SQL UNION 和UNION ALL 操作符
Reference: http://www.w3school.com.cn/sql/sql_union.asp
SQL UNION 操作符
UNION 操作符用於合併兩個或多個 SELECT語句的結果集。
請注意,UNION 內部的 SELECT 語句必須擁有相同數量的列。列也必須擁有相似的數據類型,但列名可以不同。同時,每條 SELECT 語句中的列的順序必須相同。
SQL UNION 語法
SELECT column_name(s) FROM table_name1
UNION
SELECT column_name(s) FROM table_name2
註釋:默認地,UNION操作符選取不同的值。如果允許重複的值,請使用UNION ALL。
SQL UNION ALL 語法
SELECT column_name(s) FROM table_name1
UNION ALL
SELECT column_name(s) FROM table_name2
另外,UNION 結果集中的列名總是等於 UNION 中第一個SELECT 語句中的列名。
下面的例子中使用的原始表:
Employees_China:
E_ID |
E_Name |
01 |
Zhang, Hua |
02 |
Wang, Wei |
03 |
Carter, Thomas |
04 |
Yang, Ming |
Employees_USA:
E_ID |
E_Name |
01 |
Adams, John |
02 |
Bush, George |
03 |
Carter, Thomas |
04 |
Gates, Bill |
使用 UNION 命令
實例
列出所有在中國和美國的不同的僱員名:SELECT E_Name FROM Employees_China
UNION
SELECT E_Name FROM Employees_USA
結果
E_Name |
Zhang, Hua |
Wang, Wei |
Carter, Thomas |
Yang, Ming |
Adams, John |
Bush, George |
Gates, Bill |
註釋:這個命令無法列出在中國和美國的所有僱員。在上面的例子中,我們有兩個名字相同的僱員,他們當中只有一個人被列出來了。UNION 命令只會選取不同的值。
UNION ALL
UNION ALL 命令和 UNION 命令幾乎是等效的,不過 UNION ALL 命令會列出所有的值。
SQL Statement 1
UNION ALL
SQL Statement 2
使用 UNION ALL 命令
實例:
列出在中國和美國的所有的僱員:
SELECT E_Name FROM Employees_China
UNION ALL
SELECT E_Name FROM Employees_USA
結果
E_Name |
Zhang, Hua |
Wang, Wei |
Carter, Thomas |
Yang, Ming |
Adams, John |
Bush, George |
Carter, Thomas |
Gates, Bill |
union all 和order by
Reference: http://www.linuxidc.com/Linux/2011-05/35796.htm union all 和 order by
有時候,我們會將經過排序(order by)後的結果集與其他經過排序的結果集進行合併(union or union all),比如:select * from tb where length(id)=5 order by id desc
union all
select * from tb where length(id)=10 order by id asc
通常情況下,上面的查詢將會得到下面的錯誤提示:
ORA-00933: SQL command not properly ended
錯誤指向union關鍵字這裏。
create table t as
select 'china' col_1,'america' col_2,'canada' col_3,-1 status from dual union all
select '花生','瓜子','綠豆',0 from dual union all
select '牙膏','牙刷','杯子',3 from dual union all
select '芍藥','牡丹','月季',1 from dual union all
select '優樂美','香飄飄','炸雞',2 from dual
需求:有如上表t,status字段的取值範圍:[-1,3],我們想要做的是,按照這樣的方式排序0,1,2,3,-1 解法:更具題義,我們需要將status分爲兩個區域(>0 和<0) ,然後分別對每一個區域內的數據進行order by排序 ,於是有下面的查詢
select col_1,col_2,col_3,status
from t
where status >= 0
order by status --1
union
select col_1,col_2,col_3,status
from t
where status < 0
order by status --2
不幸的是,正如剛剛開始時我提示的一樣,我們得到了下面的錯誤提示: ORA-00933: SQL command not properly ended
如果將第一個select語句的order by子句去掉,得到的又不是我們想要的結果,此時得到的結果是將整個結果集進行排序。如果將兩個排序子句都去掉的話,雖然按照status爲正負數分開了,但是沒有排序
下面我們來看看正確的答案吧!
