BETWEEN operator vs. >= AND <=: Is there a performance difference?

后端 未结 5 754
旧巷少年郎
旧巷少年郎 2020-12-19 00:17

These two statements are logically equivalent:

SELECT * FROM table WHERE someColumn BETWEEN 1 AND 100

SELECT * FROM table WHERE someColumn >= 1 AND someC         


        
相关标签:
5条回答
  • 2020-12-19 00:53

    There's no performance benefit, it's just easier to read/write the first one.

    0 讨论(0)
  • 2020-12-19 01:01

    No, no performance benifit. Its just a little candy.

    If you were to check a query comparison, something like

    DECLARE @Table TABLE(
            ID INT
    )
    
    SELECT  *
    FROM    @Table
    WHERE   ID >= 1 AND ID <= 100
    
    SELECT  *
    FROM    @Table 
    WHERE   ID BETWEEN 1 AND 100
    

    and check the execution plan, you should notice that it is exactly the same.

    0 讨论(0)
  • 2020-12-19 01:04

    Oh, but you all referring to the case when search value is on the left side of the where clause.

    Did anybody look at differences when is in the other side of the clause.

    SELECT * FROM table WHERE @date BETWEEN someCol1 AND someCol2

    SELECT * FROM table WHERE someCol1 >= @date AND someCol2 <= @date

    0 讨论(0)
  • 2020-12-19 01:08

    No benefit, just a syntax sugar.

    By using the BETWEEN version, you can avoid function reevaluation in some cases.

    0 讨论(0)
  • 2020-12-19 01:19

    Hmm, here was a surprising result. I don't have SQL Server here, so I tried this in Postgres. Obviously disclaimers apply: this won't necessarily give the same results, your mileage may vary, consult a physician before using. But still ...

    I just wrote a simple query in two different ways:

    select *
    from foo
    where (select code from bar where bar.barid=foo.barid) between 'A' and 'B'
    

    and

    select *
    from foo
    where (select code from bar where bar.barid=foo.barid)>='A'
    and (select code from bar where bar.barid=foo.barid)<='B'
    

    Surprisingly to me, both had almost identical run times. When I did an EXPLAIN PLAN, they gave identical results. Specifically, the first query did the lookup against bar twice, once for the >= test and again for the <= test, just like the second query.

    Conclusion: In Postgres, at least, BETWEEN is indeed just syntactic sugar.

    Personally, I use it regularly because it is clearer to the reader, especially if the value being tested is an expression. Figuring out that two complex expressions are identical can be a non-trivial exercise. Figuring out that two complex expressions SHOULD BE identical even though they're not is even more difficult.

    0 讨论(0)
提交回复
热议问题