导出数据时,在SQL拼接处,提示 oracle ORA-01704: string literal too long
sql:
WITH already_in AS
(SELECT distinct REGEXP_SUBSTR('-999,CX201702210003,CX201702270003,……,CX201702270004', '[^,]+', 1, level) col from dual
connect by level <= length(regexp_replace(''-999,CX201702210003,CX201702270003,……,CX201702270004', '[^,]*')) + 1)
select pro.*, u.user_name PROJECT_MANAGER_NAME
from cp_p pro left join cdc_u u on u.user_id = pro.PROJECT_MANAGER
where 1 = 1 and exists (select * from already_in al where al.col = pro.project_code)
AND pro.project_manager not in (select info.user_id from cdc_user_info info where info.user_name in ('fff'))
(SELECT distinct REGEXP_SUBSTR('-999,CX201702210003,CX201702270003,……,CX201702270004', '[^,]+', 1, level) col from dual
connect by level <= length(regexp_replace(''-999,CX201702210003,CX201702270003,……,CX201702270004', '[^,]*')) + 1)
select pro.*, u.user_name PROJECT_MANAGER_NAME
from cp_p pro left join cdc_u u on u.user_id = pro.PROJECT_MANAGER
where 1 = 1 and exists (select * from already_in al where al.col = pro.project_code)
AND pro.project_manager not in (select info.user_id from cdc_user_info info where info.user_name in ('fff'))
拼接的字符串超长导致“string” 抛异常。减少了拼接字符后,顺利执行。
注意:在拼接sql时,用于某类型字段的拼接,需要注意其长度。事先预测一下需要拼接的字符最长会达到多少,超长的话可以使用clob类型字段。
尽量避免拼接超长字符用在sql中,如必须,可以声明一个事务变量,或者分页处理(或分段连接处理)。
网上有相关处理方式,引用一下:
有两种方法可以解决:
1.使用存储过程,把超长文本保存在一个变量中,然后再insert update
declare
v_clob clob :='一个长文本';
begin
insert into table values(a,3,:clob);
end;
2.字符串拼接,update使用字符串拼接
update mall_config set category_info='安全防护:3003,' where id=1;
update mall_config set category_info=category_info||'|标准件:1040140,1035382,' where id=1;