MySQL的：一個奇怪的查詢：選擇具有多條記錄

Question

我有這樣的一個表中的字段：

---------- 
id | name | tel | email 
---------- 
1 john 0241 re@yah 
2   0534 re@rra 
3   435  fd@rar 
4 geo  43435 re@eae 
5   2347 ui@ear 
6   678  re@yaya 

我要讓這樣的查詢：

SELECT tel, email FROM table WHERE name='geo' 

和輸出是這樣的：

---------- 
    tel | email 

---------- 

43435 re@eae 
2347 ui@ear 
678  re@yaya 

非常感謝你提前

Answer 1

SELECT tel, name FROM 
(SELECT id FROM table WHERE name = 'geo') a 
LEFT JOIN table b ON 
b.id = a.id 
WHERE b.id >= a.id AND b.id < (SELECT id from table WHERE name > '' 
ORDER BY ID ASC 
LIMIT 1) c 

Answer 2

SELECT tel, email 
FROM table1 
JOIN (SELECT MAX(startid) startid, 
      IFNULL(MAX(endid), (SELECT MAX(id) FROM table1)) endid 
    FROM (SELECT id startid, NULL endid FROM table1 WHERE name = 'geo' 
      UNION 
      SELECT NULL, MIN(id)-1 FROM table1 
      WHERE name != '' 
      AND id > (SELECT id FROM table1 WHERE name = 'geo')) x) y 
ON id BETWEEN startid AND endid 

SQLFIDDLE

Answer 3

假設ID正在增加，但沒有名稱的下一個條目屬於同一個，這將是最簡單的更新您的表。

方法如下：

Test data: 
/* 
drop table yourTable; 
CREATE TABLE yourTable 
    (`id` int, `name` varchar(40), `tel` int, `email` varchar(20)) 
; 

INSERT INTO yourTable 
    (`id`, `name`, `tel`, `email`) 
VALUES 
    (1, 'john', 0241, 're@yah'), 
    (2, NULL, 0534, 're@rra'), 
    (3, NULL, 435, 'fd@rar'), 
    (4, 'geo', 43435, 're@eae'), 
    (5, NULL, 2347, 'ui@ear'), 
    (6, NULL, 678, 're@yaya'), 
    (7, 'anything', 789, 'whatever') 
; 
#*/ 


UPDATE yourTable yt1 INNER JOIN (
SELECT 
yt.*, 
COALESCE(name, @prev) AS newName, 
CASE WHEN name IS NOT NULL THEN @prev:=name END 
FROM 
yourTable yt 
, (SELECT @prev:=NULL) v 
ORDER BY id 
) yt2 ON yt1.id = yt2.id 
SET yt1.name = yt2.newName; 

SELECT * FROM yourTable; 

結果：

id name  tel email 
1 john  241 re@yah 
2 john  534 re@rra 
3 john  435 fd@rar 
4 geo   43435 re@eae 
5 geo   2347 ui@ear 
6 geo   678 re@yaya 
7 anything 789 whatever 

然後將查詢應該可以正常工作。

Answer 4

SELECT i.id, i.tel, i.email 
FROM 
    tableX AS t 
    JOIN 
    tableX AS i 
     ON i.id >= t.id 
     AND i.id < COALESCE(
      (SELECT n.id 
      FROM tableX AS n 
      WHERE n.id > t.id 
       AND n.name <> '' 
      ORDER BY n.id ASC 
       LIMIT 1 
     ), 2147483647) 
WHERE 
    t.name = 'geo' ; 

在SQL-Fiddle-1（日Thnx @Barmar）

Answer 5

使用用戶變量測試： -

SELECT tel, email 
FROM 
(
    SELECT Sub1.id, Sub1.name, Sub1.tel, Sub1.email, @name:=IF(name IS NULL, @name, name) AS aName 
    FROM (SELECT * FROM geoname ORDER BY id) Sub1 
    CROSS JOIN (SELECT @name:="") Sub2 
) Sub3 
WHERE aName = 'geo' 

由於Barmar強調，如果名字字段爲空時，在空它，而不是這僅適用於''。如果它需要應付空白以及NULL： -

SELECT tel, email 
FROM 
(
    SELECT Sub1.id, Sub1.name, Sub1.tel, Sub1.email, @name:=IF(name IS NULL OR name = "", @name, name) AS aName 
    FROM (SELECT * FROM Table1 ORDER BY id) Sub1 
    CROSS JOIN (SELECT @name:="") Sub2 
) Sub3 
WHERE aName = 'geo' 

有一點戲劇。挺喜歡通​​過Barmar的解決方案，但有一個發揮，如果我能簡化並消除了幾個子查詢的： -

SELECT a.* 
FROM geoname a 
INNER JOIN 
(
    SELECT a.id AS MinId, IFNULL(MIN(b.id), MAX(c.id)) AS MaxId 
    FROM geoname a 
    LEFT OUTER JOIN geoname b 
    ON a.id < b.id AND b.name IS NOT NULL 
    CROSS JOIN geoname c 
    WHERE a.name = 'geo' 
    GROUP BY a.id 
) Sub1 
ON a.id BETWEEN Sub1.MinId AND Sub1.MaxId 

稍微快似乎和有一個簡單的解釋。

或者用空格以及NULLS

SELECT a.* 
FROM geoname a 
INNER JOIN 
(
    SELECT a.id AS MinId, IFNULL(MIN(b.id), MAX(c.id)) AS MaxId 
    FROM geoname a 
    LEFT OUTER JOIN geoname b 
    ON a.id < b.id AND b.name IS NOT NULL OR b.name = '' 
    CROSS JOIN geoname c 
    WHERE a.name = 'geo' 
    GROUP BY a.id 
) Sub1 
ON a.id BETWEEN Sub1.MinId AND Sub1.MaxId