解法一: 如果是union all前後分別排序的話,在外面包一層select就起作用了。
<span style="background-color: rgb(255, 255, 0);">select * from ( </span>
select col_1,col_2,col_3,status
from t
where status >= 0
order by status)
union all
<span style="background-color: rgb(255, 255, 0);">select * from (</span>
select col_1,col_2,col_3,status
from t
where status < 0
order by status)
結果: COL_1 COL_2 COL_3 STATUS
------ ------- ------ ----------
花生 瓜子 綠豆 0
芍藥 牡丹 月季 1
優樂美 香飄飄 炸雞 2
牙膏 牙刷 杯子 3
china americacanada -1
解法二:
select * from t
order by
decode(status,
-1,1,
3,2,
2,3,
1,4,
0,5) desc
這可是一個很妙的排序,本人首次看到在order by語句中可以使用decode()函數來排序。
同理,我們也可以使用case語句來排序:解法三:
select * from t
order by
case status
when -1 then 5
when 3 then 4
when 2 then 3
when 1 then 2
else 1
end
最後,union和unionall中都支持orderby和groupby排序和分組子句
union all 和order by, 兩個連用的話會是,查看下一段資料。
另外一段資料:union all and order by
這段資料裏面涉及到的英文資料的鏈接很好。
Reference: http://blog.sina.com.cn/s/blog_6cb0deff0100t4l8.html
遇到的問題:一個sql中,union了幾個子查詢。單獨執行每個子查詢都沒問題,但union後執行,報
ORA-00904: "xxx": invalid identifier
關於union的使用:
SQL: UNION Query:
http://www.techonthenet.com/sql/union.php
SQL: UNION ALL Query:
http://www.techonthenet.com/sql/union_all.php
前面已經提到:union的各個子查詢要有相同數量的列,且對應位置的列必須具有相同的數據類型;但列的名字可以不同。
thediffrence between UNION ALL and UNION is that UNION will attempt to eliminateduplicates.
關於order by的使用:
SQL: ORDER BY Clause
http://www.techonthenet.com/sql/order_by.php
Example #3
You can also sort by relative position in the result set, where the first fieldin the result set is 1. The next field is 2, and so on.
SELECT supplier_city
FROM suppliers
WHERE supplier_name = 'IBM'
ORDER BY 1 DESC;
This wouldreturn all records sorted by the supplier_city field in descending order, sincethe supplier_city field is in position #1 in the result set.
union中order by的使用:
You have to use the Order By at the end of ALL the unions.
the ORDER BY is considered to apply to the whole UNION result(it's effectivelygot lower binding priority than the UNION).
The ORDER BY clause just needs to be the last statement, after you've done allyour unioning. You can union several sets together, then put an ORDER BY clause after the last set.
所以,只能在union的最後一個子查詢中使用order by,而這個order by是針對整個union後的結果集的。So:
如果unoin的幾個子查詢列名不同,如
select supplier_id, supplier_name
from suppliers
UNION
select company_id, company_name
from companies
ORDER BY ?;
這裏的問號如果是company_name,則執行整個查詢會報“company_name:invalididentifier”(當然,單獨執行第二個含orderby的子查詢是沒有問題的);這是因爲unioning後結果集的列名是以第一個參加union的子查詢的列名爲準的;orderby針對的是整個unioning後的結果集。對整個查詢結果來說,無”company_name“這個字段。
如果是supplier_name,則單獨執行第二個含order by的子查詢是會報“supplier_name:invalididentifier”的,而執行整個查詢是沒有問題的,因爲orderby針對的是unioning後的整個結果集,而這“整個結果集”是有supplier_name這列的(以第一個union子查詢的列名作爲unioning後整個結果集的列名)
爲了避免這樣事情的發生,可以:
solution#1: 使用列序號代替實際列名。如:
select supplier_id, supplier_name
from suppliers
UNION
select company_id, company_name
from companies
ORDER BY 2;
solution#2:爲unoin的各個子查詢使用相同的列名,如:
select supplier_id as id, supplier_name as name
from suppliers
UNION
select company_id as id, company_name as name
from companies
ORDER BY name;
這樣,不管是執行整個查詢還是單獨執行包含orderby的最後一個union子查詢,都不會有問題。
Q&A:
http://p2p.wrox.com/sql-language/9505-order-union.html
Q:
引用
I have two tables, TableA and TableB defined as follows,
TableA
A1 int
A2 int
A3 int
TableB
B1 int
B2 int
B3 int
If I try to run this query, SQL Server says syntex failed at the Order By clouse. Is such Order by not allowed in SQL, Any other way to achieve this?
(Select A1, A2 from TableA)
Union All
(Select B1, B2 from TableB Order by B3)
Any help will be appreciated.
A:
引用
First of all, you can not order by a column that is not included in your SELECT list(我注:這句話是錯誤的;可以order by一個不在select列表中的column).Secondly,
when performing a UNION query the ORDER BY clause must be(我注:not “must be”!) a column index not a column
name, because a UNION query does not have column headings(although SQL Server(我注:此處泛指DBMS) pretends that it has by picking the column names used in the first
query although this is not ANSIcompliant]). Assuming you want to order the second column (A2 and B2) your query should look like this:
Code:
SELECT A1, A2
FROM TableA
UNION ALL
SELECT B1, B2
FROM TableB
ORDER BY 2
Conceptually, ORDER BY works by producing the final query table with all the queries joined together (if it is a UNION query), then it orders the query results and does not care about what is in the database.