2.2 Setting the GigaSystemBuilder Preferences

2.2.1  The GigaSystemBuilder Preferences

You do this via the overall preferences page by clicking

Window -> Preferences

Select the GigaSystemBuilder preferences on the left-hand side:

1. JeeWiz Home

   This is the directory of the JeeWiz installation (i.e. underneath there will be lib and resources directories). A default value is provided, which points to a run-time copy of JeeWiz embedded within the builder plugin. There is no need to alter this value unless you install a standalone version of JeeWiz onto your system and want to use that.

   This field sets the 'jwhome' property for JeeWiz builds.

2. GigaSpaces Home

   Initially this value will be blank: there is no default. This value is required if you want to compile or run the generated GigaSpaces application; it is not required just to do modelling. Point this to the home directory of your GigaSpaces installation, e.g. C:\GigaSpaces.

3. Java Home

   This is the location of your JDK. This must be version 1.5 or above.

   This field sets the 'JAVA_HOME' property for JeeWiz builds.

4. Custom Template Directory

   You don't need to worry about this at first. Having used the GigaSystemBuilder, you (or your architect) may want to change the way it works. If you have some custom templates that apply to all projects, you can put them in a directory and point to that directory in this entry. When you do this, the templates in this directory will override the corresponding ones in the standard product.

   The GigaSystemBuilder uses a stack of template directories, which will all be beneath the resources/ directory of the JeeWiz Home property described above. The complete list is

   ${customTemplateDirectory}
   resources/giga-spaces/control
   resources/roundTripping/control
   resources/bizobject/control
   resources/java/control
   resources/convert-base/control
   resources/base/control

   This will at least give you a place to start looking. If you want to implement a specific customisation, post a question on the GigaSystemBuilder forum.

   You can also use the customization feature for an individual project by setting the customTemplateDirectory property for an application object in the model; see the customTemplateDirectory property documentation.

2.2.2  The Generate-Time Settings

Window -> Preferences -> GigaSpaces Preferences -> Generate Time Settings

There are a number of preferences that can be amended in this section. They are

1. Ant Log Level

   The engine logs significant events (e.g., directory creation, or generation of the final output file) by default.

   Sets the level of logging that will be output during the generation; the levels are defined by the Ant log level. The default is 'Info' level; 'Verbose' and 'Debug' give increasing levels of logging.

   The Verbose level is sufficient to diagnose most problems. The Debug level produces much more output, and is a last resort.

   By default the log is sent to the console window. For verbose and debug levels, you will probably need to send the log output to a text file via the 'Create Log File' option.

   This field sets the '-verbose' and '-debug' flags to the invoked Ant. This will normally be necessary if Verbose or Debug generate logging levels are used.

2. Create Log File

   Sends the logging from the generation process to an output file, rather than to the console.

3. Log File

   This defines the file that logging is to be sent to, if the "Create Log File" tickbox is checked. A relative file specification will be relative to the Eclipse project directory.

4. Generate Log Level

   This defines the generated log level in the code. The options are:

   • Info This level of logging is used for significant non-error events, like the application being ready to start.
   • Debug This is used for normal events in main-line processing code, like a file being successfully opened or an entry being written into a space.
   • Detailed Debug This is used for highly detailed events. This will record all method entries and exits, and (if a method has requested it) log the return value from methods that return a value. From detailed debug level trace, it should be possible to trace most of the flow of an application
5. Create Aggregate Specification File

   If set, JeeWiz will create a file containing the aggregate specification. The "aggregate specification" is the original specification, in JeeWiz simple XML format, plus any XML added by executable patterns. This gives a complete 'platform-specific model' plus any embedded code.

   In addition, the aggregate specification file contains cross-reference information on the templates used, the objects firing executable patterns and the objects they create, and the Velocity methods declared on model objects.

   Getting the aggregate specification file is useful for pattern developers to debug patterns.

6. Create Aggregate Specification File

   This is the path for the aggregate specification file. This fields sets the 'dump' property to the invoked Ant.

7. Trace Parse

   When #parse() is being traced, JeeWiz adds comments to the Java and XML files it generates, identifying which templates generated the files. This is particularly useful for debugging generation scripts when the main generation template calls to other templates (via the #parse() construct in Velocity). If this field is set, the JeeWiz engine 'traceParse' property is set to "true"; otherwise it is left unset.