|Oracle® OLAP Java API Developer's Guide
11g Release 2 (11.2)
|PDF · Mobi · ePub|
This chapter introduces the Oracle OLAP Java application programming interface (API). The chapter includes the following topics:
The Oracle OLAP Java API is an application programming interface that provides access to the online analytic processing (OLAP) technology in Oracle Database with the OLAP option. This topic lists operations that an OLAP Java API client application can perform, describes the classes in the OLAP Java API, describes the objects in a dimensional data model, and discusses organizing data for online analytical processing.
For a description of the advantages of OLAP technology, see Oracle OLAP User's Guide. That document describes the capabilities that Oracle OLAP provides for the analysis of multidimensional data by business intelligence and advanced analytical applications. It describes in depth the dimensional data model, and it discusses the database administration and management tasks related to Oracle OLAP.
Manage OLAP transactions with the database.
Implement a dimensional data model using OLAP metadata objects.
Create and maintain analytic workspaces.
Create logical metadata objects and map them to relational sources.
Deploy the metadata objects as an analytic workspace or as relational tables and views and commit the objects to the database.
Explore the metadata to discover the data that is available for viewing or for analysis.
Construct analytical queries of the multidimensional data. Enable end users to create queries that specify and manipulate the data according to the needs of the user (for example, selecting, aggregating, and calculating data).
Modify queries, rather than totally redefine them, as application users refine their analyses.
Retrieve query results that are structured for display in a multidimensional format.
For more information on some of these operations, see "Tasks That an OLAP Java API Application Performs".
The OLAP Java API has classes that represent the following types of objects.
Build items, processes, specifications, and commands
Cursors that retrieve the data of a query
Expressions that specify data objects, such as a column in a relational table or view, or that specify a function or command that operates on data
Table 1-1 lists packages that contain the majority of the OLAP Java API classes. These packages are under the
oracle.olapi package. The table contains brief descriptions of the package contents.
Contains classes that represent cursor managers and cursors that retrieve the data specified by a
Contains classes that represent data sources and cursor specifications. You use
Contains classes that represent metadata objects, classes that map the metadata objects to relational data sources, and classes that deploy the metadata objects in an analytic workspace or in relational database structures. For a description of these packages, see Chapter 2, "Understanding OLAP Java API Metadata". For information on using the classes in these packages, see Chapter 3, "Discovering Metadata" and Chapter 4, "Creating Metadata and Analytic Workspaces".
Contains classes that represent the items and commands that specify how Oracle OLAP builds analytic workspace objects and classes that implement a syntax for creating SQL-like expressions. You use
Contains classes that represent transactions with Oracle OLAP in an Oracle Database instance. You use
The OLAP Java API also has packages organized under the
oracle.express package. These packages date from the earliest versions of the API. The classes that remain in these packages are mostly
Exception classes for exceptions that occur during interactions between Oracle OLAP and a client application.
For information on obtaining the OLAP Java API software and on the requirements for using it to develop applications, see Appendix A, "Setting Up the Development Environment."
Data warehousing and OLAP applications are based on a multidimensional view of data. This view is implemented in a dimensional data model that includes the following dimensional objects. For more detailed information about all of these concepts, see Oracle OLAP User's Guide and Oracle Warehouse Builder Concepts.
Cubes are containers for measures that have the same set of dimensions. A cube usually corresponds to a single relational fact table or view. The measures of a cube contain facts and the dimensions give shape to the fact data. Typically, the dimensions form the edges of the cube and the measure data is the body of the cube. For example, you could organize data on product units sold into a cube whose edges contain values for members from time, product, customer, and channel dimensions and whose body contains values from a measure of the quantity of units sold and a measure of sales amounts.
The OLAP concept of a cube edge is not represented by a metadata object in the OLAP Java API, but edges are often incorporated into the design of applications that use the OLAP Java API. Each edge contains values of members from one or more dimensions. Although there is no limit to the number of edges on a cube, data is often organized for display purposes along three edges, which are referred to as the row edge, column edge, and page edge.
