MySQL 的CASE WHEN 語(yǔ)句使用說(shuō)明

2020-01-19 00:03:45

字體：大中小

來(lái)源：轉(zhuǎn)載

供稿：網(wǎng)友

mysql數(shù)據(jù)庫(kù)中CASE WHEN語(yǔ)句。

case when語(yǔ)句，用于計(jì)算條件列表并返回多個(gè)可能結(jié)果表達(dá)式之一。

CASE 具有兩種格式：

簡(jiǎn)單 CASE 函數(shù)將某個(gè)表達(dá)式與一組簡(jiǎn)單表達(dá)式進(jìn)行比較以確定結(jié)果。

CASE 搜索函數(shù)計(jì)算一組布爾表達(dá)式以確定結(jié)果。
兩種格式都支持可選的 ELSE 參數(shù)。

語(yǔ)法
簡(jiǎn)單 CASE 函數(shù)：

復(fù)制代碼代碼如下:

CASE input_expression
    WHEN when_expression THEN result_expression
        [ ...n ]
    [ 
        ELSE else_result_expression
    END

CASE 搜索函數(shù)：

復(fù)制代碼代碼如下:

CASE    
WHEN Boolean_expression THEN result_expression
        [ ...n ]
    [ 
        ELSE else_result_expression
    END

參數(shù)
input_expression

是使用簡(jiǎn)單 CASE 格式時(shí)所計(jì)算的表達(dá)式。Input_expression 是任何有效的 Microsoft? SQL Server? 表達(dá)式。

WHEN when_expression

使用簡(jiǎn)單 CASE 格式時(shí) input_expression 所比較的簡(jiǎn)單表達(dá)式。When_expression 是任意有效的 SQL Server 表達(dá)式。Input_expression 和每個(gè) when_expression 的數(shù)據(jù)類型必須相同，或者是隱性轉(zhuǎn)換。

占位符，表明可以使用多個(gè) WHEN when_expression THEN result_expression 子句或 WHEN Boolean_expression THEN result_expression 子句。

THEN result_expression

當(dāng) input_expression = when_expression 取值為 TRUE，或者 Boolean_expression 取值為 TRUE 時(shí)返回的表達(dá)式。
result expression 是任意有效的 SQL Server 表達(dá)式。

ELSE else_result_expression

當(dāng)比較運(yùn)算取值不為 TRUE 時(shí)返回的表達(dá)式。如果省略此參數(shù)并且比較運(yùn)算取值不為 TRUE，CASE 將返回 NULL 值。Else_result_expression 是任意有效的 SQL Server 表達(dá)式。Else_result_expression 和所有 result_expression 的數(shù)據(jù)類型必須相同，或者必須是隱性轉(zhuǎn)換。

WHEN Boolean_expression

使用 CASE 搜索格式時(shí)所計(jì)算的布爾表達(dá)式。Boolean_expression 是任意有效的布爾表達(dá)式。

結(jié)果類型

從 result_expressions 和可選 else_result_expression 的類型集合中返回最高的優(yōu)先規(guī)則類型。有關(guān)更多信息，請(qǐng)參見數(shù)據(jù)類型的優(yōu)先順序。

結(jié)果值

簡(jiǎn)單 CASE 函數(shù)：
計(jì)算 input_expression，然后按指定順序?qū)γ總€(gè) WHEN 子句的 input_expression = when_expression 進(jìn)行計(jì)算。

返回第一個(gè)取值為 TRUE 的 (input_expression = when_expression) 的 result_expression。

如果沒有取值為 TRUE 的 input_expression = when_expression，則當(dāng)指定 ELSE 子句時(shí) SQL Server 將返回 else_result_expression；若沒有指定 ELSE 子句，則返回 NULL 值。
CASE 搜索函數(shù)：
按指定順序?yàn)槊總€(gè) WHEN 子句的 Boolean_expression 求值。

返回第一個(gè)取值為 TRUE 的 Boolean_expression 的 result_expression。

如果沒有取值為 TRUE 的 Boolean_expression，則當(dāng)指定 ELSE 子句時(shí) SQL Server 將返回 else_result_expression；若沒有指定 ELSE 子句，則返回 NULL 值。

