This example uses the sample file "Telephone-Mod.xml". You can either place this file in the repository, or reference it from any location on your disk. When the file contents are fairly static, it is preferable to place the file in the repository so that it can be deployed to the server and maintained along with the data source. If the file contents are dynamic, you might prefer to reference the file in situ, rather than copy it into the repository. Alternatively, you could have a web server provide the XML data on demand.
Create an XML DataSource using the Add DataSource Wizard and give it a unique name. For the URL value, you can use repository:/some-path (reads from the repository), file:/some-path (reads from any location on your machine) or http://host/some-path to get the data from a web server. You can also use the file chooser [...] to choose a file: location.
Choose Next and click the XPath Builder button. This will load the XML document from the URL you've just provided. You will see Figure 7.6, “XPath Builder”. Enter the Root XPath that will identify the record locations. For this example, use
/TelephoneBook/Telephone
which will select each of the Telephone elements that are children of Telephone book. After keying in the root XPath, press Enter.
Note
In this example, there is another inner Telephone element nested within Telephone, so we can't use //Telephone or it would select those too. In any case, // XPath operations will be slower than /, so avoid them unless you need to select from a combination of depths in the XML tree.
When you press Enter, you will see a number of empty (grey) rows appear in the XPath Builder table. We now need to choose what attributes we want to extract. Expand the XML tree shown on the left to see the children of the first Telephone element. Drag the EntryID element over the first column in the table and let go. You should now see the EntryID values have been filled in to that column. Repeat the process, dragging LoginName into the <New> column. This shows the LoginName values and adds another <New> column for your next drop. You can also enter the XPath expressions directly into the table if you need more complex constructs.
The next elements we want are the Street and City from inside Address. Therefore, expand the Address element and drag these in turn over the <New> column. You should now see Figure 7.7, “XPaths Completed”. Note the builder has inserted the correct XPaths: Address/Street and Address/City for you, based on the Root XPath we defined earlier.
You can now choose Ok to close the XPath Builder. Back in the wizard you can see in Figure 7.8, “XML DataSource Schema”, the chosen fields along with the default type String, which is correct in this case. If your values are integers or dates etc. you can adjust the data types (or any of the other information) here.
When the wizard has closed, the new data source will be loaded into the workspace. You can test it by clicking the Load Data button. The output should appear as shown in Figure 7.9, “DataSource Results”.
