1. update （select .....） set column1 = column2;
update (select iy.company_name company_name1, cc.company_name_jc company_name2 from income_year_item iy , city_company cc where iy.company_code = cc.code ) set company_name1 = company_name2;
2.只能单行子查询s
update CITY_PROJECT_SCALE_INFO c set (c.value) = ( select d.value from CITY_PROJECT_SCALE_INFO @test d where d.project_id = '7d7fd580a06240b2a9137dc2bbe831e9' and d.project_id = c.project_id and c.company_code = d.company_code) where exists ( select 1 from CITY_PROJECT_SCALE_INFO @test d where d.project_id = '7d7fd580a06240b2a9137dc2bbe831e9' and d.project_id = c.project_id and c.company_code = d.company_code)
请你尤其注意这里的where子句，你可以尝试不写where子句来执行以下这句话，你将会使得CITY_PROJECT_SCALE_INFO中的很多值变成空。
这是因为在oracle的update语句中如果不写where子句,oracle将会默认的把所有的值全部更新，即使你这里使用了子查询并且某在值并不能在子查询里找到，你就会想当然的以为,oracle或许将会跳过这些值吧，你错了，oracle将会把该行的值更新为空。
3.使用merg inot 语句
--更新生产基础字段merge into city_cfg_data_column_common cf1 using city_cfg_data_column_common2 cf2 on (cf1.resourceid = cf2.resourceid) when matched then update set cf1.template_type = cf2.template_type, cf1.chinese_name = cf2.chinese_name, cf1.column_name = cf2.column_name, cf1.column_type = cf2.column_type, cf1.column_size = cf2.column_size when not matched then insert (cf1.resourceid,cf1.template_type,cf1.chinese_name,cf1.column_name,cf1.column_type,cf1.column_size ,cf1.is_can_edit,cf1.is_unique,cf1.is_can_cover,cf1.show_order) values (cf2.resourceid,cf2.template_type,cf2.chinese_name,cf2.column_name,cf2.column_type,cf2.column_size ,cf2.is_can_edit,cf2.is_unique,cf2.is_can_cover,cf2.show_order)