下面分享一些mysql case when語(yǔ)句的例子。

A. 使用帶有簡(jiǎn)單 CASE 函數(shù)的 SELECT 語(yǔ)句
在 SELECT 語(yǔ)句中，簡(jiǎn)單 CASE 函數(shù)僅檢查是否相等，而不進(jìn)行其它比較。

例子，使用 CASE 函數(shù)更改圖書分類顯示。

復(fù)制代碼代碼如下:

USE pubs
GO
SELECT      Category = 
         CASE type
            WHEN 'popular_comp' THEN 'Popular Computing'
            WHEN 'mod_cook' THEN 'Modern Cooking'
            WHEN 'business' THEN 'Business'
            WHEN 'psychology' THEN 'Psychology'
            WHEN 'trad_cook' THEN 'Traditional Cooking'
            ELSE 'Not yet categorized'
         END,
      CAST(title AS varchar(25)) AS 'Shortened Title',
      price AS Price
FROM titles
WHERE price IS NOT NULL
ORDER BY type, price
COMPUTE AVG(price) BY type
GO

注釋，后來(lái)我試了一下不讓用category=。

我使用的代碼為：

復(fù)制代碼代碼如下:

SELECT  
  case gender
 WHEN 1 THEN 'NAN'
 WHEN 0 THEN 'NV'
end  as gender
FROM
t_swidy_day_nutrient

結(jié)果集：

Category               Shortened Title              Price
------------------- ------------------------- --------------------------
Business               You Can Combat Computer S 2.99
Business               Cooking with Computers: S 11.95
Business               The Busy Executive's Data 19.99
Business               Straight Talk About Compu 19.99

                                                 avg
                                                 ==========================
                                                 13.73

Category               Shortened Title              Price
------------------- ------------------------- --------------------------
Modern Cooking         The Gourmet Microwave        2.99
Modern Cooking         Silicon Valley Gastronomi 19.99

                                                 avg
                                                 ==========================
                                                 11.49

Category               Shortened Title              Price
------------------- ------------------------- --------------------------
Popular Computing      Secrets of Silicon Valley 20.00
Popular Computing      But Is It User Friendly? 22.95

                                                 avg
                                                 ==========================
                                                 21.48

Category               Shortened Title              Price
------------------- ------------------------- --------------------------
Psychology             Life Without Fear            7.00
Psychology             Emotional Security: A New 7.99
Psychology             Is Anger the Enemy?          10.95
Psychology             Prolonged Data Deprivatio 19.99
Psychology             Computer Phobic AND Non-P 21.59

                                                 avg
                                                 ==========================
                                                 13.50

Category               Shortened Title              Price
------------------- ------------------------- --------------------------
Traditional Cooking Fifty Years in Buckingham 11.95
Traditional Cooking Sushi, Anyone?               14.99
Traditional Cooking Onions, Leeks, and Garlic 20.95

                                                 avg
                                                 ==========================
                                                 15.96

(21 row(s) affected)

B. 使用帶有簡(jiǎn)單 CASE 函數(shù)和 CASE 搜索函數(shù)的

SELECT 語(yǔ)句
在 SELECT 語(yǔ)句中，CASE 搜索函數(shù)允許根據(jù)比較值在結(jié)果集內(nèi)對(duì)值進(jìn)行替換。

例子：根據(jù)圖書的價(jià)格范圍將價(jià)格（money 列）顯示為文本注釋。

復(fù)制代碼代碼如下:

USE pubs
GO
SELECT       'Price Category' = 
         CASE 
            WHEN price IS NULL THEN 'Not yet priced'
            WHEN price < 10 THEN 'Very Reasonable Title'
            WHEN price >= 10 and price < 20 THEN 'Coffee Table Title'
            ELSE 'Expensive book!'
         END,
      CAST(title AS varchar(20)) AS 'Shortened Title'
FROM titles
ORDER BY price
GO

結(jié)果集：

Price Category           Shortened Title
--------------------- --------------------
Not yet priced           Net Etiquette
Not yet priced           The Psychology of Co
Very Reasonable Title The Gourmet Microwav
Very Reasonable Title You Can Combat Compu
Very Reasonable Title Life Without Fear
Very Reasonable Title Emotional Security:
Coffee Table Title       Is Anger the Enemy?
Coffee Table Title       Cooking with Compute
Coffee Table Title       Fifty Years in Bucki
Coffee Table Title       Sushi, Anyone?
Coffee Table Title       Prolonged Data Depri
Coffee Table Title       Silicon Valley Gastr
Coffee Table Title       Straight Talk About
Coffee Table Title       The Busy Executive's
Expensive book!          Secrets of Silicon V
Expensive book!          Onions, Leeks, and G
Expensive book!          Computer Phobic And
Expensive book!          But Is It User Frien