Measures contain fact data in a cube. The measure values are organized and identified by dimensions. Measures are usually multidimensional. Each measure value is identified by a unique set of dimension members. This set of dimension members is called a tuple.
Dimensions contain lists of unique values that identify and categorize data in a measure. Commonly-used dimensions are customers, products, and times. Typically, a dimension has one or more hierarchies that organize the dimension members into parent-child relationships.
By specifying dimension members, measures, and calculations to perform on the data, end users formulate business questions and get answers to their queries. For example, using a time dimension that categorizes data by month, a product dimension that categorizes data by unit item, and a measure that contains data for the quantities of product units sold by month, you can formulate a query that asks if sales of a product unit were higher in January or in June.
Hierarchies are components of a dimension that organize dimension members into parent-child relationships. Typically, in the user interface of a client application, an end user can expand or collapse a hierarchy by drilling down or up among the parents and children. The measure values for the parent dimension members are aggregations of the values of the children.
A dimension can have more than one hierarchy. For example, a time dimension could have a calendar year hierarchy and a fiscal year hierarchy. A hierarchy can be level-based or value-based.
In a level-based hierarchy, a parent must be in a higher level than the children of that parent. In a cube, the measure values for the parents are typically aggregated from the values of the children. For example, a time dimension might have levels for year, quarter, and month. The month level contains the base data, which is the most detailed data. The measure value for a quarter is an aggregation of the values of the months that are the children of the quarter and the measure value for a year is the aggregation of the quarters that are children of the year. Typically each level is mapped to a different column in the relational dimension table.
In a value-based hierarchy, the parent and the child dimension members typically come from the same column in the relational table. Another column identifies the parent of a member. For example, a value hierarchy could contain all employees of a company and identify the manager for each employee that has one. All employees, including managers, would come from the same column. Another column would contain the managers of the employees.
A level typically corresponds to a column in a dimension table or view. The base level is the primary key.
Attributes contain information related to the members of a dimension. An end user can use an attribute to select data. For example, an end user might select a set of products by using an attribute that has a descriptive name of each product. An attribute is contained by a dimension.
A query is a specification for a particular set of data. The term query in the OLAP Java API refers to a
Source object that specifies a set of data and can include aggregations, calculations, or other operations to perform using the data. The data and the operations on it define the result set of the query. In this documentation, the general term query refers to a
The API has a
Query class in the
oracle.olapi.syntax package. A
Query represents a multirow, multicolumn result set that is similar to a relational table, a SQL
SELECT statement, or an OLAP function. You use a
Query object in mapping a dimension or measure to a relational table or view.
In the OLAP Java API, the dimensional data objects are represented by Multidimensional Model (MDM) classes. These classes are in the
oracle.olapi.metadata.mdm package. Related classes are in the
oracle.olapi.metadata package and the other packages under it. For detailed information about those classes, see Chapter 2, "Understanding OLAP Java API Metadata".
The OLAP Java API makes it possible for Java applications (including applets) to access data that resides in an Oracle data warehouse. A data warehouse is a relational database that is designed for query and analysis, rather than for transaction processing. Warehouse data often conforms to a star schema, which is a dimensional data model for a relational database. A star schema consists of one or more fact tables and one or more dimension tables. The fact tables have columns that contain foreign keys to the dimension tables. Typically, a data warehouse is created from a transaction processing database by an extraction transformation transport (ETT) tool, such as Oracle Warehouse Builder.
For the data in a data warehouse to be accessible to an OLAP Java API application, a database administrator must ensure that the data warehouse is configured according to an organization that is supported by Oracle OLAP. The star schema is one such organization, but not the only one. See Oracle OLAP User's Guide for information about supported data warehouse configurations.
Once the data is organized in the warehouse, you can use an OLAP Java API application to design an OLAP dimensional data model of cubes, measures, dimensions, and so on, and to create the logical OLAP metadata objects that implement the model. You map the metadata objects to data in the warehouse and build an analytic workspace. Building the analytic workspace populates the OLAP views and other storage structures with the data that the OLAP metadata objects represent.
