A Mitrol database is formed of the three following elements : files, fields and relations,whose definitions are permanently
stored in Mitrol's data dictionary which eliminates potential data inconsistencies and reduces maintenance efforts. Indeed, through the
use of the Data Definition Language, modifications applied to the Data Dictionary affect all occurences of the data.
Besides, within a database file, each object for which you store data is represented by a record. For example, the PART file stores data on
parts ; each individual part you track has a corresponding record in the PART file. Once your database structure is established, you can use
Mitrol requests to create, modify, print and delete records in database files. The type of information you can store in each record depends
on the fields defined in your database files. Each field stores a characteristic, or attribute, of the entity represented in the file. A typical
PART file stores the attributes of individual parts, including their identifying numbers, on-hand quantities, and descriptions. Thus, the
PART file includes such fields as PAR-PART-NUMBER, PAR-ON-HAND, PAR-DESCRIPTION and so on.
Some fields identify individual records in a file. These keyfields correspond to the identifier by which
records are filed in a manual system. For example, part information is usually filed in order by part number. To look up information for
a specific part, you open the file and look for that number. In Mitrol, the field PAR-PART-NUMBER is usually the direct access keyfield for PART records.
Relations specify how files in your database are associated. It is the logical association between database and
records. Relations allow access to records either directly or by reference to an associated file. Much of Mitrol's power comes from this ability to associate records through relations.
Relations are of two types : direct and indirect. A direct relation gives you direct access to a record by
means of the keyfield and index. Mitrol does not have the notion of a primary index - all keys have equal weight , and partial indices are
supported. An indirect relation associates one database file with another.
In a well designed database, the relations between files correspond to the logical associations between the entities
represented by the individual files. For example, in a typical MITROL database, the VENDOR file is linked to the PURCHASE-ORDER file, so that given
a vendor you can retrieve information about each purchase order placed with that vendor. Similarly, given a purchase order you can retrieve information about the vendor with which it is placed.
Use direct relations to start selecting records in a database, and, by means of indirect relations,
describe a path to any record with which it is associated. You can continue the path from that record to other records linked on another
indirect relation gathering related information from many database files.
You must define a field for use by a relation before you define that relation. Using a field as the key for
relation identifies that field as a keyfield. Keyfields may never have their value reset without either having the record deleted and
recreated, or else having the relation link broken and then relinked. Files, fields and relations together form the database in which you
store and retrieve data about your company's operations. In order to define files, fields and relations, you'll use the Data Definition
Language (DDL) which is part of Mitrol's integrated fourth generation language.
Using the Data Manipulation Language (DML), you'll be able to easily access files and fields through a path,
thus filling in and using your database.