sql join - only select top row from 2nd table

问题

Bit of an sql noob, have a list in table a of customercodes/phone numbers, and table b has all the call records.

I want to select the most recent call from table b for each of the customercodes/phone numbers in table a.

So far I have:

SELECT     A.CustomerCode, A.PhoneNumber, B.StartTime
FROM         tableA A INNER JOIN
             tableB B ON ( A.PhoneNumber = B.PhoneNumber 
                           AND A.CustomerCode = B.CustomerCode )
ORDER BY A.CustomerCode, A.CLI, B.StartTime DESC

But that is bringing up all the results from TableB. I literally just want to get the most recent call if that makes sense?

回答1:

You can join the calls table on itself, and specify that no later row may exist. Like:

SELECT       A.CustomerCode, A.PhoneNumber, B.StartTime
FROM         tableA A 
INNER JOIN   tableB B 
ON           A.PhoneNumber = B.PhoneNumber 
AND          A.CustomerCode = B.CustomerCode
LEFT JOIN    tableB later
ON           B.PhoneNumber = later.PhoneNumber 
AND          B.CustomerCode = later.CustomerCode
AND          later.StartTime > B.StartTime
WHERE        later.PhoneNumber is null

The condition later.PhoneNumber is null says there can't be a later row. If there are multiple rows with the same StartTime, this will select all of them.

This allows you to select all columns from the latest call.

回答2:

If you only want the starttime from table b, you could use a group by

SELECT     A.CustomerCode, A.PhoneNumber, max(B.StartTime)
FROM         tableA A INNER JOIN
             tableB B ON A.PhoneNumber = B.PhoneNumber 
                         AND A.CustomerCode = B.CustomerCode
GROUP BY A.CustomerCode, A.PhoneNumber
ORDER BY A.CustomerCode, A.CLI

回答3:

Something like this?

SELECT A.CustomerCode, A.PhoneNumber, Max(B.StartTime) AS MaxOfStartTime
FROM A INNER JOIN B ON A.PhoneNumber = B.PhoneNumber
GROUP BY A.CustomerCode, A.PhoneNumber;

回答4:

SELECT A.CustomerCode, A.PhoneNumber, B.StartTime
FROM   tableA A INNER JOIN tableB B 
       ON A.PhoneNumber = B.PhoneNumber AND A.CustomerCode = B.CustomerCode
WHERE  B.StartTime = (SELECT MIN(StartTime) FROM B 
                      WHERE PhoneNumber = A.PhoneNumber 
                      AND CustomerCode = A.CustomerCode)
ORDER BY A.CustomerCode, A.CLI, B.StartTime DESC

I believe this will give your desired result, provided B.StartTime is unique in the set related to each row from table A. If you're dealing with a DateTime and events not too close together, this should be reliable.

The only other way I know is using CROSS APPLY in MSSQL 2005+

来源：https://stackoverflow.com/questions/1819003/sql-join-only-select-top-row-from-2nd-table

标签

sql

join

distinct

distinct-values