今天一個客戶問怎樣把表中相同鍵值對應的文本按照一定順序拼接起來。如果使用SQL實現將非常麻煩，並且效率低下。GP4.1以后提供了一個函數array_agg可以方便快捷，高效的實現該功能
比如原始查詢是
test1=# select n_regionkey,n_name from nation;
n_regionkey | n_name
-------------+---------------------------
3 | UNITED KINGDOM
1 | BRAZIL
2 | VIETNAM
4 | IRAN
4 | SAUDI ARABIA
4 | JORDAN
0 | ALGERIA
4 | EGYPT
4 | IRAQ
0 | ETHIOPIA
1 | ARGENTINA
3 | ROMANIA
2 | CHINA
2 | JAPAN
0 | MOROCCO
2 | INDONESIA
1 | UNITED STATES
0 | MOZAMBIQUE
3 | RUSSIA
3 | GERMANY
1 | CANADA
1 | PERU
0 | KENYA
2 | INDIA
3 | FRANCE
我們要把相同地區的代碼的國家，按照字母順序拼接在一起，放在一行當中，並且顯示地區代碼。通過arrary_agg可以很方便的實現該功能
test1=# select n_regionkey,array_agg(trim(n_name) order by n_name) from nation group by n_regionkey;
n_regionkey | array_agg
-------------+--------------------------------------------------
2 | {CHINA,INDIA,INDONESIA,JAPAN,VIETNAM}
4 | {EGYPT,IRAN,IRAQ,JORDAN,"SAUDI ARABIA"}
1 | {ARGENTINA,BRAZIL,CANADA,PERU,"UNITED STATES"}
3 | {FRANCE,GERMANY,ROMANIA,RUSSIA,"UNITED KINGDOM"}
0 | {ALGERIA,ETHIOPIA,KENYA,MOROCCO,MOZAMBIQUE}