How to build the query SQL for indexes work based on data - dict.?

Hello

I want to build the index 'create' - statement to all indexes have worked-based on the data dictionary.

For example:
create indexes tab1.ind1
on tab1 (lower (col3), lower (col1), col2, col4)
Tablespace "index."

I know I'll find all needed information views all_indexes, all_ind_columns and all_ind_expressions. During the generation of the create statement for the index of the order of the indexed columns must be correct (the order of the indexed column can be achieved using the "position_colonne" column of seen all_ind_columns and all_ind_expressions). But what would be the best way to provide all the necessary information in the right order?
One way would be a PLSQL procedure that puts it all together - but is it possible to use a single SQL statement?

Any help will be appreciated.

Rgds
JH

You can try using DBMS_METADATA. GET_DDL:

SQL> show user;
USER is "TEST"
SQL> drop table t;

Table dropped.

SQL> create table t(x varchar2(10));

Table created.

SQL> create index i on t(upper(x));

Index created.

SQL> select dbms_metadata.get_ddl('INDEX','I','TEST') from dual;

DBMS_METADATA.GET_DDL('INDEX','I','TEST')
--------------------------------------------------------------------------------

  CREATE INDEX "TEST"."I" ON "TEST"."T" (UPPER("X"))
  PCTFREE 10 INITRANS 2 MAXTRANS 255
  STORAGE(INITIAL 65536 NEXT 1048576 MINEXTENTS 1 MAXEXTENTS 2147483645
  PCTINCREASE 0 FREELISTS 1 FREELIST GROUPS 1 BUFFER_POOL DEFAULT)
  TABLESPACE "USERS"

Tags: Database

Similar Questions

write the query sql for the following situation?
We have a table like format following...

A AND B
...................
10 30
50 40
60 80
90 100

so to write the query and get the result like formats following...

RESULT
............
30
50
80
100

Try this... He's one of the interview question. I can't write the query...

If a knowledge write the query...
Hello
Try this
```
select (case when a>b then a else b end)   "Result"  from ;
```

How to pass the query parameters for the task "execute the external command"?

Hello
We are on the VSM 9 R8. Although VSM offers plenty of connectors, however we must incorporate VSM with other 3rd party applications using powershell, protocols SOAP.
I do this via the task run the external procedure. However, I am only able to send standard variables outside the tasks for example [request: REQUEST_DESC]. This task can not pass my custom settings which I take on my portal to the user for example, currency (Datacae id is CURR).
All the world is facing this limitation before?
Concerning
Ankit

It's all exactley. Although just with the help of [EXTFLDV_500xxx] would give the value of the field of the task at hand. You could use [REQUEST: EXTFLDV_500xxx]. I prefer [REQUEST_NO$ CR_REQUEST.] EXTFLDV_500xxx]

Build the query to select date with status

Hi guys,.
Grateful if you can advise me on how to build the query on below scenario:
Table A
Date
11.44.39.000000000 12-OCT-14 AM ASIA/SINGAPORE
11.44.35.000000000 16-SEP-14 AM ASIA/SINGAPORE
11.44.42.000000000 22-SEP-14 AM ASIA/SINGAPORE
The result of the query:
Date Status
11.44.39.000000000 12-OCT-14 AM ASIA / SINGAPORE inactive
11.44.35.000000000 16-SEP-14 active AM ASIA/SINGAPORE
11.44.42.000000000 22-SEP-14 AM ASIA/SINGAPORE inactive
Basically, the logic, I wanted is
Active principles: today date or max (date) < sysdate
Otherwise will be inactive
And there is 1 active date in time.
Thanks in advance

Hello

2753165 wrote:

Hi guys,.

Grateful if you can advise me on how to build the query on below scenario:

Table A

Date

11.44.39.000000000 12-OCT-14 AM ASIA/SINGAPORE

11.44.35.000000000 16-SEP-14 AM ASIA/SINGAPORE

11.44.42.000000000 22-SEP-14 AM ASIA/SINGAPORE

The result of the query:

Date Status

11.44.39.000000000 12-OCT-14 AM ASIA / SINGAPORE inactive

11.44.35.000000000 16-SEP-14 active AM ASIA/SINGAPORE

11.44.42.000000000 22-SEP-14 AM ASIA/SINGAPORE inactive

Basically, the logic, I wanted is

Active principles: today date or max (date)<>

Otherwise will be inactive

And there is 1 active date in time.

Thanks in advance

Sorry, we don't know what you want.

Are you saying that 1 row (maximum) can be assigned status = 'Active' and if 2 or more lines are eligible, the last of them will be called 'Active'?

If so:

CASE

WHEN tmstmp<=>

OR ROW_NUMBER () OVER (PARTITION BY CASE

WHEN tmstmp<=>

THEN "could be."

ANOTHER 'No Way'

END

ORDER BY tmstmp DESC

) = 1

THEN "active."

ELSE 'inactive '.

END

If you would care to post CREATE TABLE and instructions INSERT for some samples, so I could test it.

DATE is not a very good name for a column, especially if the column is a TIMESTAMP, not a DATE. I called the TMSTMP instead of the DATE column.

Web site created by Muse. The greats until tweek of today. Web site insists now for open index.html in the tablet. Don't know how product, nor how to change the default value for pc of office/index.html. Would advise. Bob

Web site created by Muse. The greats until tweek of today. Web site insists now for open index.html in the tablet. Don't know how product, nor how to change the default value for pc of office/index.html. Would advise. Bob

Hi Bob

I think that you have found the solution but if still you are facing the question, please give me the url of the site that opens another provision in the tablet.

Please make sure desktop redirection is checked in the site properties page layout.

Thank you

Sanjit

procedure that will dynamically build the query data and table Medallion

Hi people,

I write a procedure that dynamically build the query data and insert in the table "dq_summary".
enforcement procedure with success and data not inserted into the table 'dq_summary '.

I have thin problem in code attached below
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

FOR rep IN cur_di_attr
      LOOP
        dbms_output.put_line ('d: ');   
        
        BEGIN
          EXECUTE IMMEDIATE 'SELECT table_name FROM ' || sum_tab || ' WHERE id = ' || rep.attribute_id INTO rep_tab;
          dbms_output.put_line ('rep_tab: '||rep_tab);
          run_query := run_query || ' ' || rep_tab || ' WHERE ' || nvl(wh_cond, '1 = 1');
          EXECUTE IMMEDIATE run_query INTO end_rslt;
        
          EXECUTE IMMEDIATE 'UPDATE dq_summary SET ' || prop || '_' || p_code || ' = ' || end_rslt || ' WHERE attribute_id = ' || rep.attribute_id;
          dbms_output.put_line ('e: ');      
          dbms_output.put_line ('rep_tab: '||rep_tab);
          dbms_output.put_line ('end_rslt: '||end_rslt);
          dbms_output.put_line ('f: '); 
        EXCEPTION
          WHEN no_data_found THEN
            rep_tab := '';
            sum_tab := '';
        END;  
      
      END LOOP;

-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
but in the procedure below must be run several times

create or replace
PROCEDURE DQ_REPORT_PROC
AS
  prop                              di_proposition.pro_name%type;
  col_var                           VARCHAR2(100);
  p_code                            dq_parameter.para_code%type;
  sum_tab                           di_proposition.summary_table%type;
  run_query                         dq_parameter.run_query%type;
  wh_cond                           dq_parameter.where_cond%type;
  end_rslt                          VARCHAR2(20);
  rep_tab                           VARCHAR2(50);
  v_error_msg                       VARCHAR2(200);   
  v_error_code                      VARCHAR2(200);  
  v_object_name                     VARCHAR2(50)                          DEFAULT 'DQ_REPORT_PROC';
  v_iss_no                          VARCHAR2(20)                          DEFAULT NULL;
  CURSOR cur_di_prop IS 
    SELECT upper(replace(replace(pro_name, ' '),'-')) pro_name
      FROM di_proposition;
  
  CURSOR cur_di_para IS
    SELECT upper(para_code) para_code, run_query, where_cond
      FROM dq_parameter;
  
  CURSOR cur_di_attr IS 
    SELECT attribute_id
      FROM dq_summary;
BEGIN
  
  DELETE FROM dq_summary;

  INSERT INTO dq_summary (attribute_id, entity_name, attribute_name, data_champ) 
    SELECT a.attribute_id, b.entity_name, a.attribute_name, a.data_champ
      FROM di_attribute_master a, di_entity_master b
     WHERE a.entity_id = b.entity_id;

  FOR c_prop IN cur_di_prop
  LOOP
    prop := c_prop.pro_name;
    
