Tips for designing model reports

You can use the capabilities of your IBM® Rational® product in conjunction with those of Business Intelligence and Reporting Tools (BIRT) to fine-tune data sources, data sets, and report designs for UML and EMF models. The following tips provide helpful information for designing model reports, as well as links to sites where you can find additional information.
Working through the report design process with a sample model
For a step-by-step demonstration of how to design a report, see the tutorial named "Create a BIRT report design for a UML model." The tutorial shows how to create the report design and add commonly used model elements to the report.
Adding common types of model information to reports
To include information in a report, you must add the corresponding element to the report design. For certain types of information that are commonly included in reports, such as model element documentation, you must follow special procedures. For more information, see the topic "Adding common types of model information to reports."
Finding the available custom XPath functions
For the XPath functions that IBM Rational provides, see the topic "Additional XPath functions for EMF and UML data sets." When these functions are not sufficient to build the data set that you need, you can write your own XPath functions and use the extension point named org.eclipse.jet.xpathFunctions to register them. For information, see the topic "Writing custom XPath functions."
See also the context help for the Column Mapping page of the Data Set wizard, where you specify XPath functions.
Troubleshooting XPath expression
You might encounter XPath expressions that appear valid in the Data Set dialog box, but that do not yield any results when you preview them.
The Data Set dialog box only validates XPath syntax. It does not catch errors such as a reference to navigation paths that do not exist in the model.
Joining data sets
If you join data sets and need a key to uniquely identify model objects, the best key is the URI of the model object, because it is unique across resources. The getURI() XPath function returns the URI of any object in the model. The resolveURI() function resolves URIs.
For example, if you have a Class data set that has a URI columnn (getURI(.)) and an Operation data set that has a containerURI column (getURI(..)), you can join the two data sets using the URI and containerURI.
For additional information, see the topic "How to join data sets" in the Field Guide to BIRT.
Preventing prompts to load unavailable fonts when reports are generated
If report users receive prompts to load unavailable fonts when they generate a report, modify the report design by setting the font to "automatic" or by specifying a font in the Properties view for each element. Do not use a default font such as default serif.
For example, if a report requires Japanese text, the use of default serif in the report design can result in an unexpected prompt to download a Chinese language font.
Selecting the best instance models for designing a report
When you design a report, make sure that the instance models that you use at design time are representative of your data, yet small enough to not impede the speed of your report design tasks. The models might be reloaded multiple times throughout the design stage, so using smaller models saves time.
Alternate approaches for starting report design
In addition to starting a report design using the Report Explorer window, you can use the BIRT Report wizard.
To open the Report Explorer window, click Window > Show View > Other, expand Reporting, and double-click Report Explorer.
To open the BIRT New Report wizard, click File > New, expand Business Intelligence Reporting Tools, and double-click Report.
Use the Report Explorer window if you are basing your report on an existing report design.
Deploying a BIRT report as a plug-in
To deploy a BIRT report as a plug-in, create a plug-in that contains the report and that extends the extension point named in the plug-in.
To create the plug-in: start a new plug-in project and ensure that the activator is in the /src folder. Add the RPTDESIGN file that contains the report to the project, and set the build configuration for the project to include the report in the output. Ensure that the build configuration includes the bin directory in the binary build. Add as a dependency, and set as an extension point for the plug-in. To the extension point, add a category for the report, and then add the report itself to the category. For the URL for the report, use a platform plug-in path: platform:/plugin/<plug-in name>/<path to RPTDESIGN file>.
To deploy the plug-in, export the plug-in, copy the output JAR file to the SDP\plugins folder of your IBM Rational product installation directory, and restart the product.
Note: Some information, such as links to Eclipse documentation or to developer guides, is available only from the help topics that are installed with the product.