32
<DisplayInfo
labelText="First Copy Columns"
defaultLabelText="First Copy Columns"
smallIcon="default_small.gif"
largeIcon="default_large.gif" />
</ActivityNode>
</ExplorerNodeList>
</ExplorerFolder>
</ExplorerNodeList>
</ExplorerPage>
</ExplorerTree>
To put additional nodes in the folder, include the additional
ActivityNodeModel elements at the same level as the one above. For
multiple folders, include additional
ExplorerFolder elements at the
same level as the one above.
Save this description in a file named extension.iml in the extension
directory. An XML file with all three descriptions of the copy nodes is
available in the programming examples directory.
Try the Node If you have successfully compiled the code, placed it in a properly
located jar file, and created the XML description of the examples
library, you’ll now be able to try the new node.
Start Spotfire Miner. The Explorer should now have a page titled
Programming Examples containing a folder with the First Copy
Columns node. This node can be dragged onto the worksheet,
connected, and executed just like any of the built-in nodes.
Simple C++
Version
In the first implementation, we do the computation completely in
Java. Let’s now construct an implementation in C++. We will need a
Java class extending
EngineNode to calculate the meta data and create
the Java proc object, a Java class extending
CNKProc indicating the
C++ class to use, and a C++ class extending
CNKProc performing the
computation.
Note that
CNKProc is the name of both a Java class and a C++ class.
The C++ classes discussed in the section The TIBCO Spotfire
Pipeline on page 50 all have corresponding Java classes with the same
name and purpose.
Kommentare zu diesen Handbüchern