The cursor concept is an alternative to the named and unnamed result tables which can be generated with the select statement. It is supported for compatibility reasons; but its use is more complicated than that of the result tables.
A result table is specified by the statement declare cursor, and created and opened for processing by the open statement. A cursor determines a position within the result table from which a table row can be accessed by means of a fetch statement.
The result table exists until a corresponding close statement is executed or until the program is terminated.
The statement declare cursor is not written into the trace file, because it is not sent to the ADABAS kernel. This is only done for the open statement.
See also: