Question

我尝试在 data.stackexchange.com 上撰写一个查询，这可以让我回答问题的答案，这些用户在撰写超过1000次的帖子和拥有徽章的帖子的用户比例是多少某一类。

问题：如果我写AND ViewCount > 1000，我会得到与WHERE ViewCount > 1000相同的结果我不明白为什么会发生

使用WHERE的查询：

SELECT 
TOP 1000 
Class, (Count(Badges.UserId)* 100 / (SELECT COUNT(*) From Badges)) AS Percentage 
FROM Badges 
WHERE Badges.UserId IN (SELECT Posts.OwnerUserId FROM Posts 
INNER JOIN 
Badges ON Posts.OwnerUserId = Badges.UserId 
WHERE ViewCount > 1000) 
GROUP BY Class 
ORDER BY Class

使用AND的查询：

SELECT 
TOP 1000 
Class, (Count(Badges.UserId)* 100 / (SELECT COUNT(*) From Badges)) AS Percentage 
FROM Badges 
WHERE Badges.UserId IN (SELECT Posts.OwnerUserId FROM Posts 
INNER JOIN 
Badges ON Posts.OwnerUserId = Badges.UserId 
AND ViewCount > 1000) 
GROUP BY Class 
ORDER BY Class

输出（出于某种原因在两种情况下都是相同的）：

1级 - 2％

2级--15％

3级--45％

Answer 1

您正在使用AND连接ON JOIN条件中的两个表达式。专注于内部查询：

SELECT Posts.OwnerUserId FROM Posts 
INNER JOIN Badges ON Posts.OwnerUserId = Badges.UserId 
WHERE ViewCount > 1000

对战：

SELECT Posts.OwnerUserId FROM Posts 
INNER JOIN Badges
ON (
    Posts.OwnerUserId = Badges.UserId 
    AND ViewCount > 1000
)

这两个查询在逻辑上是等价的。

Answer 2

如果您对以下相同的情况感到困惑，请查看此内容： 在SQL SERVER中

CREATE TABLE #a(catId INT, catName VARCHAR(50))
CREATE TABLE #b(productId INT, catId INT)

INSERT INTO #a VALUES 
(1, 'Test1'),
(2, 'Test2'),
(3, 'Test3'),
(4, 'Test4')

INSERT INTO #b VALUES 
(5, 1),
(6, 2),
(7, 3),
(8, 1)

--Return data what exists in both tables
SELECT a.*, b.productId
FROM #a a
INNER JOIN #b b ON b.catId = a.catId
AND b.catId < 3

--Same as above
SELECT a.*, b.productId
FROM #a a
INNER JOIN #b b ON b.catId = a.catId
WHERE b.catId < 3

--Same as above
SELECT a.*, b.productId
FROM #a a
LEFT JOIN #b b ON b.catId = a.catId
WHERE b.catId < 3

--Return all exists in first table though not exist in second table
SELECT a.*, b.productId
FROM #a a
LEFT JOIN #b b ON b.catId = a.catId
AND b.catId < 3

单个AND条件提供与WHERE条件相同的输出

2 个答案: