Find ranges from a series of numbers in SQL/Oracle

2019-08-12 01:28发布

问题:

I have a table containing a series of numbers 1,2,3,4,5,11,12,13,14,15,101,102,103,104,105,510,511,512,513,515,516,517.
I want an SQL Query, Procedure or Function so that I can get the ranges in the following formats.

From-To:

1-5
11-15
101-105
510-517

OR

1-5, 11-15, 101-105, 510-517

回答1:

You could do it using ROW_NUMBER analytic function. See Find range of consecutive values in a sequence of numbers or dates.

For example,

Range

SQL> with data(num) as(
  2      select  1 from dual union
  3      select  2 from dual union
  4      select  3 from dual union
  5      select  5 from dual union
  6      select  6 from dual union
  7      select  7 from dual union
  8      select 10 from dual union
  9      select 11 from dual union
 10      select 12 from dual union
 11      select 20 from dual
 12  )
 13  select min(num)||'-'|| max(num) as "range"
 14  from (select num,
 15           num-Row_Number() over(order by num)
 16           as rn
 17           from data)
 18  group by rn
 19  order by min(num);

range
-------------------------------------------------
1-3
5-7
10-12
20-20

SQL>

List

SQL> with data(num) as(
  2      select  1 from dual union
  3      select  2 from dual union
  4      select  3 from dual union
  5      select  5 from dual union
  6      select  6 from dual union
  7      select  7 from dual union
  8      select 10 from dual union
  9      select 11 from dual union
 10      select 12 from dual union
 11      select 20 from dual
 12  )
 13  SELECT listagg(range, ',') WITHIN GROUP(
 14  ORDER BY min_num) AS "list"
 15  FROM
 16    (SELECT MIN(num) min_num,
 17      MIN(num)
 18      ||'-'
 19      || MAX(num) range
 20    FROM
 21      (SELECT num, num-Row_Number() over(order by num) AS rn FROM DATA
 22      )
 23    GROUP BY rn
 24    );

list
-------------------------------------------------------------------------
1-3,5-7,10-12,20-20

SQL>

Update OP wants a solution in PL/SQL to store the list in a PL/SQL variable.

Setup

SQL> CREATE TABLE t AS
  2  SELECT *
  3  FROM
  4    ( WITH data(num) AS
  5    ( SELECT 1 FROM dual
  6    UNION
  7    SELECT 2 FROM dual
  8    UNION
  9    SELECT 3 FROM dual
 10    UNION
 11    SELECT 5 FROM dual
 12    UNION
 13    SELECT 6 FROM dual
 14    UNION
 15    SELECT 7 FROM dual
 16    UNION
 17    SELECT 10 FROM dual
 18    UNION
 19    SELECT 11 FROM dual
 20    UNION
 21    SELECT 12 FROM dual
 22    UNION
 23    SELECT 20 FROM dual
 24    )
 25  SELECT * FROM DATA);

Table created.

PL/SQL block

SQL> SET SERVEROUTPUT ON
SQL> DECLARE
  2    v_list VARCHAR2(100);
  3  BEGIN
  4    SELECT listagg(RANGE, ',') WITHIN GROUP(
  5    ORDER BY min_num)
  6    INTO v_list
  7    FROM
  8      (SELECT MIN(num) min_num,
  9        MIN(num)
 10        ||'-'
 11        || MAX(num) range
 12      FROM
 13        (SELECT num, num-Row_Number() over(order by num) AS rn FROM t
 14        )
 15      GROUP BY rn
 16      );
 17    dbms_output.put_line(v_list);
 18  END;
 19  /
1-3,5-7,10-12,20-20

PL/SQL procedure successfully completed.

SQL>


回答2:

The common solution exploits the fact that those values should be sequential:

select min(serialnum), max(serialnum), count(*)
from
 (
   select 
      serialnum,
      -- this returns a meaningless value, but the same meaningless value for sequential numbers
      serialnum - row_number() over (order by serialnum) as dummy
   from tab
 ) dt
group by dummy