Overview:
Entities are anything that have two essential characteristics. First,
entities must be uniquely identifiable, and second, entities must be
describable. A database is simply a collection of related entities
consisting of lists of unique instances. An individual entity within a database
is referred to as an instance. Entities consist of a list of two or more attributes.
If an attribute serves to identify an instance, it is a key attribute.
Attributes that describe an instance are non-key attributes. Each
attribute
contains a datum or fact about the instance. Key attributes serve to create
relationships between database tables. Single table databases are flat
databases, however, those with two or more tables having defined relationships
are known as relational databases. The names of databases, entities, and
attributes are their identifiers. A database table, each of its rows
(instances), and each of its columns (attributes) are considered individual database
objects (dbo) and can be manipulated and interrogated in a variety of ways.
Domain Integrity:
Domain integrity, the method of ensuring the accuracy of the data contained in
the database, is accomplished by the configuration of each attribute. Factors
include datatype, length, range of values, nullability, and default value. There
are six categories of datatypes: exact numeric, approximate numeric, date and
time, character, binary, and special. Length is concerned with the number of
bytes allowed for the attribute. Nullability concerns whether or not an
attribute can be set to null or unknown values. Default values are preset values
that are inserted into the attribute when nulls are not allowable.
Identifier Naming Conventions:
By convention, entities are expressed as plural nouns (e.g. CUSTOMERS or
VENDORS), and attributes are expressed as singular nouns (e.g. NAME or PHONE).
A complete specification consists of the entity identifier combined with the
attribute identifier, separated by a period (e.g. CUSTOMERS.NAME).
Identifiers that require multiple words are generally separated by underscores
(e.g. HOME_ADDRESS).
Identifier Naming Requirements:
Identifiers for database objects (dbo) must conform to the following
requirements.
- An identifier must not exceed 128 characters (bytes) in length.
- The first character must be either a letter [A-Z], underscore [_],
asperand [@], or octathorpe [#].
- Remaining characters must be either letters [A-Z], numbers [0-9],
underscores [_], asperands [@], or octathorpes [#].
- An identifier cannot contain spaces [ ] or other special characters not
listed above.
- Any identifier that does not conform to these rules must be
delimited with either
single quotes [' '] or square brackets [ ].
- Any identifier that does conform to these rules may be delimited
with either single quotes [' '] or square brackets [ ].
- No identifier can be a reserved SQL keyword.
Notation Methods:
Two common methods of diagramming entities and attributes are shown below:
Manual notation method:
ENTITY (attribute1* attribute2 attribute3 attribute4)
Note that the primary key is underlined and asterisked and all foreign keys
are underlined.
Entity/Relationship method:
ENTITY
attribute1*
attribute2
attribute3
attribute4
Note: an E/R Diagram consists of three boxes. The entity name goes in the top box, key attributes
(underlined) in the middle box, and non-key attributes in the lower box. The primary key is also asterisked.
|