You can also use Analytic Workspace Manager to do the same tasks. See Oracle OLAP User's Guide for information about creating an analytic workspace with Analytic Workspace Manager.
An OLAP Java API application can get the OLAP metadata objects created either by Analytic Workspace Manager or through the OLAP Java API. It can use the metadata objects to create queries that operate on the data in the warehouse.
The collection of warehouse data in an analytic workspace is the data store to which the OLAP Java API gives access. Of course, the scope of the data that a user has access to is limited by the privileges granted to the user by the database administrator.
In addition to ensuring that data and metadata have been prepared appropriately, you must ensure that application users can make a JDBC connection to the data store and that users have database privileges that give them access to the data. For information about specifying privileges, see Oracle OLAP User's Guide. For information about establishing a connection, see Chapter 3, "Discovering Metadata".
Oracle OLAP metadata objects organize and describe the data that is available to a client application. The metadata objects contain other information, as well, such as the data type of the data. However, you cannot retrieve data directly from a metadata object. To specify the data that you want, you must create a query. In specifying the data, you usually must specify one or more dimension member values. To retrieve the specified data, you create a
Cursor. This topic briefly describes those actions.
Another way that you can query the data contained in OLAP metadata objects is through SQL queries of the views that Oracle OLAP creates for the metadata objects. For information about querying these views, see "Using OLAP Views" in Chapter 2, "Understanding OLAP Java API Metadata".
Queries are represented by
oracle.olapi.data.source.Source objects. You get a
Source from a metadata object and use that
Source object in specifying the data that you want to get.
Source classes have methods for selecting and performing operations on the data. You can use the methods to manipulate data in any way that the user requires. For information about
Source objects, see Chapter 5, "Understanding Source Objects" and Chapter 6, "Making Queries Using Source Methods".
The members of an Oracle OLAP dimension are usually organized into one or more hierarchies. Some hierarchies have parent-child relationships based on levels and some have those relationships based on values. The value of each dimension member must be unique.
The OLAP Java API uses a three-part format to uniquely identify a dimension member. The format contains the hierarchy, the level, and the value of the dimension member, and thereby identifies a unique value in the dimension. The first part of a unique value is the name of the hierarchy object, the second part is the name of the level object, and the third part is the value of the member in the level. The parts of the unique value are separated by a value separation string, which by default is double colons (
::). The following is an example of a unique member value of a level named
YEAR in a hierarchy named
CALENDAR_YEAR in a dimension named
The third part of a unique value is the local value. The local value in the preceding example identifies the calendar year 2001.
To retrieve the data specified by a
Source, you create an
oracle.olapi.data.cursor.Cursor for that
Source. You then use this
Cursor to request and retrieve the data from the data store. You can specify the amount of data that the
Cursor retrieves in each fetch operation (for example, enough to fill a 40-cell table in the user interface). Oracle OLAP then efficiently manages the timing, sizing, and caching of the data blocks that it retrieves for your application, so that you do not need to do so. For information about
Cursor objects, see Chapter 8, "Understanding Cursor Classes and Concepts" and Chapter 9, "Retrieving Query Results".
The examples of OLAP Java API code in this documentation are excerpts from a set of example programs that are available on the Oracle Technology Network (OTN) Web site. One example,
CreateAndBuildAW.java, has methods that create and build an analytic workspace. Another example,
SpecifyAWValues, calls the methods of
CreateAndBuildAW.java and specifies values, such as names for the metadata objects and names of columns of relational tables for mapping the metadata objects to data sources. The analytic workspace produced by these examples is named
GLOBAL_AWJ. Other examples query that analytic workspace. The metadata objects in the analytic workspace are mapped to columns in relational tables that are in the Global schema.
From the OTN Web site, you can download a file that contains SQL scripts that create the Global schema and a file that contains the example programs. The OTN Web site is at
To get either file, select Sample Code and Schemas in the Download section of the Web page. To get the sample schema, select Global Schema 11g. To get the example programs, select Example Programs for Documentation and then select Download the Example Programs for 11g Release 2 (11.2) to download the compressed file that contains the examples.
The example programs are in a package structure that you can easily add to your development environment. The classes include a base class that the example program classes extend, and utility classes that they use. The base class is
BaseExample11g.java. The utility classes include
Context11g.java class has methods that create a connection to an Oracle Database instance, that store metadata objects, that return the stored metadata objects, and that create
Cursor objects. The
CursorPrintWriter.java class is a
PrintWriter that has methods that display the contents of
GLOBAL_AWJ, which is the analytic workspace that contains the other objects.
PRODUCT_AWJ, which is a dimension for products. It has one hierarchy named PRODUCT_PRIMARY. The lowest level of the hierarchy has product item identifiers and the higher levels have product family, class, and total products identifiers.
CUSTOMER_AWJ, which is a dimension for customers. It has two hierarchies named SHIPMENTS and MARKETS. The lowest level of each hierarchy has customer identifiers and higher levels have warehouse, region, and total customers, and account, market segment, and total market identifiers, respectively.
TIME_AWJ, which is a dimension for time values. It has a hierarchy named CALENDAR_YEAR. The lowest level has month identifiers, and the other levels have quarter and year identifiers.
CHANNEL_AWJ, which is a dimension for sales channels. It has one hierarchy named CHANNEL_PRIMARY. The lowest level has sales channel identifiers and the higher level has the total channel identifier.
UNITS_CUBE_AWJ, which is a cube that contains the measures COST, SALES, and UNITS. COST has values for the costs of product units. SALES has the dollar amounts for the sales of product units. UNITS has values for the quantities of product units sold. The cube is dimensioned by all four dimensions. The aggregation method for the cube is
SUM, in which each the value for each parent is the sum of the values of the children of the parent.
PRICE_CUBE_AWJ, which is a cube that contains the measures UNIT_COST and UNIT_PRICE. UNIT_COST has the costs of the units. UNIT_PRICE has the prices of the units. The cube is dimensioned by the PRODUCT_AWJ and TIME_AWJ dimensions. The aggregation method for the cube is
AVG, in which the value for each parent is the average of the values of the children of the parent.
For an example of a program that discovers the OLAP metadata for the analytic workspace, see Chapter 3, "Discovering Metadata".
Connects to the data store and creates a
DataProvider and a
Creates or discovers metadata objects.
Deploys, maps, and builds metadata objects, as needed.
Specifies queries that select and manipulate data.
Retrieves query results.
The rest of this topic briefly describes these tasks, and the rest of this guide provides detailed information about how to accomplish them.
You connect to the data store by identifying some information about the target Oracle Database instance and specifying this information in a JDBC connection method. Having established a connection, you create a
DataProvider and use it and the connection to create a
UserSession. For more information about connecting and creating a
UserSession, see "Connecting to Oracle OLAP" in Chapter 3.
You use the
DataProvider to get an
MdmMetadataProvider gives access to all of the metadata objects in the data store. You next obtain the
MdmRootSchema object by calling the
getRootSchema method of the
MdmRootSchema object contains all of the OLAP metadata objects in the database. From the
MdmRootSchema, you get the
MdmDatabaseSchema objects for the schemas that the current user has permission to access. An
MdmDatabaseSchema represents a named Oracle Database user as returned by the SQL statement
SELECT username FROM all_users.
MdmDatabaseSchema, you can discover the existing metadata objects that are owned by the schema or you can create new ones. Methods such as
getDimensions get all of the measures or dimensions owned by the
MdmDatabaseSchema. Methods such as
findOrCreateCube get an analytic workspace or cube, if it exists, or create one if it does not already exist.
From a top-level metadata object contained by the
MdmDatabaseSchema, such as an analytic workspace, cube, or dimension, you can get the objects that it contains. For example, from an
MdmPrimaryDimension, you can get the hierarchies, levels, and attributes that are associated with it. Having determined the metadata objects that are available to the user, you can present relevant lists of objects to the user for data selection and manipulation.
For a description of the metadata objects, see Chapter 2, "Understanding OLAP Java API Metadata". For information about how you can discover the available metadata, see Chapter 3, "Discovering Metadata".
If you create a new
MdmPrimaryDimension, you must deploy it as an analytic workspace object or as a relational OLAP (Rolap) object. To deploy a cube, you call an
MdmCube method such as
findOrCreateAWCubeOrganization. To deploy a
dimension, you call an
MdmPrimaryDimension method such as
If you create a new metadata object that represents data, you must specify an
Expression that maps the metadata object to a relational source table or view, or that Oracle OLAP uses to generate the data. For objects that are contained by an analytic workspace, you can build the metadata objects after mapping them. For information on creating metadata, deploying, mapping, and building metadata objects, see Chapter 4, "Creating Metadata and Analytic Workspaces".
An OLAP Java API application can construct queries against the data store. A typical application user interface provides ways for the user to select data and to specify the operations to perform using the data. Then, the data manipulation code translates these instructions into queries against the data store. The queries can be as simple as a selection of dimension members, or they can be complex, including several aggregations and calculations involving the measure values that are specified by selections of dimension members.
The OLAP Java API object that represents a query is a
Source. Metadata objects that represent data are extensions of the
MdmSource class. From an
MdmSource, such as an
MdmMeasure or an
MdmPrimaryDimension, you can get a
Source object. With the methods of a
Source object, you can produce other
Source objects that specify a selection of the elements of the
Source, or that specify calculations or other operations to perform on the values of a
If you are implementing a simple user interface, then you might use only the methods of a
Source object to select and manipulate the data that users specify in the interface. However, if you want to offer your users multistep selection procedures and the ability to modify queries or undo individual steps in their selections, then you should design and implement
Template classes. Within the code for each
Template, you use the methods of the
Source classes, but the
Template classes themselves allow you to dynamically modify and refine even the most complex query. In addition, you can write general-purpose
Template classes and reuse them in various parts of your application.
When users of an OLAP Java API application are selecting, calculating, combining, and generally manipulating data, they also want to see the results of their work. This means that the application must retrieve the result sets of queries from the data store and display the data in multidimensional form. To retrieve a result set for a query through the OLAP Java API, you create a
Cursor for the
Source that specifies the query.
You can also get the SQL that Oracle OLAP generates for a query. To do so, you create a
SQLCursorManager for the
Source instead of creating a
generateSQL method of the
SQLCursorManager returns the SQL specified by the
Source. You can then retrieve the data by means outside of the OLAP Java API.
Because the OLAP Java API was designed to deal with a multidimensional view of data, a
Source can have a multidimensional result set. For example, a
Source can represent an
MdmMeasure that is dimensioned by four
MdmPrimaryDimension objects. Each
MdmPrimaryDimension has an associated
Source. You can create a query by joining the
Source objects for the dimensions to the
Source for the measure. The resulting query has the
Source for the measure as the base and it has the
Source objects for the dimensions as outputs.
Cursor for a query
Source has the same structure as the
Source. For example, the
Cursor for the
Source just mentioned has base values that are the measure data. The
Cursor also has four outputs. The values of the outputs are those of the
Source objects for the dimensions.
To retrieve all of the items of data through a
Cursor, you can loop through the multidimensional
Cursor structure. This design is well adapted to the requirements of standard user interface objects for painting the computer screen. It is especially well adapted to the display of data in multidimensional format.
For more information about using
Source objects to specify a query, see Chapter 5, "Understanding Source Objects". For more information about using
Cursor objects to retrieve data, see Chapter 8, "Understanding Cursor Classes and Concepts". For more information about the
SQLCursorManager class, see Oracle OLAP Java API Reference.