Experience of Niuke SQL

Cattle from SQL Experience of topic brushing

Topic link ： Cattle from SQL piece

1. Ranking ： Can pass select To implement , That is to define a variable , So as to realize each accumulation ：select @t_rank:=0, Through this, we can realize that the ranking will not repeat because of the same score ; Corresponding ones with the same ranking due to the same score can also be achieved through this , That is, through Two variables , One to store the ranking , One to store scores （ To achieve the same score , For the same place ）, The specific way of writing ：

   select grade,
   case 
   when @mid=grade then @cnt
   when @mid:=grade then @cnt:[email protected]+1
   end t_rank
   from grades,(select @cnt:=0,@mid:=null) r
   order by grade desc;
   

   In this way , Be careful := and = The difference between

   But in MySQL8.0 Added window function , It can easily solve the above problems ：row_number()dense_rank()rank()

   The first is general sorting ; The second is the sort with the same score , There is no separation ; The third is the sort with the same score , But there are gaps

   usage ：dense_rank() over(order by xxxxx) Other uses are similar

2. Related cumulative problems ： That is, the sum of the current field plus the data of the previous field

   • You can also take define Variable to solve , But this will always make some inexplicable mistakes （ It's not that the answer is incorrect ）
   • You can use the method of table connection , Use a table as the main table , A supplementary table , Some non equivalent relations can be used to realize
   • You can use the open window function ：sum( Accumulated fields ) over(order by xxxxxx) This method is very simple

3. alter Create index ：

   • Primary key ：alter table table_name add primary key (column_name);
   • only ：alter table table_name add unique index_name (column_name);
   • alter Create multiple indexes ：alter table table_name add .......,add.......;

4. Force index ：force index(index_name), Usually occurs in from After table name

5. Add fields to the created table ：alter table table_name add column Back and create Create a table ;

6. replace(A,B,C)： It means will a Used in b String use c Instead of

7. Add foreign key constraints to the created table ：alter table table_name add constraint foreign key ( Field ) references Table name ( Field name );, Notice the parentheses here

8. In the use of update While updating the data , If the updated data is limited to a field of another table , Then you can use in Keyword to implement ,in You can add data of a field later

9. Modify the name of the table ：alter table table_name rename to/as new_table_name;

10. in and exists The difference between ：

    • in：in Whether the query matches the subquery or the existing record ; Execute subquery first , Then form a Cartesian product with the appearance , To match each other .in It will only be queried once
    • exists：exists Is to query whether this record exists , It doesn't pay attention to the fields of the query , Only care about whether this field exists . First, traverse the surface , Then match with the inner table , If it exists , Add the record to the result set ; Usually, there are two tables in the sub query , There is only one table that will make mistakes

11. Query when a string appears in a field ：(length(column_name)-length(replace(column_name,'xxx','')))/ The length of a single string ;

12. see MySQL Ranking of records in each group in , You can use the open window function , Just sort those , Only need over Add a partition by Grouping field that will do ,partition by It is usually used in conjunction with order by

13. MySQL Returns the month ：month()

14. If you want to connect all the data in the Group , Use group_concat( Fields to be connected ,'x'),x Usually connected punctuation , Usually ,; This function also needs cooperation group by To use

15. MySQL in 3 Insert in ：

    • insert into： Means insert data ,database Will check the primary key, When the inserted data is repeated , You're going to report a mistake
    • replace into： Represents inserting and replacing data , In the demand table There is primary key or unique when , If database The record already exists , The data will be replaced ; If there is no , Simply insert
    • insert ignore into： Indicates that if the inserted data is repeated , Will ignore the current new data ,（ Presence in table primary key or unique）

16. While deleting data , Cannot find at the same time , You need to use the middle table

17. trigger ： Triggered by events , Can only be created in a permanent table , Cannot create in view or In the temporary table ,

    Format ：

    create trigger trigger_name trigger_time trigger_event on table_name for each row
    trigger_body;
    

    trigger_time： Trigger time ,after or before

    trigger_event： Triggering event ,insert update delete

    trigger_body： Trigger subject , Can combine begin…end Compound statements to use , Because the trigger subject is also SQL sentence , Also need to ;, have access to old and new To reference the record content that occurs in the trigger

    • old： Express delete and update, Delete or update previous data
    • new： Express update and insert, New data updated or inserted

    The specific use old.column_name