A wrapper for a web-based information source is a special kind of wrapper in that the underlying source involves websites and applications, while the native results are in the form of an XML document. The basic functionality of such wrappers include the following:
When the output XML template is defined, the wrappers for each information source will be generated dynamically. Wrappers are not developed prior to this because each information source has a vast array of services, while different broker or mediator will only use a specific part of the service. Foe example, AmazonSearch also provides information on publishers and publishing years for books. However, since the broker will not need those information, there is no need for the wrapper to produce these information.
The XML parser parses the XIB Description. Given the output XML template, the output bindings, input/output descriptions and a query from the user, a WebL script is generated to extract information from the source. The WebL interpreter interprets the WebL script and extracts the pertinent data. Finally, the XML generator transforms the output of the WebL interpreter to the XML format according to the output XML schema.