SQL database: summary of knowledge points, no suspension at the end of the period

Chapter one
1 The characteristics of database ： Database data has permanent storage 、 There are three basic characteristics of organization and sharing .
2 What is? DBMS？（DBMS It's a database management system ： Database management system is a system software specially used for data management and maintenance .）
3 Independence of data （ What is physical independence , What is logical independence ？）
Physical independence ： It refers to when the storage location or storage structure of data changes , Does not affect the characteristics of the application .
Logical independence ： When the content of information expressing the real world changes , Nor does it affect the characteristics of the application .
Chapter two
1 The three elements of the data model ： That is, the basic structure of data should be included when describing data 、 Data constraints （ These two belong to the static characteristics of data ） And define operations on data （ It belongs to the dynamic characteristics of data ） Three aspects .
2 Concept layer data model ： It refers to the elements with application value and their relationships in the abstract reality system （ Commonly used entities - Connection model ）
Entity ： It is a collection of real world objects that have common properties and can be distinguished from each other .
attribute ： A data item that describes the nature or characteristics of an entity or relationship , All instances belonging to an entity have the same properties .
contact ： It refers to the relationship between the internal attributes of an entity , The relation between entities usually refers to the relation between the attributes of different entities .
There are usually three types of connections between two entities ：① One to one contact （1：1)② One to many connection （1：n）③ Many to many contact （m：n）
3 Organization level data model ： hierarchical model 、 Mesh model 、 relational model .
4 What is the external mode ？（ External mode is also called user mode or sub mode , It is a partial description of the overall data of interest to users in the real system , It is used to meet the data needs of different users of the database . A database can have multiple external schemas .）
5 What is a pattern ？（ Also known as logical mode , It is a description of the logical structure and characteristics of all data in the database , Is a common data description for all users . A database has only one schema .）
6 What is the inner mode ？（ Also known as storage mode . Is the underlying representation of the entire database , It describes the storage structure of data , For example, sequential storage ,B Tree storage or hashing , How indexes are organized , Whether it is encrypted, etc . There is only one inner mode .）
The third chapter
1. Candidate key ： An attribute or minimum group of attributes that uniquely identifies a tuple in a relationship . There can be multiple candidate keys on a relationship .
2. Primary key ： Also called primary key or primary key . Is an attribute or attribute group in a table , Used to uniquely identify a tuple . Can consist of one attribute , It can also be composed of multiple attributes .
3. Main attribute ： Attributes contained in any candidate code .
4. Non primary property ： Attributes not included in any candidate code .
5. What are the integrity constraints of the relational schema ？
① Entity integrity ： Entity integrity is to ensure that each tuple in the relationship is identifiable and unique . It means that all tables in a relational database must have primary keys , Moreover, records without primary key value and records with the same primary key value are not allowed in the table .
② Referential integrity ： Restricting the value of a column in one relationship to the value range of a column in another relationship is called referential integrity .
③ User defined integrity ： Also known as domain integrity or semantic integrity . It refers to the database constraints defined for a specific application field , It reflects that the data involved in a specific application must meet the requirements of application semantics .
6. What is foreign key ： set up F It's the relationship R One or a set of attributes of , If F And relationships S The main code of corresponds to , said F It's the relationship R The outer code of （Foreign Key）, And call the relationship R For reference relations （Referencing Relation）, Relationship S For the referenced relationship （Referenced Relation） Or purpose relationship （Target Relation）. Relationship R And relationship S It doesn't have to be a different relationship .
7. What is choice σ： It selects certain tuples from the specified relationships to form a new relationship , The selected tuple satisfies the specified
The logical condition of .
8. What is projection ∏： It is to select the vertical direction of the specified relationship , And form a new relationship . It involves two processes ：1） Select the specified attribute 2） Delete duplicate lines , Form a new relationship
9. What is connection ： Used to connect two related relationships , To create a new relationship .
10. What are the connection requirements ： There are two commonly used connections ： Equivalent connection and natural connection .
Natural connection is a special kind of connection , It requires two relations are compared must be the same component attribute group , And remove duplicate attribute columns from the results .
The difference between natural connection and equivalent connection is ：
Natural join requires equal components to have a common attribute name , Equivalent connection does not require ;
Natural join requires that duplicate attribute names be removed , Equivalent connections don't do that .
11. Basic relational algebra should be able to write .
Eg： Check and repair c02 Student number and grades of the students of course No .
∏sno, grade(σcno=‘c02’(SC))
Inquiry information department c04 Names and grades of students in course No .
∏sname, grade(σcno=‘c04’(SC)σsdept=‘ Information Department ’(Student))
The inquiry has been revised to 2 The name of the student in the semester course 、 Department and selected course number .
∏sname, sdept,cno(σsemester=2(Course) SC Student)
Query the student number and name of students who have taken all courses .
∏sno, sname(Student (SC ÷ ∏cno(Course)))
The query selected “VB” Course and test score is greater than or equal to 80 Student name of 、 Department and grades .
∏Sname, Sdept, Grade(σCname=‘VB’∧ Grade>=80(CourseSCStudent))
or ：
∏Sname, Sdept, Grade(σCname=‘VB’(Course)σGrade>=80(SC)Student)
Among all the students , The query did not select “ Computer culture ” The name of the student and his / her department .
∏Sname, Sdept(Student)- ∏Sname, Sdept（σCname=‘ Computer culture ’ (CourseSC
Student) )
or ：
∏Sname, Sdept(Student)- ∏Sname, Sdept(σCname=‘ Computer culture ’
(Course)SCStudent)
The fifth chapter
1.SQL It can be divided into four parts according to its functions ： Data definition function 、 Data control function 、 Data query function 、 Data manipulation function .
2. Define the base table ： Use SQL In language CREATE TABLE Statements for , Its general format is ：
CREATE TABLE < Table name > (< Name > < data type > [ Column level integrity constraint definition ]{,< Name > < data type > [ Column level integrity constraint definition ］ … }[, Table level integrity constraint definition ] )
Eg：
example ． by Student Table to add “ major ” Column , This column is defined as ：Spec char(10), Allow space .
ALTER TABLE Student
ADD Spec char(8) NULL
example ． Add a new “ major ” The type of the column is changed to char(20).
ALTER TABLE Student
ALTER COLUMN Spec char(20)
example ． Delete the newly added “ major ” Column .
ALTER TABLE Student
DROP COLUMN Spec
3. constraint ：
NOT NULL： The limit column value is not empty .
DEFAULT： The default value for a given column .
UNIQUE： Limit column values are not repeated .
CHECK： Limit the value range of the column .
PRIMARY KEY： Specify this column as the primary code .
FOREIGN KEY： Define this column as a foreign code that references other tables .
4. Delete relation table ：DROP TABLE < Table name >
5. Modify the relation table ：
ALTER TABLE < Table name >
[ ALTER COLUMN < Name > < New data types >] -- Change column definition
[ADD < Name > < data type > < constraint >] -- Add new column
[DROP COLUMN < Name >] -- Delete column
[ADD [constraint < Constraint name >] Constraint definition ]– Adding constraints
[DROP [constraint] < Constraint name >] -- Delete constraints
6. Data integrity ： Data integrity refers to the correctness and consistency of data . Data integrity constraints are designed to prevent the existence of data that does not conform to semantics in the database . These semantic constraints imposed on data are data integrity constraints .
7. Object of integrity constraint ： Column level constraints 、 Tuple constraint 、 Relational constraints
Column level constraints ： For data types 、 data format 、 Value range or value set 、 NULL constraint .
Tuple constraint ： Is a constraint on the relationship between fields in a tuple .
Relational constraints ： Between several tuples 、 Constraints on the connections between relationships .
8.PRIMARY KEY constraint
example ： Add master code constraints to the employee table and the worksheet respectively .
ALTER TABLE Employee list
ADD CONSTRAINT PK_EMP PRIMARY KEY ( Employee number )
ALTER TABLE Worksheet
ADD CONSTRAINT PK_JOB PRIMARY KEY ( Job number )
9.UNIQUE constraint
example ． For employees “ Phone number ” Column add UNIQUE constraint .
ALTER TABLE Employee list
ADD CONSTRAINT UK_SID UNIQUE ( Phone number )
10.FOREIGN constraint
example ． Add a foreign code reference constraint to the job number of the employee table , This column refers to the job number column of the worksheet .
ALTER TABLE Employee
ADD CONSTRAINT FK_job_id FOREIGN KEY ( Job number ) REFERENCES Worksheet ( Job number )
11.DEFAULT constraint
example ． The default value for defining the salary of the employee table is 1000.
ALTER TABLE Employee
ADD CONSTRAINT DF_SALARY DEFAULT 1000 FOR Wages
12.CHECK constraint
example 1． In the employee table , The salary of a restricted employee must be greater than or equal to 1000 Constraints .
ALTER TABLE Employee
ADD CONSTRAINT CHK_Salary CHECK ( Wages >= 1000 )
example 2． Add a constraint that restricts the minimum wage of the payroll to be less than or equal to the maximum wage .
ALTER TABLE Work
ADD CONSTRAINT CHK_Job_Salary CHECK ( minimum wage <= Maximum wage )
Chapter six
1. Must master sub query 、 grouping （ Including aggregate functions ）、 Table connections 、 Additions and deletions 、DELETE、UPDATE、 String matching query
① Subquery

