Tek-Tips is the largest IT community on the Internet today!

Members share and learn making Tek-Tips Forums the best source of peer-reviewed technical information on the Internet!

  • Congratulations strongm on being selected by the Tek-Tips community for having the most helpful posts in the forums last week. Way to Go!

Concatenate? 1

Status
Not open for further replies.

klunde

IS-IT--Management
Dec 15, 2004
63
SE
I have this sql code

Code:
select source_name, round(sysdate-sup_date) as nday, count(article_id) as arts
from articles_contents_xml
where round(sysdate-sup_date) <=14
group by 
source_name, round(sysdate-sup_date)"
order by source_name, nday

which gives me this list:
Code:
Source_name	nday	arts
Adresseavisen	1	17
Adresseavisen	3	38
Adresseavisen	4	35
Adresseavisen	5	29
Adresseavisen	6	33
Aftenposten 	1	14
Aftenposten 	4	23

I would like to transform this into a new list like this:
Code:
Source_name	Day1	day2	Day3	Day4	Day5	Day6
Adresseavisen	17	-	38	35	29	33
Aftenposten 	14	-	-	23	-	-

Can I use concat() with my Oracle 10 and change the SQL to accomplish this?

</Morten>
 
I think the function you are looking for is called COLLECT rather than CONCAT. Concat is just another name for the || function which joins strings together.

You could do it using something like:

Code:
create table testdat (source_name varchar2(30), nday number, arts number)

insert into testdat values ('Adresseavisen',    1,    17);
insert into testdat values ('Adresseavisen',   3,    38);
insert into testdat values ('Adresseavisen',    4,    35);
insert into testdat values ('Adresseavisen',    5,    29);
insert into testdat values ('Adresseavisen',    6,    33);
insert into testdat values ('Aftenposten',     1,   14);
insert into testdat values ('Aftenposten',     4,    23);

select source_name,
case when instr(long_string, '1:') <> 0 then substr(long_string, instr(long_string, '1:')+2,instr(long_string, ',' ,instr(long_string, '1:')) - instr(long_string, '1:')-2) end as day1,
case when instr(long_string, '2:') <> 0 then substr(long_string, instr(long_string, '2:')+2,instr(long_string, ',' ,instr(long_string, '2:')) - instr(long_string, '2:')-2) end as day2,
case when instr(long_string, '3:') <> 0 then substr(long_string, instr(long_string, '3:')+2,instr(long_string, ',' ,instr(long_string, '3:')) - instr(long_string, '3:')-2) end as day3,
case when instr(long_string, '4:') <> 0 then substr(long_string, instr(long_string, '4:')+2,instr(long_string, ',' ,instr(long_string, '4:')) - instr(long_string, '4:')-2) end as day4,
case when instr(long_string, '5:') <> 0 then substr(long_string, instr(long_string, '5:')+2,instr(long_string, ',' ,instr(long_string, '5:')) - instr(long_string, '5:')-2) end as day5,
case when instr(long_string, '6:') <> 0 then substr(long_string, instr(long_string, '6:')+2,instr(long_string, ',' ,instr(long_string, '6:')) - instr(long_string, '6:')-2) end as day6
from
(select source_name, tab_to_string(CAST(collect(nday||':'||arts) AS t_varchar2_tab)) as long_string
from testdat
group by source_name)
 
I forgot to mention that you need to create a couple of database objects for that to work.

Code:
CREATE OR REPLACE TYPE t_varchar2_tab AS TABLE OF VARCHAR2(4000);
/

CREATE OR REPLACE FUNCTION tab_to_string (p_varchar2_tab  IN  t_varchar2_tab,
                                          p_delimiter     IN  VARCHAR2 DEFAULT ',') RETURN VARCHAR2 IS
  l_string     VARCHAR2(32767);
BEGIN
  FOR i IN p_varchar2_tab.FIRST .. p_varchar2_tab.LAST LOOP
    l_string := l_string || p_varchar2_tab(i) || p_delimiter;
  END LOOP;
  RETURN l_string;
END tab_to_string;
/
 
Morten,

There are many ways to resolve your need. Here is one of the ways:
Code:
select source_name
      ,substr(nvl(to_char(max(decode(nday,1,arts,null))),'-'),1,4) Day1
      ,substr(nvl(to_char(max(decode(nday,2,arts,null))),'-'),1,4) Day2
      ,substr(nvl(to_char(max(decode(nday,3,arts,null))),'-'),1,4) Day3
      ,substr(nvl(to_char(max(decode(nday,4,arts,null))),'-'),1,4) Day4
      ,substr(nvl(to_char(max(decode(nday,5,arts,null))),'-'),1,4) Day5
      ,substr(nvl(to_char(max(decode(nday,6,arts,null))),'-'),1,4) Day6
  from (select source_name, round(sysdate-sup_date) as nday, count(article_id) as arts
          from articles_contents_xml
         where round(sysdate-sup_date) <=14
         group by source_name, round(sysdate-sup_date)
         order by source_name, nday)
 group by source_name
/

SOURCE_NAME   DAY1 DAY2 DAY3 DAY4 DAY5 DAY6
------------- ---- ---- ---- ---- ---- ----
Adresseavisen 17   -    38   35   29   33
Aftenposten   14   -    -    23   -    -
Let us know if this satisfactorily resolves your need.

[santa]Mufasa
(aka Dave of Sandy, Utah, USA)
[I provide low-cost, remote Database Administration services: www.dasages.com]
“Beware of those that seek to protect you from harm or risk. The cost will be your freedoms and your liberty.”
 
Thank you both for very good help. I went for the second method since I'm writing up this page in an asp page and therefor didn't want to create new database objects.

A small? follow-up question. Would it be possible to change the sql code to allow n-number of days?

</Morten>
 
Klunde said:
Would it be possible to change the sql code to allow n-number of days?
Yes, one method to achieve your result is to use a technique called "SQL Writing SQL". In the sample code, below, I use SQL*Plus and SQL commands to write the necessary code to accommodate "n-number of days". First, let's look at some sample data that extends the data beyond your original 6 rows:
Code:
select source_name, round(sysdate-sup_date) as nday, count(article_id) as arts
from articles_contents_xml
where round(sysdate-sup_date) <=14
group by 
source_name, round(sysdate-sup_date)"
order by source_name, nday;

SOURCE_NAME         NDAY       ARTS
------------- ---------- ----------
Adresseavisen          1         17
Adresseavisen          3         38
Adresseavisen          4         35
Adresseavisen          5         29
Adresseavisen          6         33
Adresseavisen          8         43
Adresseavisen         10         53
Adresseavisen         12         63
Aftenposten            1         14
Aftenposten            4         23
Next, here is the code, which I saved into a script named "tt_507.sql", that accommodates n-number of days-worth of data:
Code:
set pagesize 0
set feedback off
set echo off
set termout off
col max_days new_value max_days
select max(nday) max_days
  from (select round(sysdate-sup_date) as nday
          from articles_contents_xml
         where round(sysdate-sup_date) <=14
         group by source_name, round(sysdate-sup_date));
spool temp.sql
set termout on
prompt Select source_name
select ',substr(nvl(to_char(max(decode(nday,'||cnt||',arts,null))),''-''),1,6) Day'||cnt
  from (select level cnt from dual connect by level <= &max_days);
prompt   from (select source_name, round(sysdate-sup_date) as nday, count(article_id) as arts
prompt           from articles_contents_xml
prompt          where round(sysdate-sup_date) <=14
prompt          group by source_name, round(sysdate-sup_date)
prompt          order by source_name, nday)
prompt  group by source_name
prompt /
spool off
set pagesize 35
@temp
Here are the results of running "tt_507.sql":
Code:
SQL> @tt_507
Select source_name
,substr(nvl(to_char(max(decode(nday,1,arts,null))),'-'),1,6) Day1
,substr(nvl(to_char(max(decode(nday,2,arts,null))),'-'),1,6) Day2
,substr(nvl(to_char(max(decode(nday,3,arts,null))),'-'),1,6) Day3
,substr(nvl(to_char(max(decode(nday,4,arts,null))),'-'),1,6) Day4
,substr(nvl(to_char(max(decode(nday,5,arts,null))),'-'),1,6) Day5
,substr(nvl(to_char(max(decode(nday,6,arts,null))),'-'),1,6) Day6
,substr(nvl(to_char(max(decode(nday,7,arts,null))),'-'),1,6) Day7
,substr(nvl(to_char(max(decode(nday,8,arts,null))),'-'),1,6) Day8
,substr(nvl(to_char(max(decode(nday,9,arts,null))),'-'),1,6) Day9
,substr(nvl(to_char(max(decode(nday,10,arts,null))),'-'),1,6) Day10
,substr(nvl(to_char(max(decode(nday,11,arts,null))),'-'),1,6) Day11
,substr(nvl(to_char(max(decode(nday,12,arts,null))),'-'),1,6) Day12
from (select source_name, round(sysdate-sup_date) as nday, count(article_id) as arts
from articles_contents_xml
where round(sysdate-sup_date) <=14
group by source_name, round(sysdate-sup_date)
order by source_name, nday)
group by source_name
/

SOURCE_NAME   DAY1   DAY2   DAY3   DAY4   DAY5   DAY6   DAY7   DAY8   DAY9   DAY10  DAY11  DAY12
------------- ------ ------ ------ ------ ------ ------ ------ ------ ------ ------ ------ ------
Adresseavisen 17     -      38     35     29     33     -      43     -      53     -      63
Aftenposten   14     -      -      23     -      -      -      -      -      -      -      -
There are some fairly complex uses of both SQL*Plus and of SQL, so let us know if you have follow-on questions regarding the code, above, and if the code satisfactorily resolves your need.


[santa]Mufasa
(aka Dave of Sandy, Utah, USA)
[I provide low-cost, remote Database Administration services: www.dasages.com]
“Beware of those that seek to protect you from harm or risk. The cost will be your freedoms and your liberty.”
 
Thanks again!

one final small question. In the first example you have

,substr(nvl(to_char(max(decode(nday,1,arts,null))),'-'),1,4) Day1

while in the last example you have

,substr(nvl(to_char(max(decode(nday,1,arts,null))),'-'),1,6) Day1

Why 4 on the first and 6 on the second?

</Morten>
 
Morten said:
Why 4 on the first and 6 on the second?
In your first example, you wanted to accommodate column headings up to "Day6" (4 chrs.) in length. In your more recent specifications, you wanted "n-number of days", so I allowed for "Day999" (6 chrs.) of output. [wink]

[santa]Mufasa
(aka Dave of Sandy, Utah, USA)
[I provide low-cost, remote Database Administration services: www.dasages.com]
“Beware of those that seek to protect you from harm or risk. The cost will be your freedoms and your liberty.”
 
ah... got it.

Thanks again for your excellent help!

</Morten>
 
My pleasure ! Good luck, Morten.

[santa]Mufasa
(aka Dave of Sandy, Utah, USA)
[I provide low-cost, remote Database Administration services: www.dasages.com]
“Beware of those that seek to protect you from harm or risk. The cost will be your freedoms and your liberty.”
 
Status
Not open for further replies.

Part and Inventory Search

Sponsor

Back
Top