Stmts SELECT with a Having Clause

Similar Questions

• Query with the having clause

  Hi all

  I have this select statement to help to filter different from zero records.

  Select GroupName, Nom_liste,

  Sum ((a_cost/1000) * degradation * quantity * total_count * index) such as total_cost,.

  FROM MyTable

  having sum ((a_cost/1000) * degradation * quantity * total_count * index) <>0

  Group GroupName, Nom_liste

  Having done so the query to run more slowly.

  How can I filter records with zero total_cost (they can be negative or greater than zero).

  Thank you!

  Hello

  user9542267 wrote:

  Hi all

  I have this select statement to help to filter different from zero records.

  Select GroupName, Nom_liste,

  Sum ((a_cost/1000) * degradation * quantity * total_count * index) such as total_cost,.

  FROM MyTable

  having sum ((a_cost/1000) * degradation * quantity * total_count * index) <> 0

  Group GroupName, Nom_liste

  Having done so the query to run more slowly.

  How can I filter records with zero total_cost (they can be negative or greater than zero).

  Thank you!

  The query you posted should not take more time with or without the HAVING clause.  The right of the comma before 'from' causes a compilation error very quickly anyway.

  If you get no result other than an error message, then this is not the code you are running.  Show the code you actually run, or a simplified version that has the same problem.

  For all performance issues, see the FAQ forum:

  https://community.Oracle.com/message/9362003#9362003

  If the problem was really in the HAVING clause, then you can try something like this:

  WITH got_total_cost AS
  (
  SELECT GroupName, Nom_liste
  , SUM ((a_cost/1000) * degradation * quantity * total_count * index) SUCH as total_cost
  FROM MyTable
  GROUP BY GroupName, Nom_liste
  )
  SELECT *.
  OF got_total_cost
  WHERE as total_cost <> 0
  ;

  but it would be very surprising he made all the difference.  I suspect that the problem is not really in the HAVING clause.

• try to get the SUM of all of the charges with the having clause

  Hi Oracle users.

  I'm trying to accomplish to get a sum of a series of charges and it should be easy, but I can't get it.

  For example, I have the following.

  Select ACCTLOC,  count(ACCTLOC) AS TOTALP
  FROM BOX_ACCTS_TBL 
   WHERE TIME_REQ > SYSDATE-120
   Group By ACCTLOC
    having (count(ACCTLOC) > 1)
   
   the results are
   
  ACCTLOC   TOTALP
   
  Philly     15970
  NY          8623
  Tacoma        3
  SanFran     195
  Hartford    5
  Miami       4374
  
    
  

  How can I get the SUM of this group? Do I put this in PL SQL and do a procedure to carry out the selection? Is this the way to go on this?
  I appreciate all the comments you can give.

  Thank you!

  with

  query_result as

  (select 'Philly' acctloc, 15970 totalp Union double all the)

  Select "NY", 8623 Union double all the

  Select "Tacoma", 3 double Union all

  Select "SanFran", 195 double Union all

  Select "Hartford", 5 Union double all the

  Select 'Miami', double 4374

  )

  Select acctloc, totalp, sum (totalp) on the_sum, avg (totalp) (on the_avg), ratio_to_report (totalp) (percentages))

  of query_result

  ACCTLOC	TOTALP	THE_SUM	THE_AVG	PERCENTAGES
  Philly	15970	29170	4861.66666666666666666666666666666666667	.547480287967089475488515598217346588961
  NY	8623	29170	4861.66666666666666666666666666666666667	.295611930065135413095646211861501542681
  Tacoma	3	29170	4861.66666666666666666666666666666666667	.000102845389098388755570791909496057593
  SanFran	195	29170	4861.66666666666666666666666666666666667	.006684950291395269112101474117243743572
  Hartford	5	29170	4861.66666666666666666666666666666666667	.000171408981830647925951319849160095989
  Miami	4374	29170	4861.66666666666666666666666666666666667	.149948577305450805622214604045251971203

  Concerning

  Etbin

• Cannot wrap a select with order by clause in a file.

