インデックスが重複してないかチェックするSQL

gistfile1.sql

select
 Z.TABLE_SCHEMA,
 Z.TABLE_NAME,
 Z.idx_column,
 Z.CNT
from 
(
select 
 Y.TABLE_SCHEMA,
 Y.TABLE_NAME,
 Y.idx_column,
 count(Y.idx_column) as CNT
from
(
select 
 X.TABLE_SCHEMA,
 X.TABLE_NAME,
 X.INDEX_NAME,
 X.NON_UNIQUE,
 rtrim(group_concat(concat(X.COLLATION,X.COLUMN_NAME) order by X.SEQ_IN_INDEX asc SEPARATOR '')) as idx_column 
from
 (
  select
    Y.TABLE_SCHEMA,
    Y.TABLE_NAME,
    Y.INDEX_NAME,
    Y.NON_UNIQUE,
    case
    when Y.COLLATION = 'A' then '+'
    when Y.COLLATION = 'D' then '-'
    else '*' 
    end as COLLATION, 
    Y.COLUMN_NAME,
    Y.SEQ_IN_INDEX
    from
     (
      select * from information_schema.STATISTICS
     ) as Y
    where
    Y.TABLE_SCHEMA <> 'mysql'
    order by 
      Y.TABLE_SCHEMA,
      Y.TABLE_NAME,
      Y.INDEX_NAME,
      Y.SEQ_IN_INDEX
) as X
group by
   X.TABLE_SCHEMA,
   X.TABLE_NAME,
   X.INDEX_NAME,
   X.NON_UNIQUE
order by
   X.TABLE_SCHEMA,
   X.TABLE_NAME,
   X.INDEX_NAME
) as Y
group by
  Y.TABLE_SCHEMA,Y.TABLE_NAME,Y.idx_column
) AS Z
where 
  Z.CNT > 1;