The select tc ard_id, t.u pload_time from CHRDATAUPLOAD t where
T.u pload_time=(select Max (t1. Upload_time) from CHRDATAUPLOAD t1 where t1. Card_id=tc ard_id)


Said card_id note: here is the title of id, in addition to the Max (), what is more convenient, more faster way?

CodePudding user response:

SELECT * FROM

(SELECT * FROM CHRDATAUPLOAD WHERE card_id=? The ORDER BY upload_time desc)

WHERE ROWNUM=1 sorted before the first

CodePudding user response:

- habit to write so, don't know whether I better, scan only one table,

 select Arthur c. ard_id, c.u pload_time 
The from (select t. *, 
- the descending Max first 
Row_number () over (partition by tc ard_id order by t.u pload_time desc) rn 
The from CHRDATAUPLOAD t) c 
Where an rn=1; 

CodePudding user response:

The select tc ard_id, t.u pload_time from CHRDATAUPLOAD t where
(tc ard_id, t.u pload_time) in (select t1 card_id, Max (t1) upload_time) from CHRDATAUPLOAD t1 group by t1. The card

CodePudding user response:

reference 1st floor xiaovhao response:

SELECT * FROM

(SELECT * FROM CHRDATAUPLOAD WHERE card_id=? The ORDER BY upload_time desc)

WHERE ROWNUM=1 sorted before the disaster area the first

this should be a faster way

CodePudding user response:

At present is not particularly large amount of data, try to use the method of the second floor, thank you

CodePudding user response:

The SELECT CARD_ID, UPLOAD_TIME
The FROM (SELECT tc ARD_ID,
T.U PLOAD_TIME,
MAX (UPLOAD_TIME) OVER (PARTITION BY CARD_ID) AS UPLOAD_TIME_
The FROM CHRDATAUPLOAD T) T
WHERE UPLOAD_TIME_=UPLOAD_TIME;

CodePudding user response:

reference

CodePudding user response:

Reference

CodePudding user response:

refer to the original poster had been based around the Java response:

select tc ard_id, t.u pload_time from CHRDATAUPLOAD t where
T.u pload_time=(select Max (t1. Upload_time) from CHRDATAUPLOAD t1 where t1. Card_id=tc ard_id)


Said card_id note: here is the title of id, in addition to the Max (), what is more convenient, more faster way?

This writing has a problem, if all upload_time, estimation and the result is different as you might expect,
The most simple and crude way:
The select tc ard_id, MAX (t.u pload_time) upload_time from CHRDATAUPLOAD t GROUP BY tc ard_id;