Back to Contents

Records

Summary:

See also: Variables, Arrays, Data Types, Database Schema File.

Definition

Purpose:

A record defines a structured variable.

Syntax 1:

DEFINE variable RECORD
  member { datatype | LIKE [dbname:]tabname.colname }
  [,...]
 END RECORD

Syntax 2:

DEFINE variable RECORD LIKE [dbname:]tabname.*

Notes:

  1. variable defines the name of the record, it can be a single variable or an array definition.
  2. member is an identifier for a record member variable.
  3. datatype can be any data type, a record definition, or an array definition.
  4. dbname identifies a specific database schema file.
  5. tabname identifies a database table defined in the database schema file.
  6. colname identifies a database column defined in the database schema file.

Usage:

A record is an ordered set of variables (called members), where each member can be of any data type, a record, or an array. Records whose members correspond in number, order, and data type compatibility to a database table can be useful for transferring data from the database to the screen, to reports, or to functions.

In the first form (Syntax 1), record members are defined explicitly. In the second form (Syntax 2), record members are created implicitly from the table definition in the database schema file. The database columns defined as SERIAL will define an INTEGER variable, while SERIAL8/BIGSERIAL columns define BIGINT variables.

Note that when using the LIKE clause, the data types are taken from the database schema file during compilation. Make sure that the database schema file of the development database corresponds to the production database, otherwise the records defined in the compiled version of your programs will not match the table structures of the production database. Statements like SELECT * INTO record.* FROM table would fail.

In the rest of the program, record members are accessed by a dot notation (record.member). The notation record.member refers to an individual member of a record. The notation record.* refers to the entire list of record members. The notation record.first THRU record.last refers to a consecutive set of members. (THROUGH is a synonym for THRU).

Records can be passed as function parameters, and can be returned from functions. However, when passing records to functions, you must keep in mind that the record is expanded as if each individual member would have been passed as parameter. See BDL Stack for more details.

It is possible to compare records having the same structure with the equal operator: record1.* = record2.*

Examples

Example 1:

01 MAIN
02   DEFINE rec RECORD
03               id INTEGER,
04               name VARCHAR(100),
05               birth DATE
06             END RECORD
07   LET rec.id = 50
08   LET rec.name = 'Scott'
09   LET rec.birth = TODAY
10   DISPLAY rec.*
11 END MAIN

Example 2:

01 SCHEMA stores
02 DEFINE cust RECORD LIKE customer.*
03 MAIN
04   SELECT * INTO cust.* FROM customer WHERE customer_num=2
05   DISPLAY cust.*
06 END MAIN