REGEXP_REPLACE 不同

REGEXP_REPLACE with DISTINCT

我正在尝试将 distinct 与 REGEXP_REPLACE 结合使用,但返回了 0 行。

我在 MySQLP v8.0

中创建了一个测试 table
   CREATE TABLE phone(
   id serial primary key,
   phone_number char(25));

   INSERT INTO phone (phone_number)
   VALUES ('(423) 330-9999');

   INSERT INTO phone (phone_number)
   VALUES ('(423)3309999');

   INSERT INTO phone (phone_number)
   VALUES ('423-330-1111)');

   INSERT INTO phone (phone_number)
   VALUES ('1-423-330-6666');

   INSERT INTO phone (phone_number)
   VALUES ('1A423*330*1111');

   INSERT INTO phone (phone_number)
   VALUES ('5553301111');

-- 然后

select 
   REGEXP_REPLACE(phone_number, '[^0-9]', '',1, 0, 'm') as clean_phone
from phone

--- 工作正常 -> clean_phone 4233309999 4233309999 4233301111 14233306666 14233301111 5553301111

--- 计数

select 
   count(REGEXP_REPLACE(phone_number, '[^0-9]', '',1, 0, 'm')) as 
   clean_phone
from phone

--- 工作正常 -> clean_phone 6

-- 不同 clean_phone

select 
   distinct(REGEXP_REPLACE(phone_number, '[^0-9]', '',1, 0, 'm')) as 
   clean_phone
from phone

--- returns 空 -> clean_phone

我不明白为什么 distinct 不起作用?

Distinct 不是一个函数,因此您不需要 distinct() 而只需要 distinct

select  distinct REGEXP_REPLACE(phone_number, '[^0-9]', '',1, 0, 'm') as 
   clean_phone
from phone

.

    select distinct  clear_phone from(
    select   REGEXP_REPLACE(phone_number, '[^0-9]', '',1, 0, 'm')  clear_phone
   from phone ) t 

如果错误仍然存​​在,您可以尝试使用 insert/select ofn a dummy table

  insert into dummy_table(clear_phone)
  select   REGEXP_REPLACE(phone_number, '[^0-9]', '',1, 0, 'm')  
  from phone;

  select distinct clear_phone from dummy_table;