② Inquire about

③ Insert

④ Update data

⑤ Delete

⑥ grouping + Aggregate functions +HAVING Clause

⑦ Table connections

⑧ String matching query

2. The difference between inner connection and outer connection ：（ The internal connection of two tables must meet the conditions , Connect a table externally to meet the requirement of reallocation ）
Internal connection ： Inner connection is one of the most commonly used connection types . When using internal connections , If the related fields of two tables meet the connection conditions , Then extract the data from the two tables and combine them into a new record . Be careful ： Connection fields in connection conditions must be comparable , That is, it must be a column with the same semantics , Otherwise, the comparison will be meaningless .
External connection ： Only the data in one table must meet the connection conditions , The data in another table may not meet the connection conditions .
Chapter vii.
1. Clustered index ： Physical sorting of data by index key .
2. Nonclustered indexes ： Do not physically sort data .
3. Create and delete indexes
① Create index

② Delete index

4. Definition of view ： It is a logical window composed of data selected from the basic table of the database , Is a combination of partial row and column data of a basic table . A view is a virtual table .
Chapter viii.
1. Nontrivial functional dependencies ： If X→Y, but Y Not included in X, said X→Y It's a nontrivial functional dependency .
2. Trivial function depends on ： If X→Y, but Y Included in X, said X→Y It's trivial functional dependencies .
3. Completely function dependent ： If X→Y, And for the X An arbitrary proper subset of X‘ There are X‘—/→Y, said Y The complete function depends on X, Write it down as ：.
4. Part of the function depends on ： If X‘→Y establish , said Y Some functions depend on X, Write it down as ：.
5. The transfer function depends on ： If X→Y（ Nontrivial function dependency , also Y—/→X）、Y→Z, said Z The transfer function depends on X .
6. Problems arising from functional dependencies ： Data redundancy problem 、 Data update problem 、 Data insertion problem 、 Data deletion problem
7. Relationship normalization ： There will be “ Bad ” The theory of transforming functional dependency patterns into good relational patterns .
8. First normal form ： Relationships that do not contain duplicate groups .
9. Second normal form : If R(U,F)∈1NF, also R Each non primary attribute in the is completely functionally dependent on the primary code , be R(U,F)∈2NF.
10. Third normal form ： If R(U,F)∈2NF, And all non - primary attributes do not depend on the primary code , be R(U,F)∈3NF.
11.BC normal form （BCNF)： Relationship model R∈1NF. If a function depends on a set F All functional dependencies in X→Y（Y Not included in X） The left part of the contains R Any candidate key for , be R∈BCNF. In other words ,BCNF The left part of all dependencies in the must contain candidate keys .
Eg： example 1、 Take the warehouse management relation table as an example ： Warehouse ID, Store things ID, Administrators ID, Number . An administrator works in only one warehouse , A warehouse can store many items .
Semantic analysis ：
( Warehouse ID, Store things ID)→( Administrators ID, Number )
( Administrators ID, Store things ID)→( Warehouse ID, Number )
In line with the third paradigm .
( Warehouse ID)→( Administrators ID) ( Administrators ID)→( Warehouse ID), It does not conform to BCNF normal form .
decomposition method ： Decompose the warehouse management relation table into two relation tables ： Warehouse management ：StorehouseManage( Warehouse ID, Administrators ID); Warehouse ：Storehouse( Warehouse ID, Store things ID, Number ).
example 2、 There's a relationship model ：Student（ Student number , full name , Tutor No , Tutor name , Course no. , Course description , achievement ）
semantics ： A student has only one tutor , Students can choose multiple courses .
Normalize it into 3NF Of .
1． This watch is 1NF, Its functional dependency is ：
Student number → full name , Student number → Tutor No , Student number → Tutor name , Course no. → Course description ,
( Student number , Course no. )→ achievement
The main code is （ Student number , Course no. ）
There are some functional dependencies , No 2NF, First, it is decomposed into 2NF.
Student （ Student number , full name , Tutor No , Tutor name ）,
Course （ Course no. , Course description ）,
achievement （ Student number , Course no. , achievement ） Are all 2NF
2． Judge whether it is 3NF
“ Student ” Table is not 3NF, Its functional dependency is ：
Student number → full name , Student number → Tutor No , Tutor No → Tutor name ,
∴ Student ID transfer → Tutor name
Eliminate attributes that depend on the decision maker , Put them in a separate table , obtain ：
