Question

我有以下表结构：

+----------+---------+
|       Trade        |
+----------+---------+
| Trade_Id | Book_Id |
+----------+---------+

+---------+-----------+
|      Book           |
+---------+-----------+
| Book_Id | Book_Name |
+---------+-----------+

+----------+--------------+---------------+
|                 Keyword                 |
+----------+--------------+---------------+
| Trade_Id | Keyword_Name | Keyword_Value |
+----------+--------------+---------------+

现在我想从具有一个特定重复关键字的特定图书中检索交易。我有一个查询工作，但我讨厌它，因为我基本上重复where子句中的wohle查询。

select 
    t.TRADE_ID, 
    kw.KEYWORD_VALUE 
from 
    trade t
    inner join book b               on  b.BOOK_ID = t.BOOK_ID
                                        and b.BOOK_NAME = 'test123'
    inner join TRADE_KEYWORD kw     on  kw.TRADE_ID = t.TRADE_ID
                                        and kw.KEYWORD_NAME = 'ISIN'
where 
    kw.KEYWORD_VALUE in (
        select 
            kw2.KEYWORD_VALUE 
        from 
            trade t2
            inner join book b2              on  b2.BOOK_ID = t2.BOOK_ID
                                                and b2.BOOK_NAME = 'test123'
            inner join TRADE_KEYWORD kw2    on  kw2.TRADE_ID = t2.TRADE_ID
                                                and kw2.KEYWORD_NAME = 'ISIN'
        group by kw2.KEYWORD_VALUE
        having count (*) > 1)

我怎样才能更优雅地做到这一点？谢谢！

Answer 1

select TRADE_ID, KEYWORD_VALUE from
    (select 
        t.TRADE_ID, 
        kw.KEYWORD_VALUE,
        count(kw.KEYWORD_VALUE) over(partition by kw.KEYWORD_VALUE) n
    from 
        trade t
        inner join book b               on  b.BOOK_ID = t.BOOK_ID
                                            and b.BOOK_NAME = 'test123'
        inner join TRADE_KEYWORD kw     on  kw.TRADE_ID = t.TRADE_ID
                                            and kw.KEYWORD_NAME = 'ISIN'
    )                               
where n > 1;

您可以使用分析功能。它们与聚合函数类似，但为每一行提供结果（PARTITION BY类似于GROUP BY）。

获取具有重复值的行

1 个答案: