Sorting technique used by oracle for "order by" clause.

Hi All,
it could be very help to me if you provide some information about sorting technique used by oracle engine for order by clause.
Issue i am facing :
Table : xx
Line Date
1 05-06-2013 00:00:00
2 05-06-2013 00:00:00
when we query above table using order by date, it is returning line 2 prior to line 1. we would like to know why it is returning line 2 first?
Regards,
Ram

>
it could be very help to me if you provide some information about sorting technique used by oracle engine for order by clause.
>
Well ok - but be warned that many people wind up being sorry they ask that question. Hopefully Hemant's answer is what you really wanted.
See 'Linguistic Sorting and String Searching' in the Oracle® Database Globalization Support Guide
http://docs.oracle.com/cd/B28359_01/server.111/b28298/ch5lingsort.htm
Sorting will be controlled by the settings of NLS_LANGUAGE, NLS_SORT and NLS_COMP.
Here is the doc page for NLS_SORT
http://docs.oracle.com/cd/B28359_01/server.111/b28298/ch3globenv.htm#i1008393
>
NLS_SORT specifies the type of sort for character data. It overrides the default value that is derived from NLS_LANGUAGE.
NLS_SORT contains either of the following values:
NLS_SORT = BINARY | sort_name
BINARY specifies a binary sort. sort_name specifies a linguistic sort sequence.
>
And the one for NLS_COMP
http://docs.oracle.com/cd/B28359_01/server.111/b28298/ch3globenv.htm#i1008458
>
The value of NLS_COMP affects the comparison behavior of SQL operations.
You can use NLS_COMP to avoid the cumbersome process of using the NLSSORT function in SQL statements when you want to perform a linguistic comparison instead of a binary comparison. When NLS_COMP is set to LINGUISTIC, SQL operations perform a linguistic comparison based on the value of NLS_SORT. A setting of ANSI is for backward compatibility; in general, you should set NLS_COMP to LINGUISTIC when you want to perform a linguistic comparison.

Similar Messages

Using bind varaible for order by clause

Hello
Can some one suggest for the following scenario?
My order by clause will be constructed based on the selected fields in the form. And now the order by clause need to be passed as parameter to Excel report.
When I use Ref cursor and OPEN-FOR-USING clause, data is not sorted.
do we have any other alternate for this?
Please observe the example code for the same. Block below resembles the code for to select the data for excel report.
SQL> select * from t;
T
sdf
der
gdr
ghft
ytut
lkrt
rtrt
tyrt
SQL> declare
2 l_order_by VARCHAR2 (100);
3 l_test varchar2(10);
4 TYPE TEST IS REF CURSOR;
5 c_TEST test;
6 L_string VARCHAR2(2000):= 'select * from t order by :pi_order';
7 begin
8 l_order_by := ' t DEsc';
9 open c_test for l_string using l_order_By;
10 loop
11 fetch c_test into l_test;
12 exit when c_TEST%notfound;
13 dbms_output.put_line (l_test);
14 end loop;
15 close c_test;
16 end;
17 /
sdf
der
gdr
ghft
ytut
lkrt
rtrt
tyrt
PL/SQL procedure successfully completed.
Cheers
Ram Kanala

My order by clause will be constructed based on the selected fields in the formDoes this look like you need ?
SQL> var so number
SQL> exec :so := 1;
PL/SQL procedure successfully completed.
SQL> set serveroutput on
SQL> declare
2   rc sys_refcursor;
3   type emprec is table of emp%rowtype index by pls_integer;
4   erec emprec;
5 begin
6
7   open rc for 'select * from emp order by ' ||
8   'decode(:p,1,ename,2,deptno,3,sal,null),' ||
9   'decode(:p,4,ename,5,deptno,6,sal,null) desc' using :so,:so;
10   fetch rc bulk collect into erec;
11   close rc;
12
13   for i in 1..erec.count loop
14    dbms_output.put_line('Ename = ' || erec(i).ename || ', deptno = ' || erec(i).deptno || ', sal
= ' || erec(i).sal);
15   end loop;
16 end;
17 /
Ename = ADAMS, deptno = 20, sal = 1100
Ename = ALLEN, deptno = 30, sal = 1600
Ename = BLAKE, deptno = 30, sal = 2850
Ename = CLARK, deptno = 10, sal = 2450
Ename = FORD, deptno = 20, sal = 3000
Ename = JAMES, deptno = 30, sal = 950
Ename = JONES, deptno = 20, sal = 2975
Ename = KING, deptno = 10, sal = 5000
Ename = MARTIN, deptno = 30, sal = 1250
Ename = MILLER, deptno = 10, sal = 1300
Ename = SCOTT, deptno = 20, sal = 3000
Ename = SMITH, deptno = 20, sal = 800
Ename = TURNER, deptno = 30, sal = 1500
Ename = WARD, deptno = 30, sal = 1250
PL/SQL procedure successfully completed.
SQL> exec :so := 2
PL/SQL procedure successfully completed.
SQL> /
Ename = CLARK, deptno = 10, sal = 2450
Ename = KING, deptno = 10, sal = 5000
Ename = MILLER, deptno = 10, sal = 1300
Ename = JONES, deptno = 20, sal = 2975
Ename = FORD, deptno = 20, sal = 3000
Ename = ADAMS, deptno = 20, sal = 1100
Ename = SMITH, deptno = 20, sal = 800
Ename = SCOTT, deptno = 20, sal = 3000
Ename = WARD, deptno = 30, sal = 1250
Ename = TURNER, deptno = 30, sal = 1500
Ename = ALLEN, deptno = 30, sal = 1600
Ename = JAMES, deptno = 30, sal = 950
Ename = BLAKE, deptno = 30, sal = 2850
Ename = MARTIN, deptno = 30, sal = 1250
PL/SQL procedure successfully completed.
SQL> exec :so := 5;
PL/SQL procedure successfully completed.
SQL> /
Ename = BLAKE, deptno = 30, sal = 2850
Ename = TURNER, deptno = 30, sal = 1500
Ename = ALLEN, deptno = 30, sal = 1600
Ename = MARTIN, deptno = 30, sal = 1250
Ename = WARD, deptno = 30, sal = 1250
Ename = JAMES, deptno = 30, sal = 950
Ename = SCOTT, deptno = 20, sal = 3000
Ename = JONES, deptno = 20, sal = 2975
Ename = SMITH, deptno = 20, sal = 800
Ename = ADAMS, deptno = 20, sal = 1100
Ename = FORD, deptno = 20, sal = 3000
Ename = KING, deptno = 10, sal = 5000
Ename = MILLER, deptno = 10, sal = 1300
Ename = CLARK, deptno = 10, sal = 2450
PL/SQL procedure successfully completed.Rgds.

Can I use an expression in order by clause

SELECT
PERIOD_NUM, period_num *12,
period_name,
--TO_DATE(SUBSTR(PERIOD_NAME,1,3)| |SUBSTR(PERIOD_NAME,INSTR(PERIOD_NAME,'-'),3),'MON-RR'),
TO_NUMBER(TO_CHAR(TO_DATE((SUBSTR(PERIOD_NAME,1,3)| |SUBSTR(PERIOD_NAME,INSTR(PERIOD_NAME,'-')+1,2)),'MONRR'),'RRRRMM'))
FROM GL.GL_PERIODS
order by 2--TO_NUMBER(TO_CHAR(TO_DATE(SUBSTR(PERIOD_NAME,1,3)| |SUBSTR(PERIOD_NAME,INSTR(PERIOD_NAME,'-')+1,2),'MONRR'),'RRRRMM'))
desc
gl_periods is a table, period_name consists of values like oct-01_fy-01 etc and oct01_01_fy-01
I want to sort on this column based on period_name, so I am trying to convert into number like for oct-01_fy-01, the query gives
200110 and so on. the recent period comes first, But my question is I can't use the same to_number(.....) in the order by clause. oracle gives an error
ORA-01843: not a valid month.
Let me know whether anyone come across with this kind of situation

If you are sure the expression is correct, since you are using it in the select clause, give an alias and use the alias in the order by clause.
null

Using ROWID in the ORDER BY clause

Hi,
I can consider myself as a PL/SQL beginner. I need to query some existing tables created for audit scope.
These tables have more duplicate rows respect to the same audit date: the precision is at "hour and minute and second" level.
But I need to understand which is the more recent row for the same audit date. So I think to use in the SELECT statement
the clause ORDER BY in this manner:
SELECT ... FROM Audit_Table ORDER BY Audit_Date desc, ROWID desc.
I'd like to use this ORDER BY also for the ROW_NUMBER function.
Any suggests to me about this topic? Many thanks

I don't need to provide an example, I can do better by pointing you to the documentation:
http://docs.oracle.com/database/121/SQLRF/pseudocolumns008.htm#SQLRF00254
"If you delete a row, then Oracle may reassign its rowid to a new row inserted later" - for example
I can point you to asktom: https://asktom.oracle.com/pls/asktom/f?p=100:11:0::::P11_QUESTION_ID:912210644860
"ROWIDs are NOT sortable" - for example
Your example just happened to work, but you are using the rowid semantically incorrect.

Database Settings for order By Clause

Hi,
In oracle if we do a "order by" in SQL statement for field values whose values are empty, then they would be listed at the last if we dont specify any NVL() for it.
Is there any other option by which the funcionality of Order by clause with Empty field values can be changed.
Is there any settings that can be made in the existing database , so that the Order by clause does not list the empty field values at the bottom.
Thanx in advance.
Bharath

you can do wonders with decode and nvl
for instance:
order by decode(nullable_field, null, 1, 2)would make all your null fields come first, whereas
order by nvl(nullable_field, 'M')would make them appear in the middle of an alphabetical list.
Basically NULLs are not something. You have to give the ORDER BY something to replace the null.
null

Passing parameter in report for order by clause

how can we pass the name of the column on which we want the order by clause to sort as a parameter through parameter from in a report??

Hi Guptha,
We can create a bind parameter in report and we can send it as a parameter or enter in the parameter form.
The Query will be
select * from emp
order by :Order_Bu_Column
I think it will help you.
Regards,
Siva.

How to use XML Gateway for Order Import and Item Import?

Hello XML gurus,
I'm working in Oracle Apps area and do not have much knowledge on XML or web programming. Currently in my project there is a need to use Order Import and Item Import using interface tables and XML Gateway. The items/orders information will be in the XML format from Siebel system, there will be no custom processing...just process the data as-is.
I've read the XML Gateway user guide but did not understand much.
Anyone with similar implementation experience?
I'm looking for a step-by-step guide on what needs to be done in Oracle Apps side to populate the interface tables from XML file and call the import APIs?
Your inputs, suggestions, tips are highly appreciated. :)
Thx,
Jags

I have a similar requirement to automate the order import and was looking at XML Gateway. Please update if you were able to find more information.
- Ayyappa

Help for order by clause

Hi ,
i have a below table with data , when i run my select with order by option its gives me
1     Card
as a first row but i need this
12     atm
as a first row ,how this is possible .
create table test
(ID number ,
val varchar2(20)
insert into test values (1,'Card')
insert into test values (2,'Card')
insert into test values (3,'Card')
insert into test values (4,'Cash')
insert into test values (5,'Cash')
insert into test values (6,'Cash')
insert into test values (7,'Checque')
insert into test values (8,'Checque')
insert into test values (9,'Checque')
insert into test values (10,'atm')
insert into test values (11,'atm')
insert into test values (12,'atm')
select * from test
order by val asc

YOu can also use Nlssort like
SQL> select * from v$version;
BANNER
Oracle Database 10g Enterprise Edition Release 10.2.0.1.0 - Prod
PL/SQL Release 10.2.0.1.0 - Production
CORE    10.2.0.1.0      Production
TNS for 32-bit Windows: Version 10.2.0.1.0 - Production
NLSRTL Version 10.2.0.1.0 - Production
SQL> select *
2 from test
3 order by nlssort(val, 'NLS_SORT = WEST_EUROPEAN_AI') asc, id desc
4 /
        ID VAL
        12 atm
        11 atm
        10 atm
         3 Card
         2 Card
         1 Card
         6 Cash
         5 Cash
         4 Cash
         9 Checque
         8 Checque
        ID VAL
         7 Checque
12 rows selected.

Order By clause in Oracle

Hello,
Sorry I am posting it for the second time
In Oracle , the order by clause does not return the expected query result, if any of the field value in the order by clause has an empty string.
Oracle treats the empty string as null value and ORDER BY gives a result with the empty string field values listed at last.
For example :
test is a sample table containing "name" field.
Query: "select name from test order by name"
In SQL Server/Access
Result (1)
NAME
(blank)
(blank)
User1
User2
User3
In Oracle
Result (2)
NAME
User1
User2
User3
(blank)
(blank)
I know some of the Work arounds for this as listed below :
1) To use NVL in Order By Clause.
i.e., the modified query
"select name from test order by nvl(name,0)"
gives the result same as Result (1).
2) To have single blank space in the field value if it is empty.
I dont want to use either of these two options b'se it would lead to a mass change in my existing code.
Is there any way i can do a
collation order settings in the Oracle databases to get the results as in MS SQL/Access.
Or Is there any other way that i can set the
option once for this and need not worry abt it afterwards?
Can Any help me out in Solving this?
Thanks
Devi
null

For fun and out of curiosity, I tried :
SQL> select /*+ leading(t) use_nl(e t) */
2         e.deptno
3       , e.empno
4 from scott.emp e
5     , table(sys.odcinumberlist(30,10,20)) t
6 where e.deptno = t.column_value
7 ;
DEPTNO EMPNO
    30 7499
    30 7521
    30 7654
    30 7698
    30 7844
    30 7900
    10 7782
    10 7839
    10 7934
    20 7369
    20 7566
    20 7902
12 rows selected
It "seems" to work but of course we can't expect a consistent behaviour, or can we? ;)

Order By Clause with Empty Field values !

Hello,
In Oracle , the order by clause does not return the expected query result, if any of the field value in the order by clause has an empty string.
Oracle treats the empty string as null value and ORDER BY gives a result with the empty string field values listed at last.
For example :
test is a sample table containing "name" field.
Query: "select name from test order by name"
In SQL Server/Access
Result (1)
NAME
(blank)
(blank)
User1
User2
User3
In Oracle
Result (2)
NAME
User1
User2
User3
(blank)
(blank)
I know some of the Work arounds for this as listed below :
1) To use NVL in Order By Clause.
i.e., the modified query
"select name from test order by nvl(name,0)"
gives the result same as Result (1).
2) To have single blank space in the field value if it is empty.
I dont want to use either of these two options b'se it would lead to a mass change in my existing code.
Is there any way i can do a
collation order settings in the Oracle databases to get the results as in MS SQL/Access.
Can Any help me out in Solving this?
Thanks
Devi Shankar
null

Bharath,
I am moving this question to the SQL forum.
Regards,
Geoff

Materialized view and Order by clause

Hi all. I'd like to have some information regarding the order by clause used in a materialized view.
I'm using Oracle 9.2 and Win2003 server. I have a common view (my_common_view) on which a materialized view is based (my_materialized_view). Materialized view is built for fast refresh and is created in this way:
create materialized view my_materialized_view as
select * from my_common_view
order by 1,2,3;
My question is: if I query this materialized view without adding an order by clause in the statement, I will obtain always an ordered result set? So, is it useful to include the order by clause in the materialized view script or the order by has to be include in the statement used for quering the materialized view (for example like this: select * from my_materialized view order by ...)?
Thank you very much.
Ste.

SQL> create table t
2 (x int)
3 /
Table created.
SQL> create view vw
2 as
3 select *
4    from t
5   order by x
6 /
View created.
SQL> select *
2    from v$version
3 /
BANNER
Oracle9i Enterprise Edition Release 9.2.0.7.0 - 64bit Production
PL/SQL Release 9.2.0.7.0 - Production
CORE    9.2.0.7.0       Production
TNS for IBM/AIX RISC System/6000: Version 9.2.0.7.0 - Production
NLSRTL Version 9.2.0.7.0 - Production
SQL>

Sys_context and Order by clause

Hi All,
Im using Oracle 11g R2.
Is it somehow possible to use Context variable (to store column name) in ORDER BY clause? My ultimate aim is to construct Order by at runtime without using concatenation.
e.g. i created context as my_ctx for some pkg and its variable as v_ctx having value set as "TABLE_NAME"
so is it possible to use it as
select * from all_tables
order by sys_context('my_ctx', 'v_ctx');
if no, is there any work around??
Thanks,
Vivek

Hi, Vivek,
Vivek wrote:
Hi All,
Im using Oracle 11g R2.
Is it somehow possible to use Context variable (to store column name) in ORDER BY clause? My ultimate aim is to construct Order by at runtime without using concatenation.Sure, you can use SYS_CONTEXT in an ORDER BY clause, but it's unclear what you're trying to do, so I don't know if SYS_CONTEXT will help you.
e.g. i created context as my_ctx for some pkg and its variable as v_ctx having value set as "TABLE_NAME"
so is it possible to use it as
select * from all_tables
order by sys_context('my_ctx', 'v_ctx');
if no, is there any work around??Post a complete test script hat people can run to re-create the problem and test their ideas. Show the output you want from the parameters an data you give.
Think about using a CASE expression in the ORDER BY clause. You can use SYS_CONTEXT in the CASE expression if you want to.

Default where with order by clause

Hi all
How to use Default where with order by clause
for example i want to use Default_where clause with order by clause
Regards
Shahzaib ismail

Hi,
I have a similar query for where clause.
my query-find form is a multi record form like below:
Criteria - Condition - Value
Item/Category/Planner (List_Item) - Equals/Among (List_Item) - text_FIELD
Item - Equals - ITEM_1
Category - Among - ('CAT1', 'CAT2')
Planner - Equals - PL1
Find_BUTTON Clear_BUTTON
User can select any criteria and condition combination and then enter the value in "Value" text field.
My query is: how can I prepare a where clause based on user input since the user can enter any number of values/combinations? I guess I would have to use some looping to generate where clause.
Thanks
Imran

Group by and order by clause

Hi,
I have wriiten the below query and i am using group by and order by clause....but i am not getting the required result
SELECT b.ACCOUNT_REGION,CASE WHEN b.product_reference_status='Reference' THEN 'Recruited' WHEN b.product_reference_status in ('Inactive','Declined') THEN 'Inactive' WHEN b.product_reference_status ='Nominate' THEN 'Inprogress' ELSE 'Other' END product_reference_status, COUNT(decode(b.product_quarter,'Q1FY09',b.REFERENCE_ID)) Q1FY09, COUNT(decode(b.product_quarter,'Q2FY09',b.REFERENCE_ID)) Q2FY09, COUNT(decode(b.product_quarter,'Q3FY09',b.REFERENCE_ID)) Q3FY09, COUNT(decode(b.product_quarter,'Q4FY09',b.REFERENCE_ID)) Q4FY09, COUNT(decode(b.product_quarter,'Q1FY10',b.REFERENCE_ID)) Q1FY10, COUNT(decode(b.product_quarter,'Q2FY10',b.REFERENCE_ID)) Q2FY10, COUNT(decode(b.product_quarter,'Q3FY10',b.REFERENCE_ID)) Q3FY10, COUNT(decode(b.product_quarter,'Q4FY10',b.REFERENCE_ID)) Q4FY10, COUNT(b.product_quarter) Total
FROM refdump a, ref_dh_pr b
WHERE A.REFERENCE_ID=b.REFERENCE_ID AND (b.CREATED_ON - NVL(a.REFERENCE_DATE,a.CREATED))>15 group by b.account_region,product_reference_status order by 2
output is as given below
Region Status Count
LAD     Inactive     0
EMEA     Inactive     21
*APAC     Inactive     2
*APAC     Inactive     1
EMEA     Inactive     2
EMEA     Inprogress     220
LAD     Inprogress     19
LAD     Other     2
LAD     Other     0
LAD     Other     5
LAD     Recruited     182
APAC     Recruited     191
My question is
1) Why i am geeting two APAC regions and 3 LAD (marked *) regions when i have grouped by region and status
2) How can i make the status in the following order
a) Recruited
b)Inprogress
c)Inactive
d)Other
Thanks,

<i>group by b.account_region,product_reference_status</i>
Just take all the columns in the group by clause in your select segment and run the SQL, you'll come to know abt the different values in columns which group together.
*009*

Default sorting order used by oracle

Can anyone please let me know whether oracle uses any default sort order or not when executing an sql which does not have any sort order/ order by clause?
I have a sql query where no sort order is specified.
Now if I add some more columns in the select clause and execute the same query, the number of records returned are same and also the records are same.
But the order in which they are displayed is different from the previous query.
Could anyone please let me know whether oracle uses any default sort order or not when executing an sql which does not have any sort order/ order by clause
or does it depend on the columns of the table?
Thanks in advance.

To answer your direct question "why the order in the previous example differ".
When you added additional columns to the select list, that changed the query plan generated for the query. This caused Oracle to pick up the rows in a different order. There are a number of reasons this could happen, the new selected columns caused a different index (or no index) to be used for one of the tables in the query, it caused a different table to be be used as the "driving" table (i.e. the first one accessed), it caused a different join method to be used.
HTH
John

Sorting technique used by oracle for "order by" clause.

Similar Messages

Maybe you are looking for