Add the specified node or nodes to a named removal group. The purpose of this function is to identify a group of nodes. If a group name is not specified the default name
group is used. Each identity is scoped to an individual pattern. Those nodes must be created and managed solely by the pattern making the call. Adding nodes managed by other patterns will have undefined behavior. The function can be called multiple times within the pattern.
On completion of the pattern all nodes created or confirmed as belonging to a group in the previous execution of the pattern are examined. If a node does not appear in the same group in the new run it is deleted.
By default the system associates the group nodes with the trigger node. This means that the trigger node must be an inferred node or it will differ with every invocation of the pattern. This can be overcome by specifying an inferred anchor node for a named group.
You can prevent a removal group being removed using the
For example, the following code creates a group named
dbs. On the first run this contains all the discovered databases. On the second run, if any databases are not seen then they are automatically removed upon the completion of the pattern.
Changing the name of a removal group
In the course of writing a pattern you might choose to change the name of a removal group. When you upload the new pattern with the changed removal group name, and the pattern runs, the original removal group and the nodes it contains are removed. The removal group with the new name is created, but does not have the same nodes added as they have been removed. On subsequent scans the nodes are recreated and are added to the new group when the pattern runs.