(18 row(s) affected)

C. 使用帶有 SUBSTRING 和 SELECT 的 CASE 函數(shù)

例子，使用 CASE 和 THEN 生成一個(gè)有關(guān)作者、圖書標(biāo)識(shí)號(hào)和每個(gè)作者所著圖書類型的列表。

復(fù)制代碼代碼如下:

USE pubs
SELECT SUBSTRING((RTRIM(a.au_fname) + ' '+ 
      RTRIM(a.au_lname) + ' '), 1, 25) AS Name, a.au_id, ta.title_id,
      Type = 
  CASE 
       WHEN SUBSTRING(ta.title_id, 1, 2) = 'BU' THEN 'Business'
       WHEN SUBSTRING(ta.title_id, 1, 2) = 'MC' THEN 'Modern Cooking'
       WHEN SUBSTRING(ta.title_id, 1, 2) = 'PC' THEN 'Popular Computing'
       WHEN SUBSTRING(ta.title_id, 1, 2) = 'PS' THEN 'Psychology'
       WHEN SUBSTRING(ta.title_id, 1, 2) = 'TC' THEN 'Traditional Cooking'
  END
FROM titleauthor ta JOIN authors a ON ta.au_id = a.au_id

結(jié)果集：

Name                         au_id          title_id Type
------------------------- ----------- -------- -------------------
Johnson White                172-32-1176 PS3333      Psychology
Marjorie Green               213-46-8915 BU1032      Business
Marjorie Green               213-46-8915 BU2075      Business
Cheryl Carson                238-95-7766 PC1035      Popular Computing
Michael O'Leary              267-41-2394 BU1111      Business
Michael O'Leary              267-41-2394 TC7777      Traditional Cooking
Dean Straight                274-80-9391 BU7832      Business
Abraham Bennet               409-56-7008 BU1032      Business
Ann Dull                     427-17-2319 PC8888      Popular Computing
Burt Gringlesby              472-27-2349 TC7777      Traditional Cooking
Charlene Locksley            486-29-1786 PC9999      Popular Computing
Charlene Locksley            486-29-1786 PS7777      Psychology
Reginald Blotchet-Halls      648-92-1872 TC4203      Traditional Cooking
Akiko Yokomoto               672-71-3249 TC7777      Traditional Cooking
Innes del Castillo           712-45-1867 MC2222      Modern Cooking
Michel DeFrance              722-51-5454 MC3021      Modern Cooking
Stearns MacFeather           724-80-9391 BU1111      Business
Stearns MacFeather           724-80-9391 PS1372      Psychology
Livia Karsen                 756-30-7391 PS1372      Psychology
Sylvia Panteley              807-91-6654 TC3218      Traditional Cooking
Sheryl Hunter                846-92-7186 PC8888      Popular Computing
Anne Ringer                  899-46-2035 MC3021      Modern Cooking
Anne Ringer                  899-46-2035 PS2091      Psychology
Albert Ringer                998-72-3567 PS2091      Psychology
Albert Ringer                998-72-3567 PS2106      Psychology
(25 row(s) affected)

CASE 可能是 SQL 中被誤用最多的關(guān)鍵字之一。
雖然，可能以前用過這個(gè)關(guān)鍵字來(lái)創(chuàng)建字段，但是它還具有更多用法。

例如，可以在 WHERE 子句中使用 CASE。

首先，來(lái)看下 CASE 的語(yǔ)法。在一般的 SELECT 中，其語(yǔ)法如下：

復(fù)制代碼代碼如下:

SELECT <myColumnSpec> = 
CASE 
WHEN <A> THEN <somethingA> 
WHEN <B> THEN <somethingB> 
ELSE <somethingE> 
END

以上代碼，需要用具體的參數(shù)代替尖括號(hào)中的內(nèi)容。

例子：

復(fù)制代碼代碼如下:

