessed through a uniform interface: DataTable. For example, class DataExporter uses this interface to output the metric data tables, descriptive statistics, design rule violations etc.

To add new metrics (or rules or relation matrices):

• Read Chapters 7 and 8 of the SDMetrics User Manual to become familiar with SDMetrics project file formats for metamodel definitions, XMI transformations, and metric definition files.
• Most of the time, you can just add the definition of the new metric to the metric definition file, without having to change anything else.
• If your new metric requires information that is currently not contained in SDMetrics' UML metamodel
  • extend the metamodel definition file with the new metamodel elements or attributes that are needed,
  • add XMI transformations and triggers for your new metamodel elements and attributes to the XMI transformation files.
  • Add the metrics to the metric definition files.
    
• If your new metrics cannot be expressed using the capabilities of the metric definition file formalism, an extension of the metrics engine is required. The overall procedure is described in the documentation for package sdmetrics.metrics.