Three paradigms of database design

brief introduction

In order to establish less redundancy , A well structured database , Certain rules must be observed when designing the database , In relational databases, this rule is the paradigm . A paradigm is a collection of relational patterns that conform to a certain level . The relationship in relational database must meet certain requirements , That is to meet different paradigms .

At present, there are six paradigms in relational database , Generally speaking, as long as the database meets the third paradigm , And what we often use is what we call the three paradigms , First normal form , Second and third paradigms .

First normal form

Atomicity 、 Indivisibility

What does that mean ？
It requires that every field in your table cannot be further split , Are the smallest data units .

If there is the following student information form

Student information sheet

Student number	full name	Home address	College number	The name of the College	Dormitory number	Dormitory type	The number of beds in the dormitory
001	Zhang San	Xi'an, Shaanxi Province	100001	College of science	222	standard room	4
002	Li Si	Xianyang City, Shaanxi Province	100002	school of computing	333	Non standard room	6
003	Wang Wu	Baoji City, Shaanxi Province	100003	School of Marxism	666	standard room	4

The address column can be further split into provinces 、 City So this table does not conform to the first normal form

Second normal form

Second normal form It is required that on the premise of ensuring the first paradigm , requirement Each column should be related to the primary key , It cannot only be related to a part of the primary key , Require entity uniqueness .

For example, the following table ：

Student number	full name	College number	The name of the College	Dormitory number	Dormitory type	The number of beds in the dormitory
001	Zhang San	100001	College of science	222	standard room	4
002	Li Si	100002	school of computing	333	Non standard room	6
003	Wang Wu	100003	School of Marxism	666	standard room	4

The primary key is student number, college number and dormitory number
full name 、–> Student number
The name of the College —> College number
Dormitory type 、 The number of beds in the dormitory —> Dormitory number

The above is only partially dependent So it violates the second paradigm So you should split the table

Student number	full name	College number	Dormitory number
001	Zhang San	100001	222
002	Li Si	100002	333
003	Wang Wu	100003	666

College number	The name of the College
100001	College of science
100002	school of computing
100003	School of Marxism

Dormitory number	Dormitory type	The number of beds in the dormitory
222	standard room	4
333	Non standard room	6
666	standard room	4

The primary key of the first table is student number , The primary key of the second table is the college number , The primary key of the table in Chapter 3 is dormitory number

This design makes the database clearer , If you need to find the name of the College , Check directly in the college table

Third normal form

The third paradigm requires Each column is directly related to the primary key column Not indirectly , That is, limit the redundancy of columns

.

Dormitory number	Dormitory type	The number of beds in the dormitory
222	standard room	4
333	Non standard room	6
666	standard room	4

Like the above table . Dormitory number —> Dormitory type —> The number of beds in the dormitory
From these dependencies, we can choose the dormitory number —> The number of beds in the dormitory
Dormitory No —> The number of beds in the dormitory is called transmission dependence

So the table should be subdivided again

Dormitory number	Dormitory type
222	standard room
333	Non standard room
666	standard room

Dormitory type	The number of beds in the dormitory
standard room	4
Non standard room	6
standard room	4

In this way, there is no indirect dependency .

Be careful ： The repeated occurrence of primary keys and foreign keys in multiple tables does not belong to data redundancy , The repetition of non key fields is data redundancy .