Student （ Student number , full name , Tutor No ）, mentor （ Tutor No , Tutor name ）
12. There are two decomposition criteria for relational schema ： Schema decomposition has lossless connectivity ; Schema decomposition preserves functional dependencies .
What is lossless connection means that the decomposed relationship can be restored to the original relationship through natural connection , That is, the relationship obtained through natural connection is compared with the original relationship , No more information 、 Without losing information .
What is preserving functional dependency decomposition means in the process of schema decomposition , Features that cannot be lost in functional dependencies , That is, schema decomposition cannot destroy the original semantics .
additional p154 After class exercises 6、7、8.
Chapter nine
1. Concept of transactions ： A transaction is a user-defined series of data operations , These operations are performed as a complete unit of work . All statements in a transaction as a whole , Or all , Or none at all .
2. The transaction ACID features ： Atomicity （A）： Is the logical working unit of the database , Operations in transactions , Or do it all , Either not .
Uniformity （C）： The result of transaction execution must be to change the database from one consistency state to another .
Isolation, （I）： The execution of one transaction in the index database cannot be disturbed by other transactions .
persistence （D）： Once the matter is submitted , The change to the data in the database is permanent , Future operations or failures will not have any impact on the operation results of the transaction .
3. What is concurrency control ： It is necessary to use the correct method to schedule concurrent operations , Make the execution of one transaction free from the interference of other transactions , Avoid data inconsistencies .
4. Problems caused by concurrent operations ： Lost data modification 、 read “ dirty ” data 、 It can't be read repeatedly 、 produce “ ghost ” data .
5.S lock （ Shared lock ）： You can read but not modify transaction data . You can continue to add S lock , But we can't add X lock , Until released S Lock it .
6.X lock （ Exclusive lock ）： Allows you to read and modify transaction data . There are exclusive locks , No locks , Until released X Lock it .
7. Blocking agreement ：
First level blockade protocol ： On business T Add... To the data to be modified X（ Exclusive lock ） lock , Until the end of the transaction （ Including normal end and abnormal end ） Release only when . The first level blocking protocol can prevent data loss and modification , And make sure that things T It's recoverable . There is no guarantee of readability and non readability “ dirty ” data .
Secondary blockade protocol ： First level blocking protocol plus transactions T Add... To the data to be read S（ Shared lock ） lock , Release after reading S lock . It can prevent data loss and modification , It also prevents reading “ dirty ” data . There is no guarantee that data can be read repeatedly .
Three level blockade agreement ： First level blocking protocol plus transactions T Add... To the data to be read S lock , And do not release until the transaction ends . It can prevent data loss, modification and unread “ dirty ” data , It also prevents non repeatable reading .
8. Deadlock ： Two transactions wait for each other to release resources , It will cause deadlock .
9. The way to solve the deadlock problem ： One is to take certain measures to prevent deadlock ; The other is to allow deadlock , But use certain means to regularly diagnose whether there is deadlock in the system , If there is one, it shall be cancelled .
10. Deadlock is not allowed ： Deadlock prevention ：（1） One time blockade （ Each transaction locks all data to be used at once , Otherwise, we can't carry on ）（2） Sequential blockade （ Set a blocking order for data objects in advance , All transactions are blocked in this order ） It's hard to achieve .
11. Allow deadlock ： Diagnosis and removal of deadlock ：（1） Overtime method （ The waiting time has exceeded the specified time limit , A deadlock occurs . Easy to implement , However, it is prone to misjudgment and unable to handle the deadlock in time ）（2） Waiting graph
Chapter ten
1. The structural design includes ： Conceptual structural design 、 Logical structure design and physical structure design .
2. Design the whole picture E-R The conflicts caused by the model are ： Property conflict 、 name conflict 、 Structural conflict .
3. How to transform conceptual model to relational model ：

additional p186 After-school exercises 10、11
（ The last big question ： Draw by yourself according to the meaning E-R chart , Convert to conform to 3NF Of , Find the primary and foreign keys ）