Question

In a SQL SERVER 2005 database, please suppose we have a table called RUNNING_COLA_IS_AFRICA.

This table has a unique varchar(50) field called RUNNING_ID.

We have 1 million records in this table.

I would like to write a query that, sorting by RUNNING_ID ASC, and stating a range of 50, produces an output similar to the following:

RUNNING_ID_START RUNNING_ID_END
000000           000103
000104           000767
000892           001492
001576           011222
012345           013579

This means that:

a) The number of the records between 000000 and 000103 is 50;
b) The number of the records between 000104 and 000767 is 50;
c) The number of the records between 000892 and 001492 is 50;
d) The number of the records between 001576 and 011222 is 50;
e) The number of the records between 012345 and 013579 is <= 50.

Of course, min(RUNNING_ID) = 000000 and max(RUNNING_ID) = 013579 because they are ordered by RUNNING_ID ASC.

How could I achieve this in SQL SERVER 2005?

Thank you in advance for your kind help.

Answer 1

WITH
  sequenced
AS
(
  SELECT
    ROW_NUMBER() OVER (ORDER BY running_id) - 1 AS sequence_id,
    *
  FROM
    RUNNING_COLA_IS_AFRICA
)
SELECT
  sequence_id / 50     AS group_id,
  MIN(running_id)      AS running_id_first,
  MAX(running_id)      AS running_id_last,
  COUNT(*)             AS size_of_group
FROM
  sequenced
GROUP BY
  sequence_id / 50