AWS Athena（Presto）如何将值是数组的映射转置为行答案

【问题标题】：AWS Athena (Presto) how to transpose map where the values are an array to rowsAWS Athena（Presto）如何将值是数组的映射转置为行
【发布时间】：2021-09-17 18:27:41
【问题描述】：

我正在尝试将 Athena 上的复杂结构展平

我的表格中的一行看起来像

id	tags
2003301084	{repeats=[3], copayment=[CoPayment, other], concessionalstatus=[E, T], prescriberdate=[2014-06-29], previouspurchasesonsamescript=[0], age=[49], numberofscripts=[1]}

我想去：

id	tag	value
2003301084	repeats	3
2003301084	copayment	CoPayment
2003301084	copayment	other
2003301084	concessionalstatus	E
2003301084	concessionalstatus	T
2003301084	prescriberdate	2014-06-29
2003301084	previouspurchasesonsamescript	0
2003301084	age	49
2003301084	numberofscripts	1

我找不到这样深入结构的示例

【问题讨论】：

标签： amazon-athena presto

【解决方案1】：

您可以使用UNNEST 将映射键/值对提取到单独的行中，然后将各个数组元素提取到单独的行中：

WITH data(id, tags) AS
(
    VALUES (
        2003301084,
        map_from_entries(array[
            row('repeats', array['3']),
            row('copayment', array['CoPayment', 'other']),
            row('concessionalstatus', array['E', 'T']),
            row('prescriberdate', array['2014-06-29']),
            row('previouspurchasesonsamescript', array['0']),
            row('age', array['49']),
            row('numberofscripts', array['1'])
        ])
    )
),
arrays AS (
    SELECT id, tag, entries
    FROM data CROSS JOIN UNNEST(tags) AS m(tag, entries)
)
SELECT id, tag, value
FROM arrays CROSS JOIN UNNEST(entries) AS a(value)

     id     |              tag              |   value
------------+-------------------------------+------------
 2003301084 | repeats                       | 3
 2003301084 | copayment                     | CoPayment
 2003301084 | copayment                     | other
 2003301084 | concessionalstatus            | E
 2003301084 | concessionalstatus            | T
 2003301084 | prescriberdate                | 2014-06-29
 2003301084 | previouspurchasesonsamescript | 0
 2003301084 | age                           | 49
 2003301084 | numberofscripts               | 1
(9 rows)

【讨论】：