未验证 提交 df1213f3 编写于 作者: A Artem Zuikov 提交者: GitHub

Try to make if_transform_strings perf test faster (#12598)

上级 16e8c614
<test>
<preconditions>
<table_exists>hits_10m_single</table_exists>
<table_exists>hits_100m_single</table_exists>
</preconditions>
<query>SELECT length(URL) > 1000 ? 'LONG' : 'SHORT' as x FROM hits_100m_single GROUP BY x FORMAT Null</query>
<query>SELECT transform(number, [2, 4, 6], ['google', 'yandex', 'yahoo'], 'other') as x FROM numbers(1000000000) GROUP BY x FORMAT Null</query>
<query>SELECT transform(number, [2, 4, 6], ['google', 'yandex', 'yahoo'], 'other') as x FROM numbers(100000000) GROUP BY x FORMAT Null</query>
<query>SELECT length(URL) > 1000 ? 'LONG' : 'SHORT' as x FROM hits_100m_single GROUP BY x FORMAT Null SETTINGS optimize_if_transform_strings_to_enum = 1</query>
<query>SELECT transform(number, [2, 4, 6], ['google', 'yandex', 'yahoo'], 'other') as x FROM numbers(100000000) GROUP BY x FORMAT Null SETTINGS optimize_if_transform_strings_to_enum = 1</query>
</test>
Markdown is supported
0% .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑!
想要评论请 注册