Clarify the use of unqualified function names in view definitions

Both MySQL manual and MariaDB KB describe (in identical words) how unqualified table and view names are resolved within CREATE VIEW statement:

Unqualified table or view names in the SELECT statement are interpreted with respect to the default database. A view can refer to tables or views in other databases by qualifying the table or view name with the proper database name.

However, neither explains what happens with unqualified function names, and it is not obvious, as it apparently works differently from table and view names:

MariaDB [test]> create function db1.f() returns int return 1;

Query OK, 0 rows affected (0.019 sec)

MariaDB [test]> create function db2.f() returns int return 2;

Query OK, 0 rows affected (0.058 sec)

MariaDB [test]> use db1;

Database changed

MariaDB [db1]> create view db2.v as select f();

Query OK, 0 rows affected (0.053 sec)

MariaDB [db1]> select * from db2.v;

+------+

| f()  |

+------+

|    2 |

+------+

1 row in set (0.001 sec)

So, it actually uses the function from the same schema where the view resides, not from the default schema. It works the same way in all MariaDB and MySQL versions, so I hope it's intentional and just needs to be documented.