在 SQL 中删除字符串的一部分

Question

I have a table with a field with this kind of entries:

"text text text"
"text text - Cap. 125"
"text - Cap. 1"

I need to cut " - Cap. (Number)" .

As you can see in some cases there's no "Cap." and in other cases the number has more than 1 digit.

I tried to use REPLACE() and TRIM() , but REPLACE() only works for " - cap." but not for the numbers; and TRIM() cut a specific number of characters.

Also, I think I need something like WHERE field LIKE '% - CAP.%' to verify that I'm changing the right field, because not all the fields need to be changed.

Any suggestions?

Answer 1

SELECT CASE
         WHEN myField RLIKE ' - Cap\\. [[:digit:]]{1,3}$' THEN
           LEFT(myField, 1
             + CHAR_LENGTH(myField)
             - CHAR_LENGTH(' - Cap. ')
             - LOCATE(REVERSE(' - Cap. '), REVERSE(myField))
           )
         ELSE myField
       END
FROM   myTable

See it on sqlfiddle .

Answer 2

SELECT SUBSTRING_INDEX(column_name, 'Cap', 1) FROM table ; // for left
SELECT SUBSTRING_INDEX(column_name, 'Cap', -1) FROM table; // for right