    BEGIN
      SELECT distinct SUBSTR(column_name, 1, INSTR(column_name, '_')-1), summary_table
        INTO col_var, sum_tab
        FROM user_tab_cols a, di_proposition b
       WHERE a.table_name = 'DQ_SUMMARY'
         AND upper(replace(replace(b.pro_name, ' '),'-')) = prop
         AND SUBSTR(a.column_name, 1, INSTR(a.column_name, '_')-1) = upper(replace(replace(b.pro_name, ' '),'-'))
         AND upper(b.status) = 'Y';
         
         dbms_output.put_line ('col_var: '||col_var);
         dbms_output.put_line ('sum_tab: '||sum_tab);
         
    EXCEPTION
      WHEN no_data_found THEN
        col_var := '';
        sum_tab := '';
    END;

    dbms_output.put_line ('a: ');

    FOR para IN cur_di_para
    LOOP
     dbms_output.put_line ('b: ');
      p_code := para.para_code;
      run_query := para.run_query;
      wh_cond := para.where_cond;
      dbms_output.put_line ('c: ');
      FOR rep IN cur_di_attr
      LOOP
        dbms_output.put_line ('d: ');   
        
        BEGIN
          EXECUTE IMMEDIATE 'SELECT table_name FROM ' || sum_tab || ' WHERE id = ' || rep.attribute_id INTO rep_tab;
          dbms_output.put_line ('rep_tab: '||rep_tab);
          run_query := run_query || ' ' || rep_tab || ' WHERE ' || nvl(wh_cond, '1 = 1');
          EXECUTE IMMEDIATE run_query INTO end_rslt;
        
          EXECUTE IMMEDIATE 'UPDATE dq_summary SET ' || prop || '_' || p_code || ' = ' || end_rslt || ' WHERE attribute_id = ' || rep.attribute_id;
          dbms_output.put_line ('e: ');      
          dbms_output.put_line ('rep_tab: '||rep_tab);
          dbms_output.put_line ('end_rslt: '||end_rslt);
          dbms_output.put_line ('f: '); 
        EXCEPTION
          WHEN no_data_found THEN
            rep_tab := '';
            sum_tab := '';
        END;  
      
      END LOOP;    
    END LOOP;
  END LOOP; 
  COMMIT;   
EXCEPTION
      WHEN OTHERS THEN
         v_error_msg   := SQLERRM;
         v_error_code  := SQLCODE;  
         TRACKER_LOG_EXECEPTION(v_iss_no, v_object_name, CURRENT_TIMESTAMP, v_error_msg, v_error_code);
      COMMIT;        
  
END DQ_REPORT_PROC;

Published by: BluShadow on February 7, 2012 12:04
addition of {noformat}

{noformat} tags.  Please read {message:id=9360002} and learn to do this yourself in future.

903830 wrote:

I write a procedure that dynamically build the query data and insert in the table "dq_summary".
enforcement procedure with success and data not inserted into the table 'dq_summary '.

I'm sorry. But there is no kind of say that way. The code is undesirable. The approach is wrong. This will not happen. This will cause the fragmentation of memory in the shared Pool. This will lead to another session being impossible to analyze the sliders because of the fragmented memory.

Not only that. The underlying data model is questionable.

All this seems a candidate perfect as an example of how NOT to design and code and use Oracle.

Need help to build the query

Hello

I got the info from a table named X.

Col1 Col2 Col3 Col4
IT01 20 01 X
IT01 20 50 X
IT01 30 01 X
IT01 30 50 X

If the value of "X" all 4 combinations of Col2 and Col3, Col4 the result should be displayed as

Col1, Col4
IT01 X

If the actual values are as
Col1 Col2 Col3 Col4
IT01 20 01 X
IT01 20 50 X
IT01 30 01 X
30 50 IT01 Y

The result should be displayed as

Col1, Col4
IT01 Y

How will generate the query for this?

Thank you
Thomas Titus

Published by: Titus Thomas on April 20, 2010 14:48

Try this:

with test_tab as (select 'ITO1' Col1, 20 Col2, 01 Col3, 'X' Col4 from dual union all
                  select 'ITO1' Col1, 20 Col2, 50 Col3, 'X' Col4 from dual union all
                  select 'ITO1' Col1, 30 Col2, 01 Col3, 'X' Col4 from dual union all
                  select 'ITO1' Col1, 30 Col2, 50 Col3, 'X' Col4 from dual union all
                  select 'ITO2' Col1, 20 Col2, 01 Col3, 'X' Col4 from dual union all
                  select 'ITO2' Col1, 20 Col2, 50 Col3, 'X' Col4 from dual union all
                  select 'ITO2' Col1, 30 Col2, 01 Col3, 'X' Col4 from dual union all
                  select 'ITO2' Col1, 30 Col2, 50 Col3, 'Y' Col4 from dual)
--- end of mimicking your data; use SQL below:
select col1, max(col4) col4
from   test_tab
group by col1;

COL1 COL4
---- ----
ITO2 Y
ITO1 X

How to get the device model, for example app works on PlayBoook or BlackBerry 10?

How to get the device model, for example app works on PlayBoook or BlackBerry 10?

You can use this class: http://goo.gl/GtMLP for information on devices

something like this:
```
String myDeviceModel = android.os.Build.MODEL;
String myDeviceBrand = android.os.Build.BRAND;
String myDeviceDevice = android.os.Build.DEVICE;
```

How to get the different records for the date max.

Hi all
Here is the sample sql for sample table and data.

Create table student (dept_id number(10), first_name varchar2(10),last_name varchar2(10),join_date date,years_attended number(10)); insert into student values (1,'Ann','Coleman',to_date('3/7/1917','MM/DD/YYYY'),4); insert into student values (1,'Ann','Coleman',to_date('3/7/1916','MM/DD/YYYY'),5); insert into student values (2,'Rock','Star',to_date('1/1/1920','MM/DD/YYYY'),5); insert into student values (2,'Rock','Star',to_date('1/1/1921','MM/DD/YYYY'),6); insert into student values (3,'Jack','Smith',to_date('7/1/1900','MM/DD/YYYY'),3);

insert into student values (3,'Jack','Smith',to_date('7/1/1901','MM/DD/YYYY'),4);

commit;

I need to get maximum date records when the name and dep_id corresponds to. I wrote the query below and it becomes the expected result, but I'm not sure it's quite effective.

SELECT s.dept_id, s.first_name,s.years_attended

FROM (SELECT dept, MAX (join_date) join_date

STUDENT GROUP BY dept_id) x

Student JOIN s ON x.dept_id = s.dept_id AND x.join_date = s.join_date;

This above query returns records like below, and this is the goal.

DEPT_ID NAME YEARS_ATTENDED

1                         Ann                                4

2                         Rock                              6

3                         Jack                              4

Can you please let me know the query SQL I wrote is effective or not? This sample table may have less data, but I'm dealing with millions of records.

Hello

Thanks for posting CREATE TABLE and INSERT statement. This really helps.

Here's a solution. I also added a name that seems logical. In which case you can delete:

Select dept_id, first_name, last_name

, max (years_attended) Dungeon years_attended (last dense_rank order by join_date)

the student

Group of dept_id, first_name, last_name;

DEPT_ID FIRST_NAME LAST_NAME YEARS_ATTENDED

---------- ---------- ---------- --------------

1 Ann Coleman 4

Rock 2 star 6

3 Jack Smith 4

Kind regards.

Alberto

How to provide the Alias name for BOX Stmt?

Hello
How to provide the Alias name for the instruction BOX?
--
Thank you and best regards,
Badr Hari

1007980 wrote:

Hello

How to provide the Alias name for the instruction BOX?

--

Thank you best regards &,.

Badr Hari

You post any instructions BOX or a query.

You just add an alias after him, as you do for any other column you select.

Select case when deptno = 20, then 'Yes' else 'no' from the end IsItDept20 dept

ISITDEPT20

NO.

Yes

NO.

NO.

How to frame the query?

Uses oracle ebs r12

I have a table of data is

Operation_seq Type statusflag SeqNum
1 10A Y
2 20 b Y
3 30 c Y
4 40 D Y
5 50 e N
6 60 f Y

I want to write a query in which I want to take only 1 record operation_seq = 50 and Statusflag is N and all other interventions seq status is there.

My query must return only 50 operation when the status of 50 is N and any other status is Y.

Help, please.

Thank you
Lavan

LAVANKV wrote:
This statusflag can be (null or N), then how to frame the query.

Simply wrap the statusflag with NVL column, you should be good.

SQL> with t
  2  as
  3  (
  4  select 1 seqnum, 10 operation_seq, 'a' type, 'Y' statusflag from dual union all
  5  select 2 seqnum, 20 operation_seq, 'b' type, 'Y' statusflag from dual union all
  6  select 3 seqnum, 30 operation_seq, 'c' type, 'Y' statusflag from dual union all
  7  select 4 seqnum, 40 operation_seq, 'd' type, 'Y' statusflag from dual union all
  8  select 5 seqnum, 50 operation_seq, 'e' type, '' statusflag from dual union all
  9  select 6 seqnum, 60 operation_seq, 'f' type, 'Y' statusflag from dual
 10  )
 11  select seqnum, operation_seq, type, statusflag
 12    from (
 13            select t.*, count(decode(statusflag, 'Y', 1, null)) over() cnt_1, count(*) over() cnt_2
 14              from t
 15         )
 16   where operation_seq = 50
 17     and nvl(statusflag, 'N') = 'N'
 18     and cnt_2-cnt_1 = 1
 19  /

    SEQNUM OPERATION_SEQ T S
---------- ------------- - -
         5            50 e

How to write the query option in expdp
Hi Please someone help me how to write the query option in expdp... .in expdp using the query option...

where AM columnname between 5 May 12 02:57:00.000' and ' 02:59:59.999 6 May 12: ';

Please do what is necessary...
Pavan Kumar says:
QUERY = (columnname scott.test: "where between 5 May 12 02:57:00.000 h ' and ' 6 May 12 AM 02:59:59.999'")

Who will fail in databases, because you assume nls_date_format. How it is difficult to put to_date() surround channels? Rather than play with quotation marks, using one parfile thusly.
```
query=table_owner.table_1:"where business_date between to_date('20120505025700 am','yyyymmddhhmiss am') and to_date('20120505025959 am','yyyymmddhhmiss am')"
query=table_owner.table_2:"where business_date between to_date('20120505025700 am','yyyymmddhhmiss am') and to_date('20120505025959 am','yyyymmddhhmiss am')"
query=table_owner.table_3:"where business_date between to_date('20120505025700 am','yyyymmddhhmiss am') and to_date('20120505025959 am','yyyymmddhhmiss am')"
```
You do not have to have all the clauses in a single line, as they are side by side parfile, which would be enough. For this reason parfile is better than the command line in order to avoid all the back-citing dance.

How to change the PCTFREE parameter for an existing table
How to change the PCTFREE parameter for an existing table?

Guys can you help me please
Hello

Of course it is possible.
In this case you would export the table from the database (using old style exp)
Start the import with the index_file option to generate the table/index instructions
Modify the file created and updated the PCTFREE and remove any that need to be performed (default, if I remember correctly, he comments on the statements of the Index)
Run the file for the table and the index created
After that start the import with the option "ignore = yes".

Make sure that you export only this particular table using "tables =
Also in importing the "ignore = yes" will also be important data that still exists.

In my view, however, this is not what you want to do. It's more complex, sensitive failure and the table should be deleted, so users are affected unnecessarily long.

Success!
FJFranken

How to end a query running for a report?
Hello
When I run a report, the SQL query takes too long.
This causes an overload on the database.
How to stop the query execution after a while?
It is NOT possible in OBIEE 10 g.
In OBIEE 11 g, when a scheduled report task is working you can now cancel the currently running task since the report job history page.
This section 6.9, ['Cancellation of a work in progress.' | http://docs.oracle.com/cd/E23943_01/bi.1111/e22257/view_manage_rpt_his.htm#cancel_runj]
And only possible since 11.1.1.5

concerning
Jorge
p.s if this answers your question, then please mark my answer 'Proper' or 'useful '.

How to find the bpel instance in 11 g based on the index values
We have 10 GB BPEL processes where we define the 4 index values for all instances. Whenever the support request comes, ask us the values of index and based on what we're looking for the process instance.

We have now migrated bpel 10g and 11g process. How to find the bpel instance in 11 g based on the index values?
You can go there. http://soasphere.blogspot.com/2011/04/how-to-Serch-composites-for-index.html. Do a join on the cube_instance and ci_indexes tables.

How to build the query SQL for indexes work based on data - dict.?

Similar Questions

Web site created by Muse. The greats until tweek of today. Web site insists now for open index.html in the tablet. Don't know how product, nor how to change the default value for pc of office/index.html. Would advise. Bob

Maybe you are looking for