SQL and Relational Theory, 2nd Edition

CHAPTER 2

2.1 A type is a named, finite set of values—all possible values of some specific kind: for example, all possible integers, or all possible character strings, or all possible supplier numbers, or all possible XML documents, or all possible relations with a certain heading (etc., etc.). There’s no difference between a domain and a type. Note: SQL does draw a distinction between domains and types, however. In particular, it supports both a CREATE TYPE statement and a CREATE DOMAIN statement. To a first approximation, CREATE TYPE is SQL’s counterpart to the TYPE statement of Tutorial D, which I’ll be discussing in Chapter 8 (though there are many, many differences, not all of them trivial in nature, between the two). CREATE DOMAIN might be regarded, very charitably, as SQL’s attempt to provide a tiny part of the total functionality of CREATE TYPE (it was introduced in SQL:1992, while CREATE TYPE wasn’t introduced until SQL:1999); now that CREATE TYPE exists, there seems little reason to use, or even support, CREATE DOMAIN at all.

2.2 Every type has at least one associated selector; a selector is an operator that allows us to select, or specify, an arbitrary value of the type in question. Let T be a type and let S be a selector for T; then every value of type T must be returned by some successful invocation of S, and every successful invocation of S must return some value of type T. See Chapter 8 for further discussion. Note: Selectors are provided “automatically” in Tutorial ...