MySQL development considerations (Alibaba development manual)

1. Make a schedule

1.1 Field constraints that indicate whether the concept is

You have to use is_xxx How to name , The data type is unsigned tinyint( 1 Said is ,0 Indicate no )

	 Be careful ： If any field is non negative , Must be unsigned

1.2 Table name 、 Field name constraints

For table names 、 Field names must be lowercase letters or numbers , Do not start with numbers , Prohibit only numbers between two underscores , And the table name cannot use plural names

1.3 Disable reserved words , such as ：desc、range、match wait

1.4 Index name constraint

Index name	Name constraint
primary key	pk_ Field name
unique index	uk_ Field name
General index	idx_ Field name

 Full-text index ：InnoDB I won't support it ,Myisam Good support performance , Generally in char、varchar、text Column to create 

1.5 Decimal type constraint

Decimal type is decimal, No use float and double

	 reason ： At the time of storage ,float and double There is a problem of precision loss 
	 Suggest ： The range of stored data exceeds decimal The scope of the , It is recommended to split the data into integers and decimals for storage  

1.6 If the stored string length is almost equal , Use char Fixed length string type

2. Index specifications

2.1 Fields with unique characteristics on the business , Even a combination of multiple fields , You have to build a unique index

	 explain ： Building a unique index affects insert The number of degrees can be ignored , But building a unique index will increase the search speed 

2.2 More than three tables prohibited join, need join Field of , Data types must be consistent ; When multi table associated query , Ensure that the associated field needs to have an index

2.3 stay varchar Index on field , Index length must be specified , There's no need to index all fields

2.4 Page search must not be left blurred or full blurred , If you need to go to search engine to solve .

3. SQL sentence

3.1 Do not use count（ Name ） perhaps count（ Constant ） To replace count（*）

count（*） The statistical value is NULL The line of , and count（ Column ） Will not count as null Row of values 

3.2 count（distinct col） Calculate the column Division NULL Duplicate lines outside

	 Be careful ：count（distinct col1,col2） If one of the columns is all null, So even if the other column has different values , And will return to 0

3.3 When the values of a column are all NULL,count（col） The return result is 0 however sum（col） The return result of is NULL

3.4 Use isnull（） To determine whether it is null value

 explain ：null The comparison with any value is null
	1.null<>null The return value is null, instead of false
	2.null=null The return value is null, instead of true
	3.null<>1 The return result is null, instead of true

3.5 When the code writes the paging query logic , if count by 0 Should return directly , Avoid executing subsequent paging statements

3.6 Do not use foreign keys and cascades , All foreign keys must be resolved at the application layer

3.7 Stored procedures are not allowed , Stored procedures are difficult to debug and extend , There is no portability

3.8 Data revision （ When deleting or modifying ） First select, Avoid deleting by mistake , It is confirmed that the correction is correct

4. ORM mapping

4.1 When querying tables , Never use * As a list of query fields , What fields are required must be clearly stated

4.2 POJO The Boolean attribute of a class cannot be added is, Database fields must be added is-, Ask for in resultMap Mapping between fields and attributes in

4.3 Do not use resultClass When a parameter is returned , Even if all class attribute names and database fields correspond to each other , You also need to define

4.4 sql.xml Use of configuration parameters #{}

4.5 It is not allowed to take it directly HashMap And HashTable As the output of the query result set