Question

如何获取匹配列的名称？（例如姓名/类型/国家/城市）

我的表是：

USER
ID int
NAME varchar
TYPE varchar
COUNTRY varchar
CITY varchar

我的sql：

SELECT *
FROM USER 
WHERE NAME LIKE '%".$search."%' 
    OR TYPE LIKE '%".$search."%' 
    OR COUNTRY LIKE '%".$search."%' 
    OR CITY LIKE '%".$search."%'

Answer 1

你可以这样做：

SELECT *
,   COALESCE(
    CASE WHEN NAME LIKE '%".$search."%' THEN 'name' END
,   CASE WHEN NAME TYPE '%".$search."%' THEN 'type' END
,   CASE WHEN COUNTRY LIKE '%".$search."%' THEN 'country' END
,   CASE WHEN CITY LIKE '%".$search."%' THEN 'city' END
) as matched
FROM USER
WHERE NAME LIKE '%".$search."%'
   OR TYPE LIKE '%".$search."%'
   OR COUNTRY LIKE '%".$search."%'
   OR CITY LIKE '%".$search."%'

如果单个行匹配多个列，matched将返回一个名称，该名称对应于按{name, type, country, city}顺序匹配的第一列。

Answer 2

以下是一种方法，即将OR扩展为明确指定哪个列匹配的UNION：

SELECT ..., GROUP_CONCAT(Match) as Match
FROM
(
    SELECT ..., 'Name' as Match
    FROM USER WHERE 
    NAME LIKE '%".$search."%' 

    UNION

    SELECT ..., 'Type' as Match
    FROM USER WHERE 
    TYPE LIKE '%".$search."%' 

    UNION

    SELECT ..., 'Country' as Match
    FROM USER WHERE 
    COUNTRY LIKE '%".$search."%' 

    UNION

    SELECT ..., 'City' as Match
    FROM USER WHERE 
    CITY LIKE '%".$search."%'
) x
GROUP BY ...

根据Frazz的评论，需要使用包装器分组来消除Match列创建的重复项，否则会阻止UNION的默认不同行为。

获取匹配的列名称

2 个答案: