Generating Wrappers and Data Sources¶
Wrappers are components responsible for offering the server overall common interface for accessing the data sources. Each search method in a base relation has an associated wrapper, which is in charge of receiving the queries issued to the base relation, transforming them into queries to the data source and obtaining the results, returning them to the logical layer of Virtual DataPort in accordance with the format specified by the base relation. Wrappers make the peculiarities of obtaining data from the sources transparent for the server.
Virtual DataPort includes the following predefined types of wrappers:
WWW: This is used to incorporate wrappers for semi-structured sources created using Denodo ITPilot into the system. These sources can be accessed from the local file system, via Web, or via FTP. HTML websites are the most important type of sources this wrapper is used for, although it can also be used for other semi-structured sources such as PDF or Word files.
JDBC: Extract data from a remote database via JDBC.
ODBC: Extract data from a remote database via ODBC.
Multidimensional Database wrappers: Extract data from multidimensional databases. There is a different wrapper for each type of database:
SAPBWBAPI: use the SAP JCo connector to connect to SAP BW and SAP BI.
ESSBASE: extract data from Oracle Essbase 9 and 11.
OLAP: extract data from Mondrian 3.x, Microsoft SQL Server Analysis Services or any other OLAP server, using their XMLA interface.
WS (Web Services): Extract data invoking operations defined by Web services.
XML: Extract data from XML files, optionally following a specific DTD or schema. These sources can be accessed via web, local file system or FTP.
JSON: Extract data from JSON files. These sources can be accessed via web, local file system or FTP.
DF: Extract data from flat text files that represent tuples using a regular format, such as using specific characters as tuple and field delimiters or disposing the tuple fields according to a regular expression. Amongst the files supported are files in CSV format. These sources can be accessed via web, local file system or FTP.
GS (Google Search): They provide access to indexes on non-structured data created using the search tool Google Search.
LDAP: They extract data from LDAP servers such as Microsoft Windows Server Active Directory.
BAPI: They invoke SAP BAPIs (Business Application Programming Interfaces) to extract data stored in SAP ERP and other SAP applications.
SALESFORCE: They retrieve data from Salesforce.com.
CUSTOM: Extract data from a source through a specific Java implementation. This type of wrapper allows ad hoc construction of a wrapper program for a specific type of source.
There are corresponding data sources elements for all wrappers, except WWW-type ones, to encapsulate certain data on data source access and configuration.
This section describes how to create and modify wrappers (and their data sources) of any type in Virtual DataPort by using VQL.
We strongly recommend you to use the Administration Tool to create data sources and wrappers.
The remainder of this section is structured as follows. The sections Valid Conversions Between Types in Wrappers and VDP Types and Specifying Paths in Virtual DataPort define aspects of general interest for the rest of the section: valid conversions of types between wrappers and base relations in Virtual DataPort and ways of specifying paths to resources. The section Creating Data Sources specifies how to add to the system data sources of the various available types. Finally, section Creating Wrappers shows how to create wrappers for each of these source types.