In row-based mode, the IP is responsible for accessing rows from the data source and then using the OpenAccess SDK SQL engine to process these rows. The IP only has to deal with row-level access. The OpenAccess SDK SQL engine handles all SQL parsing and execution.
The OpenAccess SDK SQL engine interfaces with an IP through a set of functions that are defined by the IP API. Each IP registers with the OpenAccess SDK SQL engine, its routines for performing the required functions, and its capabilities in supporting operations like SELECT, INSERT, UPDATE, DELETE.
The OpenAccess SDK SQL engine passes the IP information such as the columns referenced in the WHERE clause evaluation and the columns referenced in the SELECT part (the columns to return as part of the result). The IP has the option of passing only the required columns or all columns to the OpenAccess SDK SQL engine. The OpenAccess SDK SQL engine reduces, if necessary, the column set passed by the IP to the column set requested by the client.
The IP calls functions in the OpenAccess SDK SQL engine to evaluate rows against the SQL expression and to optimize the execution of the query. An IP can register to handle LIKE and other expressions, in which case the OpenAccess SDK SQL engine allows the IP to handle it. Otherwise, the OpenAccess SDK SQL engine will handle all expression processing.
