[MySQL] Statistics by date (day before, week, someday)

In MySQL database, you often encounter the content of the day of statistics.
For example, in the user table, the Date field is: Log_time
Statistics Day

The SQL statement is:

SELECT * from user where date (log_time) = Curdate ();

Curdate () indicates the date of the day
Statistics day before

If the data for the previous day is indicated, you cannot use Curdate ()-1, because the Curdate ()-1st period is not the month-end date for the last one months when the date is the beginning of the month.

For example: Today is June 1, theoretically curdate ()-1 is May 31, but Curdate ()-1 gets not May 31, but 0 days. Then the date of the day before the statistics can not use Curdate ()-1, MySQL database has a new method to count the data the previous day.

Statistics log SQL statements for the day before:

SELECT * from beans where date (log_time) = Date_sub (Curdate (), Interval 1 day);

In parentheses, the day before today, if the number of days before the statistics, the parentheses in the ' 1 ' is changed to the corresponding days. If you want to count months or years, change the day directly to month or year
Statistics this week

Requirements: Statistics starting from yesterday 7 days before the diary including yesterday

SELECT * from user-where date (log_time) >= date_sub (Curdate (), Interval 7 day)
and date (log_time) <= date_sub (Curdate (), Interval 1 day)

On the Internet to find the use of week statistics for a week, can only be counted to 5 days of information, not meet the requirements, so use this method.
Count One day

Statistics history a day's log, replacing Curdate () in Date_sub (Curdate (), Interval 0 day) function with a date
For example: to count the 2012-05-25 date information
Date_sub (' 2012-05-25 ', Interval 0 day)
An example of a date_sub () function:

Today is May 20, 2013.
Date_sub (' 2012-05-25 ', Interval 1 day) says 2012-05-24
Date_sub (' 2012-05-25 ', Interval 0 day) says 2012-05-25
Date_sub (' 2012-05-25 ', interval-1 day) says 2012-05-26
Date_sub (' 2012-05-31 ', interval-1 day) says 2012-06-01
Date_sub (Curdate (), Interval 1 day) represents 2013-05-19
Date_sub (Curdate (), Interval-1 day) represents 2013-05-21
Date_sub (Curdate (), Interval 1 month) represents 2013-04-20
Date_sub (Curdate (), Interval-1 month) represents 2013-06-20
Date_sub (Curdate (), Interval 1 year) represents 2012-05-20
Date_sub (Curdate (), Interval-1 year) represents 2014-05-20

[MySQL] Statistics by date (day before, week, someday)