Tell me more ×
Stack Overflow is a question and answer site for professional and enthusiast programmers. It's 100% free, no registration required.

I have a table that looks like this:

A 1 
A 2 
B 1 
B 2

And I want to produce a result set that looks like this:

A 1 2 
B 1 2

Is there a SQL statement that will do this? I am using Oracle.

Related questions:

share|improve this question
See similar question advice-using-pivot-table-in-oracle – Tony Andrews Jul 13 '09 at 16:56

7 Answers

It depends on the version of Oracle you're using. If it supports the wm_concat() function, then you can simply do something like this:

SELECT field1, wm_concat(field2) FROM YourTable GROUP BY field2;

wm_concat() basically works just like group_concat() in MySQL. It may not be documented, so fire up ye olde sqlplus and see if it's there.

If it isn't there, then you'll want to implement something equivalent yourself. You can find some instructions on how to do this in the string aggregation page at oracle-base.com.

share|improve this answer
5  
Note that wmsys.wm_concat is undocumented and unsupported. – turbanoff Dec 16 '11 at 3:46

In Oracle 10g+:

SELECT  *
FROM    (
        SELECT  *
        FROM    mytable
        MODEL
        PARTITION BY
                (grouper)
        DIMENSION BY
                (ROW_NUMBER() OVER (PARTITION BY grouper ORDER BY id) AS rn)
        MEASURES
                (val, val AS group_concat, 0 AS mark)
        RULES SEQUENTIAL ORDER (
                group_concat[rn > 1] ORDER BY rn = group_concat[CV() - 1] || ', ' || val[CV()],
                mark[ANY] ORDER BY rn = PRESENTV(mark[CV() + 1], 0, 1)
                )
        )
WHERE   mark = 1
ORDER BY
        grouper

See this article in my blog for explanations:

share|improve this answer
You are totally dependent on mytable being sorted by rn. If it is not sorted you'll get an "ORA-32637: Self cyclic rule in sequential order MODEL". So this variant is buggy. You may want to read another model clause variant on my blog: rwijk.blogspot.com/2008/05/… – Rob van Wijk Jul 29 '09 at 19:20
@Rob: rn here is just an alias for ROW_NUMBER(). I copied it from my article (which used a CTE as a rowsource) but didn't put it into the table. Thanks for noticing. – Quassnoi Jul 29 '09 at 20:36
Yes, I know. But if, for instance, you used another analytic function with a different order by, and that accompanying WINDOW SORT operation is executed last, you get the aforementioned error. So you still have to use ordered rule evaluation to make this idea work. – Rob van Wijk Jul 29 '09 at 21:15
@Rob: right again, updated, checked with 6 - ROW_NUMBER() OVER (PARTITION BY grouper ORDER BY rno) AS rn. Thanks. – Quassnoi Jul 29 '09 at 21:45

Try something like :

SELECT
    field1,
    RTRIM(REPLACE(REPLACE(XMLAgg(XMLElement("x", field2) ORDER BY field2), '<x>'), '</x>', ' ')) AS field2s
  FROM yourTable
  GROUP BY field1

Freely inspired by an answer found in this Oracle forum.

EDIT: this solution proved very resources intensive with requests involving something like 105 rows. I ended up replacing this by custom aggregate functions as suggested by John.

share|improve this answer

Pretty old topic, but it could help others since Oracle improved in the mean time.

The LISTAGG function is what you are looking for (in 11g at least)

share|improve this answer
For a full syntax example, "LISTAGG(name, ', ') WITHIN GROUP (ORDER BY name)". docs.oracle.com/cd/E11882_01/server.112/e10592/functions089.htm – Braden Anderson May 22 at 20:41

User defined aggregate functions: http://www.adp-gmbh.ch/ora/sql/user_def_agg.html

Just copy/paste and use it. Works on 9i.

share|improve this answer
SELECT a , COLLECT(b) FROM foo GROUP BY a

very useful when used in pl/sql - can be casted to a user defined collection.

share|improve this answer

If you have got 10g, then you have to go through the function below:

CREATE OR REPLACE FUNCTION get_separated_value (input_val  in  number)
  RETURN VARCHAR2
IS
  return_text  VARCHAR2(10000) := NULL;
BEGIN
  FOR x IN (SELECT col2 FROM table_name WHERE col1 = input_val) LOOP
    return_text := return_text || ' ' || x.col2 ;
  END LOOP;
  RETURN return_text;
END;
/

So, you can do like:

select col1, get_separated_value(col1) from table_name

Fiddle here

If you have got oracle 11g, you can use listagg :

SELECT 
    age,
    LISTAGG(name, ' ') WITHIN GROUP (ORDER BY name) "names"
FROM table_x
GROUP BY age

Fiddle here for Listagg

share|improve this answer

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.