Question

I have a large group of columns in a table (50+) which contain some statistics about a game.

We are simplifying our statistics, so now we want to do two things.

• Drop any columns where the name contains ‘%_nd%’
• Remove the string ‘_er’ from any column names.

Is there an easy way to do this automatically, or do I need to enter all the column names manually?

Answer 1

Well you can do the first by running the following and using the SQL it produces:

select distinct concat('alter table ',table_name,' drop column ',column_name) someSql
from information_schema.columns
where column_name like '%_nd%'

You could do the second with a similar query:

select distinct concat('alter table ',table_name,' change ',column_name, ' ', replace(column_name, '_er', ''), ' ', data_type, case when CHARACTER_MAXIMUM_LENGTH is not null then concat('(', CHARACTER_MAXIMUM_LENGTH, ')') end) someSql
from information_schema.columns
where column_name like '%_er%'