比较列和同一列时查询性能的影响

Question

I really want to know about there would be any performance impact while comparing table column with the same table column like

select id, name from person
where name = coalesce(NULLIF(:application_parameter,''),name)

So in above example if I don't recieve parameter from my application(if it's null or blank) then query will be

select id, name from person
where name = name

This is just a sample, but in my application it's really a complex query. I know there are some alternates to this like create dynamic query(append where condition only if you want it).

But I am only interested to know what will be the performance impact if I compare column with the same column.

Answer 1

Yes. This version:

where name = coalesce(NULLIF(:application_parameter,''),name)

will not take advantage of an index on name . The use of the function impedes index optimization.

It is hard to get queries to both take parameters and to use indexes. You can try:

where name = :application_parameter or :application_parameter is NULL

MySQL might optimize this correctly. Otherwise this might do the trick:

select id, name
from person
where :application_parameter is NULL
union all
select id, name
from person
where name = :application_parameter;

Answer 2

To see the performance impact in mysql you can look at the query after the optimizer had it. Do a

EXPLAIN EXTENDED SELECT aCol FROM aTable WHERE aCol = aCol;

This will generate a WARNING with the result of the optimizer. You can get it by

SHOW WARNINGS;

Here you will find:

SELECT aCol FROM aTable WHERE 1;

So the condition gets optimized away :-)