USE pubs 
GO 
SELECT 
    Title, 
    'Price Range' = 
    CASE 
        WHEN price IS NULL THEN 'Unpriced' 
        WHEN price < 10 THEN 'Bargain' 
        WHEN price BETWEEN 10 and 20 THEN 'Average' 
        ELSE 'Gift to impress relatives' 
    END 
FROM titles 
ORDER BY price 
GO

這是 CASE 的典型用法，但是使用 CASE 其實(shí)可以做更多的事情。

比方說(shuō)下面的 GROUP BY 子句中的 CASE：

復(fù)制代碼代碼如下:

SELECT 'Number of Titles', Count(*) 
FROM titles 
GROUP BY 
    CASE 
        WHEN price IS NULL THEN 'Unpriced' 
        WHEN price < 10 THEN 'Bargain' 
        WHEN price BETWEEN 10 and 20 THEN 'Average' 
        ELSE 'Gift to impress relatives' 
    END 
GO

甚至還可以組合這些選項(xiàng)，添加一個(gè) ORDER BY 子句，例如：

復(fù)制代碼代碼如下:

USE pubs 
GO 
SELECT 
    CASE 
        WHEN price IS NULL THEN 'Unpriced' 
        WHEN price < 10 THEN 'Bargain' 
        WHEN price BETWEEN 10 and 20 THEN 'Average' 
        ELSE 'Gift to impress relatives' 
    END AS Range, 
    Title 
FROM titles 
GROUP BY 
    CASE 
        WHEN price IS NULL THEN 'Unpriced' 
        WHEN price < 10 THEN 'Bargain' 
        WHEN price BETWEEN 10 and 20 THEN 'Average' 
        ELSE 'Gift to impress relatives' 
    END, 
    Title 
ORDER BY 
    CASE 
        WHEN price IS NULL THEN 'Unpriced' 
        WHEN price < 10 THEN 'Bargain' 
        WHEN price BETWEEN 10 and 20 THEN 'Average' 
        ELSE 'Gift to impress relatives' 
    END, 
    Title 
GO

注意，為了在 GROUP BY 塊中使用 CASE，查詢語(yǔ)句需要在 GROUP BY 塊中重復(fù) SELECT 塊中的 CASE 塊。

除了選擇自定義字段之外，在很多情況下 CASE 都非常有用。

稍加深入，還可以得到以前認(rèn)為不可能得到的分組排序結(jié)果集。

使用CASE WHEN進(jìn)行字符串替換處理

復(fù)制代碼代碼如下:

/* 
 mysql> select * from sales; 
 +-----+------------+--------+--------+--------+------+------------+ 
 | num | name       | winter | spring | summer | fall | category   | 
 +-----+------------+--------+--------+--------+------+------------+ 
 |   1 | Java       |   1067 |    200 |    150 |  267 | Holiday    | 
 |   2 | C          |    970 |    770 |    531 |  486 | Profession | 
 |   3 | JavaScript |     53 |     13 |     21 |  856 | Literary   | 
 |   4 | SQL        |    782 |    357 |    168 |  250 | Profession | 
 |   5 | Oracle     |    589 |    795 |    367 |  284 | Holiday    | 
 |   6 | MySQL      |    953 |    582 |    336 |  489 | Literary   | 
 |   7 | Cplus      |    752 |    657 |    259 |  478 | Literary   | 
 |   8 | Python     |     67 |     23 |     83 |  543 | Holiday    | 
 |   9 | PHP        |    673 |     48 |    625 |   52 | Profession | 
 +-----+------------+--------+--------+--------+------+------------+ 
 9 rows in set (0.01 sec) 
 mysql> SELECT name AS Name, 
     -> CASE category 
     -> WHEN "Holiday" THEN "Seasonal" 
     -> WHEN "Profession" THEN "Bi_annual" 
     -> WHEN "Literary" THEN "Random" END AS "Pattern" 
     -> FROM sales; 
 +------------+-----------+ 
 | Name       | Pattern   | 
 +------------+-----------+ 
 | Java       | Seasonal  | 
 | C          | Bi_annual | 
 | JavaScript | Random    | 
 | SQL        | Bi_annual | 
 | Oracle     | Seasonal  | 
 | MySQL      | Random    | 
 | Cplus      | Random    | 
 | Python     | Seasonal  | 
 | PHP        | Bi_annual | 
 +------------+-----------+ 
 9 rows in set (0.00 sec) 
 */ 
 Drop table sales; 
 CREATE TABLE sales( 
     num MEDIUMINT NOT NULL AUTO_INCREMENT, 
     name CHAR(20), 
     winter INT, 
     spring INT, 
     summer INT, 
     fall INT, 
     category CHAR(13), 
     primary key(num) 
 )type=MyISAM; 
 insert into sales value(1, 'Java', 1067 , 200, 150, 267,'Holiday'); 
 insert into sales value(2, 'C',970,770,531,486,'Profession'); 
 insert into sales value(3, 'JavaScript',53,13,21,856,'Literary'); 
 insert into sales value(4, 'SQL',782,357,168,250,'Profession'); 
 insert into sales value(5, 'Oracle',589,795,367,284,'Holiday'); 
 insert into sales value(6, 'MySQL',953,582,336,489,'Literary'); 
 insert into sales value(7, 'Cplus',752,657,259,478,'Literary'); 
 insert into sales value(8, 'Python',67,23,83,543,'Holiday'); 
 insert into sales value(9, 'PHP',673,48,625,52,'Profession'); 
 select * from sales; 
 SELECT name AS Name, 
 CASE category 
 WHEN "Holiday" THEN "Seasonal" 
 WHEN "Profession" THEN "Bi_annual" 
 WHEN "Literary" THEN "Random" END AS "Pattern" 
 FROM sales; 

簡(jiǎn)單語(yǔ)句

復(fù)制代碼代碼如下:

 
SELECT CASE WHEN 10*2=30 THEN '30 correct' 
WHEN 10*2=40 THEN '40 correct' 
ELSE 'Should be 10*2=20' 
END; 

多重表達(dá)式

復(fù)制代碼代碼如下:

 
SELECT CASE 10*2 
WHEN 20 THEN '20 correct' 
WHEN 30 THEN '30 correct' 
WHEN 40 THEN '40 correct' 
END; 

在SELECT查詢中使用CASE WHEN

復(fù)制代碼代碼如下:

 
/* 
mysql> SELECT Name, RatingID AS Rating, 
-> CASE RatingID 
-> WHEN 'R' THEN 'Under 17 requires an adult.' 
-> WHEN 'X' THEN 'No one 17 and under.' 
-> WHEN 'NR' THEN 'Use discretion when renting.' 
-> ELSE 'OK to rent to minors.' 
-> END AS Policy 
-> FROM DVDs 
-> ORDER BY Name; 
+-----------+--------+------------------------------+ 
| Name | Rating | Policy | 
+-----------+--------+------------------------------+ 
| Africa | PG | OK to rent to minors. | 
| Amadeus | PG | OK to rent to minors. | 
| Christmas | NR | Use discretion when renting. | 
| Doc | G | OK to rent to minors. | 
| Falcon | NR | Use discretion when renting. | 
| Mash | R | Under 17 requires an adult. | 
| Show | NR | Use discretion when renting. | 
| View | NR | Use discretion when renting. | 
+-----------+--------+------------------------------+ 
8 rows in set (0.01 sec) 
*/ 
Drop table DVDs; 
CREATE TABLE DVDs ( 
ID SMALLINT NOT NULL AUTO_INCREMENT PRIMARY KEY, 
Name VARCHAR(60) NOT NULL, 
NumDisks TINYINT NOT NULL DEFAULT 1, 
RatingID VARCHAR(4) NOT NULL, 
StatID CHAR(3) NOT NULL 
) 
ENGINE=INNODB; 
INSERT INTO DVDs (Name, NumDisks, RatingID, StatID) 
VALUES ('Christmas', 1, 'NR', 's1'), 
('Doc', 1, 'G', 's2'), 
('Africa', 1, 'PG', 's1'), 
('Falcon', 1, 'NR', 's2'), 
('Amadeus', 1, 'PG', 's2'), 
('Show', 2, 'NR', 's2'), 
('View', 1, 'NR', 's1'), 
('Mash', 2, 'R', 's2'); 
SELECT Name, RatingID AS Rating, 
CASE RatingID 
WHEN 'R' THEN 'Under 17 requires an adult.' 
WHEN 'X' THEN 'No one 17 and under.' 
WHEN 'NR' THEN 'Use discretion when renting.' 
ELSE 'OK to rent to minors.' 
END AS Policy 
FROM DVDs 
ORDER BY Name; 