Subtype Mapping

There are three ways to implement subtypes in the database. All three will be discussed in this tutorial.

Table of Contents

Objectives

This tutorial covers the following objectives:

State and apply the table, column, identifiers, relationship, and integrity constraint rules for mapping:
- Supertype and subtype implementations
- Supertype and subtype arc implementations

A carpenter who is building your dream house may know that you will use different types of light bulbs all around the house.
However, if you do not provide information on where certain types of light bulbs should be installed, you could end up with an overly bright bedroom and a dimly lit kitchen!
Mapping supertypes and subtypes makes sure that the right information gets stored with each type.

This choice produces a single table for the implementation of the supertype entity and its subtypes.
This is also called “single-table (or one-table) implementation.”
Rules:
- Tables: Only one table is created, regardless of the number of subtypes.
- Columns: The single table gets one column for each attribute of the supertype, along with the original optionality of the attribute.
Rules (cont.):
- The table also gets a column for each attribute belonging to the subtype, but the columns all become optional.
- Additionally, a mandatory column should be created to act as a discriminator column to distinguish between the different subtypes of the entity.
- The value it can take is from the set of all the subtype short names (FTE, PTE, OTR in the example).
- This discriminator column is usually called_type, which would be epe_type in the example.

Rules:
- Identifiers: Unique identifiers transform into primary and unique keys.
- Relationships: Relationships at the supertype level transform as usual. Relationships at the subtype level are implemented as optional foreign-key columns.
- Integrity constraints: A check constraint is needed to ensure that for each particular subtype, all columns that come from mandatory attributes are not null.
In the conceptual model, salary is mandatory for fulltime employees and hourly rate is mandatory for parttime employees.
When the EMPLOYEE supertype is implemented as a single table in the physical model, these attributes become optional.
A check constraint is needed to enforce the business rules modeled in the ERD.
In the example, the code for the check constraint would look like this:
- CHECK (epe_type = ‘FTE’ and salary is not null and hourly_rate is null and agy_id is null)
- OR (epe_type = ‘PTE’ and salary is null and hourly_rate is not null and agy_id is not null)
The code checks that if it is a full-time employee (epe_type = ‘FTE’), then a value must exist in the salary column and the hourly_rate and agy_id columns must be empty.
Conversely, if it is a part-time employee (epe_type = ‘PTE’), then a value must exist in hourly_rate and agy_id, but salary must be left blank.

The single-table implementation is a common and flexible implementation.
It is the one you are likely to consider first and is especially appropriate where:
- Most of the attributes are at the supertype level.
- Most of the relationships are at the supertype level.
- Business rules are globally the same for the subtypes.

This is also called “two-table implementation.”
You create a table for each of the subtypes.
So, in reality, you could have more than two tables, if you had more than two subtypes.
Rules:
- Tables: One table per first-level subtype.
- Columns: Each table gets one column for each attribute of the supertype along with its original optionality.
- Each table also gets one column for each attribute belonging to the subtype along with its original optionality.
Rules (cont.):
- Identifiers: The primary UID at the supertype level creates a primary key for each table. Secondary UIDs of the supertype become unique keys in each table.
- Relationships: All tables get a foreign key for a relationship at the supertype level, with the original optionality.
For relationships at the subtype levels, the foreign key is implemented in the table it is mapped to.
Original optionality is retained.

Subtype implementation may be appropriate when:

Subtypes have very little in common. There are few attributes at the supertype level and several at the subtype level.
Most of the relationships are at the subtype level.
Business rules and functionality are quite different between subtypes.
How tables are used is different — for example, one table is being queried while the other is being updated.

In this ERD, we have redrawn the CLOTHING supertype and its subtypes of SHIRT and SHOE as standalone entities…

…with each one having mandatory 1:1 relationships with the supertype. The relationships are in an arc.

This choice produces one table for every entity.
The supertype table has a foreign key for each subtype table.
These foreign keys represent exclusive relationships.
They are optional because only one of them can have a value for each row in the table.
Rules:
- Tables: As many tables are created as there are subtypes, as well as one for the supertype.
- Columns: Each table gets a column for all attributes of the entity it is based on, with the original optionality.
Identifiers: The primary UID of the supertype level creates a primary key for each of the tables.
- All other unique identifiers become unique keys in their corresponding tables.
Relationships: All tables get a foreign key for a relevant relationship at the entity level, with the original optionality.
Integrity constraints: Two additional columns are created in the table based on the supertype.
They are foreign-key columns referring to the tables that implement the subtypes.
The columns are optional because the foreign keys are in an arc.
An additional check constraint is needed to implement the arc.
The foreign-key columns are also unique keys because they implement a mandatory 1:1 relationship.

This implementation is rarely used, but it could be appropriate when:
- Subtypes have very little in common and each table represents information that can be used independently.
- For example, when the CLOTHING table gives all global information, and both SHOES and SHIRTS give specific information, and the combination of global and specific information is hardly ever needed.
- Business rules and functionality are quite different between all types.
- How tables are used is different.

Key terms used in this tutorial included:

In this tutorial, you should have learned how to:

State and apply the table, column, identifiers, relationship, and integrity constraint rules for mapping:
- supertype and subtype implementations
- supertype and subtype arc implementations

-1