In the previous section, when you dragged the Patient table from the Workspace to the Table Grammar window, an error icon appeared next to the table name. This is because you need to specify a grammar root for your table. The grammar root determines the "region" of the message that is to be equivalent to one row of your table, and from which you will be able to extract data.

The grammar root is normally set to be either:

• A particular segment, or
• The entire message.

If you select a particular segment to be the grammar root, you will only be able to access data in that segment.

To specify the grammar root for your table:

• In the Table Grammar panel, double-click Patient. An Information window appears, reminding you that you must set the grammar root.
• Click OK. The Grammar Root Selection window appears.

Selecting the Grammar Root

• Select the Message folder, and click OK. This enables you to select data items from the entire message.

	You must set the grammar root to be a particular segment of a message if that segment is repeated in your message and you want to extract data from it. For example, suppose that your sample message contains more than one NK1 (Next Of Kin) segment. To extract data from all of these NK1 segments, you will need to specify NK1 as the grammar root for your table. This tells Chameleon to create one table row for each NK1 segment. If you specify the entire message as the grammar root, you will not be able to extract data from repeated segments. Repeating segments are discussed in Step 7 - Repeated Segments.