Eliminate duplicate results in a select query that contains CLOB column

问题

This is the select query:

select orderid,ordernum,orderdate,orderxml from orders

The query returns multiple rows with same ordernum. I tried to use DISTINCT and Group BY but orderxml is a clob column. How can I eliminate duplicate ordernums in my query?

回答1:

You could use an analytic function to identify a single orderid for each ordernum - probably either min or max, but other functions are available, it depends on what you need - in a subquery, and then filter to only get the rows with the identified IDs:

select orderid, ordernum, orderdate, orderxml
from (
  select orderid, ordernum, orderdate, orderxml,
    max(orderid) over (partition by ordernum) as maxorderid
  from orders
)
where orderid = maxorderid;

The inline view gets all the columns and rows from your table, but adds an additional column to its result that has the maximum ID across all rows with the same order number. (You can add any other filters you want in there, or course).

The outer filter then only matches the ID for each order number that matches that maximum value.

This assumes that orderid is unique - at least for an ordernum but presumably globally. You said the orderdate is always the same for the ordernum but you could include that in the partition by if not, possibly modifying the analytic function used.

(And then investigate how and why you are getting duplicates, and try to stop them; then remove the duplicates from your table - carefully...)

来源：https://stackoverflow.com/questions/42604897/eliminate-duplicate-results-in-a-select-query-that-contains-clob-column