对查询结果进行分组以获取此格式

Question

I have the following schema:

a VARCHAR(255),
b INTEGER,
c INTEGER,
d TEXT

In my database, if d is not null, then c is 0. What I would like to do is group all columns that have equal a s and equal b s. However, I'd like to use the sum of c for the result and the non-null value of d (if there is one). For instance, if my data looked like this:

"XXX" 100 1 NULL
"XXX" 100 1 "my text"
"YYY" 5 1 NULL
"YYY" 7 0 "other text"

I would like the result to be:

"XXX" 100 2 "my text"
"YYY" 5 1 NULL
"YYY" 7 0 "other text"

Is there a query I can use that will do this?

Answer 1

You can use group_concat to concatenate the values of column d into one field

SELECT a, b, SUM(c), GROUP_CONCAT(d)
FROM table
GROUP BY a, b
ORDER BY a asc, b asc

If you have two rows like

"XXX", 100, 0, "Some Text"
"XXX", 100, 0, "Some More Text"

This will return

"XXX", 100, 0, "Some Text,Some More Text"

However if you have three rows:

"XXX", 100, 0, "Some Text"
"XXX", 100, 0, "Some More Text"
"XXX", 100, 1, NULL

You will get

"XXX", 100, 1, "Some Text,Some More Text"

(ie, the null will disappear. Use coalesce if you need to maintain a placeholder)

Answer 2

I would do this just using max() :

select a, b, sum(c) as c, max(d) as d
from table t
group by a, b;