Do like and in indexes in MySQL go

Today, let's practice
First, we create a user table to test

Like

stay email Add an index to the field to test Like keyword

Let's review Like Several ways of writing sentences
Say... In the big direction Like Statements are written in two ways , When we parted % and _.
%: Used to match the appearance of several characters （ It can also be 0 individual ）
_: Used to match the appearance of a single character
Our main fuzzy query in work is also %, Let's focus on one Like Several ways of writing

1. select * from t_user where email like ‘ Characters you know %’

2. select * from t_user where email like ‘% Characters you know ’

3. select * from t_user where email like ‘% Characters you know %’
   Let's take a look at these three cases of indexing
   Let's look at the first way of writing

   select * from t_user where email like ‘ Characters you know %’

Use explain Look at the

Obviously, this statement is indexed .
Is there a problem here ？ Let's modify the data in the database. Will this line of statement still go through the index ？
We modify the data to duplicate , Reduce the index discrimination and try again .

Re pass explain Analyze the results

adopt type Column ALL It is obvious that this statement has not gone through the index , Because at this time, in the database email Field discrimination is no longer supported 123% This way of writing is indexed , So use the full table scan .

Next, let's look at the second way

select * from t_user where email like '% Characters you know '

First look at the test data

Let's carry out this line sql

SELECT * from t_user where email LIKE '%[email protected]'

You can stop and think about whether you will go to the index , Pay attention to what I write sql And database data . This article sql The discrimination of the target data queried in the database is no problem , and email Fields are also indexed , Logically speaking, this statement will definitely follow the index .
Let's see explain Analysis results of

Judging from the results, there is no index , In fact this with mysql The index organization structure of .
% Add to the front mysql Cannot use index query , because mysql I do not know! % How many characters does it represent , So we can only scan the whole table .

The third way of writing is actually the same as the second way , Also belong to % The one in front , Also won't go index .

In
in Whether to go or not depends on the amount of data behind the condition , When the amount of data is small, the index will be used , And when in When the amount of data followed is large, the full table scan will be performed .mysql The bottom layer may have repeated data processing for the following conditions ; Refer to others' blogs to explain that the range of conditions accounts for% of the total data 30% The whole table will be scanned around , Abandon the way index is executed ; It may also involve the determination of the proportion of the results after de duplication of the following data ！

You can see this blog in detail , It's still very detailed .

MySQL Of in Whether conditions go or not index