The API provides a call-level interface to data sources. Client applications can manipulate persistent data by invoking callable methods designed for this purpose. There are three types of method calls:
Administrative method calls. Initialize the connector and customize certain aspects of its behavior and are required.
Communications method calls. Make it possible for application programs to create and terminate sessions with servers.
Data access method calls. Enable application programs to issue database requests and process the results of those requests.
The typical application follows a predictable pattern. It generally begins by initializing the API and continues by connecting to a server. After connecting (for example, establishing a client/server session), it alternately issues SQL requests and processes results until it completes its work. At that point, the application disconnects from the server and terminates the API. Complex applications, for example, those requiring parallel access to many data sources, deviate from this schema, but not greatly.
When the application issues an SQL request, the API responds by transmitting a character string to the designated server. Once the message is safely on its way, the call-level interface returns control to the caller, freeing the application process and the server process to work in parallel. A client process synchronizes its operation with a server process either explicitly (by invoking a wait method) or implicitly (by calling for data).
The API provides samples for several computer languages including C, C++, Cobol, Visual Basic, and RPG. Any third-generation language that supports call-by-reference can be used to create an API application.
| iWay Software |