Each Oracle WebLogic Server domain generally includes one metadata repository. A metadata repository is divided into a number of partitions, where each partition is independent and isolated from the others in the repository.

Each application can choose which partition to use. Two applications can also choose to share a partition.

Within a partition, you can organize the data in any way. Usually, the data is organized hierarchically like the file system of an operating system. Where a file system uses folders or directories, the Metadata Service uses name spaces or package names which form a unique name used to locate a file.

For all other Oracle Enterprise Scheduler applications, the application name and an optional package name containing the application-level metadata displays under the name space /oracle/apps/ess. For example, the metadata repository for an application named application1 can be divided into packages with the names dev, test, and production.

The metadata repository for this application has the following structure:

/oracle/apps/ess/application1/dev/metadata
/oracle/apps/ess/application1/test/metadata
/oracle/apps/ess/application1/production/metadata

Each Metadata Service method that creates a metadata object takes a required packageName argument that specifies the package part of the directory structure.

After you access an Oracle Enterprise Scheduler metadata repository you can perform different types of Metadata Service operations, including:

• Add, Update, Delete: These operations have transactional characteristics.

• Copy: These operations have transactional characteristics.

• Query: These operations have read-only characteristics and let you list metadata objects in the metadata repository.

• Get: These operations have either read-only or transactional characteristics, depending on the value of the forUpdate flag.

Because clients access the Metadata Service through a Stateless Session EJB, each method uses a reference to a MetadataServiceHandle argument; this argument stores state for Metadata Service operations. The Metadata Service open() method begins each metadata repository user transaction. In an Oracle Enterprise Scheduler application client you obtain a MetadataServiceHandle reference with the open() method and you pass the reference to subsequent Metadata Service methods. The MetadataServiceHandle reference provides a connection to the metadata repository for the calling application.

In a client application that uses the Metadata Service you must explicitly close a Metadata Service transaction by calling close(). This ends the transaction and causes the transaction to be committed or rolled back (undone). The close() not only controls the transactional behavior within the Metadata Service, but it also allows Oracle Enterprise Scheduler to release certain resources. Thus, the close() is also required for Metadata Service read-only query() and get() operations.

User applications use a Stateless Session EJB to access the Metadata Service for application level metadata operations. Using JNDI you can lookup the Metadata Service associated with an Oracle Enterprise Scheduler application.

Example 8-1 shows the JNDI lookup for the Oracle Enterprise Scheduler Metadata Service that allows you to use application level metadata. Note that the getMetadataServiceEJB() method looks up the metadata service using the name "ess/metadata". By convention, Oracle Enterprise Scheduler applications use "ess/metadata" for the EJB reference to the MetadataServiceBean.

// Demonstration on how to lookup metadata service from a Java EE application
// JNDI lookup on the metadata service EJB

import oracle.as.scheduler.core.JndiUtil;

MetadataService ms = JndiUtil.getMetadataServiceEJB();

A filter specifies a comparison or a criteria for a query. You create a filter by creating a comparison that includes a field argument (String), a comparator, and an associated value (Object). In a filter, you can use the filter methods to combine comparisons to form filter expressions.

Table 8-1 lists the comparison operators (comparator argument).

Example 8-2 shows code that creates a new filter.

Filter filter =
  new Filter(MetadataService.QueryField.PACKAGE.fieldName(),
  Filter.Comparator.NOT_EQUALS, null);

A MetadataService query returns an enumeration list of MetadataObjectIDs of the form:

java.util.Enumeration<MetadataObjectId>

Example 8-3 shows a sample routine that queries for a list of job types in the metadata.

Example 8-3, shows the following important steps for using the queryJobTypes() method:

• You need to supply a reference to a metadata repository by obtaining an instance of MetadataServiceHandle.

• You need to create a filter for the query. The filter contains the fields, comparators, and values to search for.

• You determine the field to sort by in the query using the orderBy argument, or you set the orderBy argument to null to indicate that no specific ordering is applied.

• You set the ascending argument for the query. When ordering is applied setting the ascending argument to true indicates ascending order or false indicates descending order for the result list.

Enumeration<MetadataObjectId> qryResults
            = m_service.queryJobTypes(handle, filter, null, false);