* All operations on the data tree are performed via one of the transactions: *
* These transactions provide a stable isolated view of data tree, which is * guaranteed to be not affected by other concurrent transactions, until * transaction is committed. * *
* For a detailed explanation of how transaction are isolated and how transaction-local * changes are committed to global data tree, see * {@link AsyncReadTransaction}, {@link AsyncWriteTransaction}, * {@link AsyncReadWriteTransaction} and {@link AsyncWriteTransaction#commit()}. * * *
* It is strongly recommended to use the type of transaction, which * provides only the minimal capabilities you need. This allows for * optimizations at the data broker / data store level. For example, * implementations may optimize the transaction for reading if they know ahead * of time that you only need to read data - such as not keeping additional meta-data, * which may be required for write transactions. * *
* Implementation Note: This interface is not intended to be implemented * by users of MD-SAL, but only to be consumed by them. * * @param
* Type of path (subtree identifier), which represents location in
* tree
* @param , D, L extends AsyncDataChangeListener
* Represents scope of data change (addition, replacement, deletion).
*
* The terminology for scope types is reused from LDAP.
*
*
* Note that this is done in the binding independent data tree and so the behavior
* might be counterintuitive when used with binding aware interfaces particularly
* when it comes to lists. The list itself is a node in the binding independent tree,
* which means that if you want to listen on new elements you must listen on the list itself
* with the scope of {@link #ONE}.
*
* As an example, in the below YANG snippet, listening on node with scope
* {@link #ONE} would tell you if the node-connector list was created or deleted,
* but not when elements were added or removed from the list assuming the list itself
* already existed.
*
*
* You are able to register for notifications for any node or subtree
* which can be reached via the supplied path.
*
* If path type
* You are able to register for data change notifications for a subtree even
* if it does not exist. You will receive notification once that node is
* created.
*
* If there is any preexisting data in data tree on path for which you are
* registering, you will receive initial data change event, which will
* contain all preexisting data, marked as created.
*
*
* You are also able to specify the scope of the changes you want to be
* notified.
*
* Supported scopes are:
*
* This method returns a {@link ListenerRegistration} object. To
* "unregister" your listener for changes call the "close" method on this
* returned object.
*
* You MUST call close when you no longer need to receive notifications
* (such as during shutdown or for example if your bundle is shutting down).
*
* @param store
* Logical Data Store - Logical Datastore you want to listen for
* changes in. For example
* {@link LogicalDatastoreType#OPERATIONAL} or
* {@link LogicalDatastoreType#CONFIGURATION}
* @param path
* Path (subtree identifier) on which client listener will be
* invoked.
* @param listener
* Instance of listener which should be invoked on
* @param triggeringScope
* Scope of change which triggers callback.
* @return Listener registration object, which may be used to unregister
* your listener using {@link ListenerRegistration#close()} to stop
* delivery of change events.
*/
ListenerRegistrationExamples
*
* Following is an example model with comments describing what notifications
* you would receive based on the scope you specify, when you are
* registering for changes on container a.
*
*
* container a // scope BASE, ONE, SUBTREE
* leaf "foo" // scope ONE, SUBTREE
* container // scope ONE, SUBTREE
* leaf "bar" // scope SUBTREE
* list list // scope ONE, SUBTREE
* list [a] // scope SUBTREE
* id "a" // scope SUBTREE
* list [b] // scope SUBTREE
* id "b" // scope SUBTREE
*
*
* Following is an example model with comments describing what notifications
* you would receive based on the scope you specify, when you are
* registering for changes on list list (without specifying concrete item in
* the list).
*
*
* list list // scope BASE, ONE, SUBTREE
* list [a] // scope ONE, SUBTREE
* id "a" // scope SUBTREE
* list [b] // scope ONE, SUBTREE
* id "b" // scope SUBTREE
*
*
*
* @see http://www.idevelopment.info/data/LDAP/LDAP_Resources/
* SEARCH_Setting_the_SCOPE_Parameter.shtml
*/
public enum DataChangeScope {
/**
* Represents only a direct change of the node, such as replacement of a node, addition or
* deletion. Note that, as described in {@link #ONE}, this may have counterintuitive
* interactions when viewed from a binding aware application, in particular when it
* pertains to lists.
*
*/
BASE,
/**
* Represent a change (addition,replacement,deletion) of the node or one of its direct
* children.
*
* container nodes {
* list node {
* key "id";
* leaf id {
* type string;
* }
* list node-connector {
* leaf id {
* type string;
* }
* }
* }
* }
*
*
* This scope is superset of {@link #BASE}.
*
*/
ONE,
/**
* Represents a change of the node or any of or any of its child nodes,
* direct and nested.
*
* This scope is superset of {@link #ONE} and {@link #BASE}.
*
*/
SUBTREE
}
/**
* {@inheritDoc}
*/
@Override
public AsyncReadOnlyTransactionP allows it, you may specify paths up to the leaf nodes
* then it is possible to listen on leaf nodes.
*
*
* See {@link DataChangeScope} for examples.
*