postgresql 数组函数全集

1. 分组聚合

ARRAY_AGG(参数1：字段名称) : 聚合同一个字段的多个值为一个数组对象

select product_name_in_sci_pdp,count(1),ARRAY_AGG(npi_product_name) from dim_product_mapping GROUP BY product_name_in_sci_pdp

2 数组拆分

2.1 数组拆分，还原为多行

select 'A' as fieldName1,'B' as fieldName2, "unnest"(ARRAY['CN','AR','BR','IN']) AS plant

2.2 数组拆分，重组为一个字符串

array_to_string(参数1：数组，参数2：拼接字符) : 将数组中的值拼接成新的字符串

select product_name_in_sci_pdp,count(1),array_to_string(ARRAY_AGG(npi_product_name),',') from dim_product_mapping GROUP BY product_name_in_sci_pdp

2.3 字符串拆分为数组

string_to_array(参数1：原始字符串，参数2：拆分字符) : 将原始字符串按拆分字符串拆分为一个数组

with t_level_1 as(
SELECT '123,456,789'::VARCHAR as include_plants
)
SELECT string_to_array(include_plants, ',') as plants from t_level_1

3. 数组去重

没有现成的方法，需要自己写一个方法

CREATE OR REPLACE FUNCTION "public"."array_distinct"(anyarray)RETURNS "pg_catalog"."anyarray" AS $BODY$ SELECT ARRAY(SELECT DISTINCT unnest($1));
$BODY$LANGUAGE sql VOLATILECOST 100

示例

select product_name_in_sci_pdp,count(1),array_distinct(ARRAY_AGG(npi_product_name)) from dim_product_mapping GROUP BY product_name_in_sci_pdp

4. 数组取第一个元素

需要去一下重

SELECT demand_region,(array_distinct(ARRAY_AGG(demand_factory)))[1] from dim_factory_sourcing_dependency where demand_process='Ship FG'  GROUP BY demand_region ORDER  BY demand_region

5.单个元素关联数组

使用 any() 方法实现

OR group_id = any(string_to_array(mgmt_team_group, ','))

其他的后续补充…
参考：
https://blog.csdn.net/shulinshulinzi/article/details/115245325?spm=1001.2101.3001.6650.18&utm_medium=distribute.pc_relevant.none-task-blog-2%7Edefault%7EOPENSEARCH%7ERate-18.pc_relevant_default&depth_1-utm_source=distribute.pc_relevant.none-task-blog-2%7Edefault%7EOPENSEARCH%7ERate-18.pc_relevant_default&utm_relevant_index=23