  TOGETHER SET TERMOUT OFF FEEDBACK OFF PAGESIZE 0 400 LINESIZE ECHO OFF OFF POSITION

  coil & 1;

  Select "EMPID, EMP_NAME, REGION, FROM, TO, MONTH, ACTIVITY_TYPE, PROJECT_PIN, MILESTONE_NAME, PRACTICE_ID, RESOURCE_CATEGORY, HOURS, COMMENT, percent ALLOCATION, INVOICE_TO, the STATE of the double"

  Union of all the

  select empid,resource_name,region,week_of,(week_of+7),core_month,bill_rate,project_name,project_type,milestone_name,practice_id,resource_category,total_hours_submitted,project_type,"%ALLOCATION",invoice_to,status from master_emp_allocation order by week_of;

  spool off;

  "exit";

  /

  the above block gives result below

  select empid,resource_name,region,week_of,(week_of+7),core_month,bill_rate,project_name,project_type,milestone_name,practice_id,resource_category,total_hours_submitted,project_type,"%ALLOCATION",invoice_to,status from master_emp_allocation order by week_of

  *

  ERROR at line 3:

  ORA-00904: "WEEK_OF": invalid identifier

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

  Without the order by clause it wraps the file passed as an argument.

  check

  Combining the UNION AGENDA and BY

  Columns in the ORDER BY list must be a subset of the columns in the selection list on the left side of the union.

  All the columns in the ORDER BY list should be sorted in ascending order, and they must be a prefix in the order of the columns in the target of the left side of the UNION list.

• Using the HAVING clause

  I have trouble understanding the following query:

  SELECT cust_city, COUNT (cust_last_name)

  Customers

  WHERE cust_credit_limit > 1000

  GROUP BY cust_city

  HAVING AVG (cust_credit_limit) BETWEEN 5000 AND 6000

  According to my preparation for the review, it will work successfully without errors, although I don't understand why.

  The AVG (cust_credit_limit) should also be included in with the rest of the fields to make it work correctly?

  It would be greatly appreciated if someone could explain it to me better.

  Thank you

  Sean

  5285cd35-2ff3-432e-B2FD-19285481be60 wrote:

  I have trouble understanding the following query:

  SELECT cust_city, COUNT (cust_last_name)

  Customers

  WHERE cust_credit_limit > 1000

  GROUP BY cust_city

  HAVING AVG (cust_credit_limit) BETWEEN 5000 AND 6000

  According to my preparation for the review, it will work successfully without errors, although I don't understand why.

  The AVG (cust_credit_limit) should also be included in with the rest of the fields to make it work correctly?

  It would be greatly appreciated if someone could explain it to me better.

  Thank you

  Sean

  No - the AVG function doesn't have to be in the select list. Explain why you think it should be.

  How would it make a difference if you added to the SELECTION list and ignored, then?

  When you have questions DOC the first thing you need to do is RTFM. See the HAVING clause in the RFSO of the SQL language

  http://docs.Oracle.com/CD/B28359_01/server.111/b28286/statements_10002.htm

  >

  HAVING clause

  Use of the HAVING clause to restrict groups of returns of the lines to these groups for which the specified condition is TRUE .

  >

  As the doc t says

  He HAVING condition applies to the 'groups '.

  If you use the HR. EMPLOYEES table you can see the effect of the HAVING clause. This query is similar to yours but with none HAVING clause, and the AVG (salary) added to the selection list to see which groups will have

  >

  Select department_id, count (last_name), avg (salary)
  employees
  where salary > 500
  Group by department_id

  1. DEPARTMENT_ID, COUNT (LAST_NAME), AVG (SALARY)
  2. 10,1,4400
  3. 20,2,9500
  4. 30,6,4150
  5. 40,1,6500
  6. 50,45,3475.55555555556
  7. 60,5,5760
  8. 70,1,10000
  9. 80,34,8955.88235294118
  10. 90,3,19333.3333333333
  11. 100,6,8601.33333333333
  12. 110,2,10154
  13. 1,7000

  >

  Note that 60 is the ONLY group with AVG (salary) between 5000 and 6000

  Now, try the query using the HAVING clause

  >

  Select department_id, count (last_name)

  employees

  where salary > 500

  Group by department_id

  having avg (salary) between 5000 and 6000

  DEPARTMENT_ID, COUNT (LAST_NAME)

  60.5

  >

  Only group 60 has been selected - the HAVING condition has been applied to the first query GROUPS, not the lines.

• Adding a HAVING clause in the BI tools

  Hello

  I have to build a BI reporting tool that counts students participating in academic programs.

  I am limited to display data personal due to the regulation of the right, so I do the classes on a higher leverage. (Example: the size of the location has records of the regions instead of the city).

  But the sill, I can't guarantee that the report will not show only one or two records for a combination of dimension values.

  Software Oracle BI (as Oracle BI) has the functionality to add a HAVING clause in my selection as the final select statement sent to the database of the data store will be:

  Select d.region_name

  count (*)

  location d

  f participation

  where f.location_id = d.location_id

  D.region_name group

  Having count (*) > 5

  With such a feature, I could drop the possibility of exploiting personal data.

  ARO

  Ales

  If you wanted to make a report, do you mean to use an Oracle BI answers, really?

  First of all, in Oracle BI answers, you could change your filter:

  

  And use something like this:

  

  All,

  

  Kind regards

• Use of the place where and having clause

  Hi all
  
  I have always a doubt as to the use of HAVING and WHERE clause.
  Suppose I have table T1 with a single column C1
  
  CREATE TABLE T1
  (C1 VARCHAR2 (1));
  
  who data following INSERT scripts
  
  INSERT INTO T1 VALUES ('A');
  INSERT INTO T1 VALUES ('B');
  
  INSERT INTO T1 VALUES('C');
  
  INSERT INTO T1 VALUES ('A');
  INSERT INTO T1 VALUES ('B');
  
  INSERT INTO T1 VALUES ('A');
  
  
  Now I want the result as follows
  
  
  
  C1 = COUNT (C1)
  ==============
  B = 2
  A = 3
  
  Then out of query 1 and 2, approach who is right?
  
  (1) SELECT C1, COUNT (C1) FROM T1
  WHERE C1 <>'C '.
  GROUP BY C1
  ORDER BY C1 DESC;
  
  (2) SELECT C1, COUNT (C1) FROM T1
  GROUP BY C1
  SEEN C1 <>'C '.
  ORDER BY C1 DESC;
  
  Published by: user13306874 on June 21, 2010 02:36

  In SQL, it is always better to filter data as soon as possible of the moment.
  In your example, the WHERE clause would be right now:

  SQL> explain plan for
    2  select c1,count(c1)
    3  from t1
    4  where c1 != 'C'
    5  group by c1
    6* order by c1 desc;
  
  Explained.
  
  SQL> select * from table(dbms_xplan.display);
  
  PLAN_TABLE_OUTPUT
  ----------------------------------------------------
  Plan hash value: 3946799371
  
  ----------------------------------------------------
  | Id  | Operation          | Name | Rows  | Bytes |
  ----------------------------------------------------
  |   0 | SELECT STATEMENT   |      |     5 |    10 |
  |   1 |  SORT GROUP BY     |      |     5 |    10 |
  |*  2 |   TABLE ACCESS FULL| T1   |     5 |    10 |
  ----------------------------------------------------
  
  Predicate Information (identified by operation id):
     2 - filter("C1"!='C')
  
  18 rows selected.
  
  SQL>
  

  As you can see that the filter is applied during the analysis of T1.

  Whereas in the case of HAVING:

  SQL> explain plan for
    2  select c1,count(c1)
    3  from t1
    4  group by c1
    5  having c1 != 'C'
    6* order by c1 desc;
  
  Explained.
  
  SQL> select * from table(dbms_xplan.display);
  
  PLAN_TABLE_OUTPUT
  ----------------------------------------------------
  Plan hash value: 3146800528
  
  ----------------------------------------------------
  | Id  | Operation           | Name | Rows  | Bytes |
  ----------------------------------------------------
  |   0 | SELECT STATEMENT    |      |     6 |    12 |
  |*  1 |  FILTER             |      |       |       |
  |   2 |   SORT GROUP BY     |      |     6 |    12 |
  |   3 |    TABLE ACCESS FULL| T1   |     6 |    12 |
  ----------------------------------------------------
  
  Predicate Information (identified by operation id):
     1 - filter("C1"!='C')
  
  18 rows selected.
  
  SQL>
  

  The analysis is performed after all groups have been calculated: which was calculated in vain, because it will be further filtered because of the HAVING clause.

  In general, I would like to use as a guide: If you do not use aggregate functions in your predicate in the HAVING clause, set this predicate to the WHERE of your query part.

  Published by: Toon Koppelaars June 21, 2010 11:54

• ORDER BY clause or HAVING clause... ERROR

  Hi, I request what error throws please find the solution for the application to go without danger...
  
  
  Select C.src_cd,
  D.src_nm,
  Count (*) supp_clm_count,
  (CASE WHEN A.clm_ttladjamt > 0 then A.clm_ttladjamt to another A.clm_ttlreqamt END) as the amount
  Of IWOWNER. WC_clm, a.
  IWOWNER. WC_clm_srvc B,.
  LOVELY. WC_SRC C.
  LOVELY. WC_SRC_lang D
  where A.clm_id = B.clm_id
  and A.rcv_loc_id = C.src_id
  and C.src_id = D.src_id
  and TRANSLATE (A.clm_rqst_type_cd) = "SUPPLIERCLAIM".
  and TRANSLATE (A.clm_typ_cd) in ('GUARANTEE', 'PRE-DELIVERY')
  and DATE (A.clm_create_dt) between January 1, 2000 "and on January 1, 2010"
  C.src_cd group, D.src_nm
  
  
  
  ERROR:
  -------
  
  SQLSTATE: 42803, SQLERRMC: CLM_TTLADJAMT
  Message: An expression starting with "CLM_TTLADJAMT" specified in a SELECT clause, HAVING clause, or ORDER BY clause is not specified in the GROUP BY clause or a SELECT clause, a HAVING clause ORDER BY clause with a column function and no GROUP BY clause is specified

  Hello

  With oracle, you can try this query.

  Salim cordially.

  /* Formatted on 2009/07/22 10:10 (Formatter Plus v4.8.8) */
  SELECT DISTINCT c.src_cd, d.src_nm,
                  COUNT (1) OVER (PARTITION BY c.src_cd, d.src_nm)
                                                                 supp_clm_count,
                  (CASE
                      WHEN a.clm_ttladjamt > 0
                         THEN a.clm_ttladjamt
                      ELSE a.clm_ttlreqamt
                   END
                  ) AS amount
             FROM iwowner.wc_clm a,
                  iwowner.wc_clm_srvc b,
                  showner.wc_src c,
                  showner.wc_src_lang d
            WHERE a.clm_id = b.clm_id
              AND a.rcv_loc_id = c.src_id
              AND c.src_id = d.src_id
              AND a.clm_rqst_type_cd = 'SUPPLIERCLAIM'
              AND a.clm_typ_cd IN ('WARRANTY', 'PRE-DELIVERY')
              AND a.clm_create_dt BETWEEN TO_DATE ('01/01/2000', 'dd/mm/yyyy')
                                      AND TO_DATE ('01/01/2010', 'dd/mm/yyyy')
  

• Apparently, my memory is almost full - no idea how to deal with not having a novice such am - sd card

  Apparently, my memory is almost full - no idea how to deal with not having a novice such am - sd card

  Hi sixty.

  It can also be a good idea to begin to transfer the content from your phone to a PC/Mac to help reduce the storage room.

  It is a process usually easy, just plug the unit to a PC, on the phone screen select MTP mode when it appears.

  Once finished, you just have to access the phone's internal memory and transfer pictures/videos from the DCIM folder.

  Let us know if you need help in this kind of task.

• Need help with dynamic where clause.

  Hello world.  I'm kind of a novice with PL/SQL, so please be patient with me.  I am building a web application using Weblogic Oracle 11 g and Oracle database.

  To make a long story short, I need to generate the complex dynamic WHERE clause.  I don't know what will be the name of the column or the operator of comparison or the value will be until the moment of execution.

  I'm not 100% certain, but I think I need to use dynamic SQL statements.  Unless there is a better way to do it.

  I'm sure that using dynamic SQL opens me up to SQL injection attacks if I use simple string concatenation.  I would like to use bind variable if possible.

  In any case, it seems like trying to build a dynamic where clause does not seem to work as I want it would be...

  bv1 := 'colA = 5 and colB >= 9';
  
  query_str := 'SELECT * FROM view WHERE :bv1';
  
  OPEN CUR FOR query_str USING bv1;
  

  The code above pours out when I try to run it.

  Is there a way to build dynamic where the provisions without the use of concatenation of simple string that leaves me open to attacks by SQL injection?

  Thank you

  Hello.  I have read briefly upward on things, you suggested, but it seems a bit abstract for me since I was a novice.

  Can you give us an explanation or an example of how to use it to accomplish what I need to do?

  Do ypu really plan create a query with a WHERE clause that contains a predicate for all possible columns?

  Variable bind offer performance when the same query is executed several times with different values for bind variables.

  Select * from emp where empno =: v_empno

  After the first run Oracle will always be this query in the library cache. The next running can use the version of the cache even if the bind value is different.

  Select * from emp where empno =: v_empno;

  Select * from emp where ename =: v_ename;

  Select * from emp where sal =: v_sal;

  Each one is a DIFFERENT query. Yes - each of them uses a variable binding, but queries are different.

  So, if your username select EMPNO, an '=', and a value from the first query can be executed. But the next time that the user can select the ENAME column so that 'stored' first request may not be used given that the query does not match.

  The third time your user selects SAL so none of the first two queries can be used in the library cache. The use of bind variables gives performance during the SAME QUERY is used multiple times with DIFFERENT values for the variable binding. Oracle is not to analyze the hard query everything simply because the VALUE is different.

  So bind variables do not help you if most of the queries were completely DIFFERENT sets of predicates. And SYS_CONTEXT will not help either since that simply provides a way to define binds the values of the variables.

  Just for the simple example above, there are at least SIX possible combinations of columns that could choose your user name: empno, ename, sal, empno ename, empno & sal, ename & sal

• Unable to set keyframes in the audio clip selected with the pen tool - can no longer see the volume control line after developing the clip. What could be the problem? This seems to be the problem on any project that I opened.

  Cannot set keyframes in the audio clip selected with the pen tool - cannot see even the line volume control more to set keyframes on having developed the clip. What could be the problem?  I use the first CC pro @.

  You can use the Ctrl key and then click with the regular selection tool, clicking with ctrl will make a new keyframe. in this way, have no need for the tools. CTRL has other great shortcuts that will stop you having to spend retouching pads tools too.

  If the automation keyframe line is missing, you can switch on and off in the menu sequence.

  

• Problem with order by clause

  Hai all,
  
  I have problem with order by clause,
  
  My query is
  
  "Select number from table1 order number CSA.
  
  and the output as
  
  1
  10
  12
  13
  15
  17
  19
  2
  20
  21
  22
  .
  ........
  
  But if we give the order that it should display below the only good?
  
  1
  2
  10
  12
  13
  15
  17
  19
  20
  21
  22 ........
  
  Help me please why it is not displayed as she. and how to make the statement to display as in the second case. Thanks in advance.
  
  Kind regards
  Uraja

  In your table, the column is VARCHAR2 data type.

  It is a string, even if it is to store numeric values, it will get condemn as strings

  Try the below

  select your_column from table1 order by to_number(your_column) asc 
  

  Published by: JAC on November 19, 2012 16:35

• How to create a procedure for REF CURSOR output with any WHERE clause?

  I have a requirement like this: I have a huge question that need to reuse in my code over 10 times. This SQL has about 50 lines. Here for these 10 odd times sometimes changes in WHERE clause (columns are the same). So I can't create a view as SQL is not static.
  
  I thought to write a procedure with a para of WHERE_CLAUSE entry. I pulled out a refcursor sys by adding where clause. But I can't do it because you cannot add a clause like this where clause.
  
  i.e.

  PROCEDURE dynamyic_query (p_where_clause IN VARCHAR2, p_out_query OUT SYS_REFCURSOR ) IS
  BEGIN
  
    OPEN p_out_query FOR SELECT ......... FROM table WHERE || ' ' || p_where_clause;
  
  END;

  The foregoing gives an error.
  
  How to handle such a situation? Any help would be greatly appreciated.

  Hello

  Channa wrote:
  ... However, when I change the procedure like this:

  PROCEDURE FORMS_TEXT_DYN_SQL_TEST(p_where_cluase IN VARCHAR2, p_out_cursor OUT SYS_REFCURSOR) IS
  v_stmt VARCHAR2(1000);
  BEGIN
  v_stmt := 'SELECT tname FROM tab_test WHERE tname LIKE ''%ABS_V4%'' AND :y';
  
  OPEN p_out_cursor FOR v_stmt using p_where_cluase;
  
  END;
  

  And run this block of code:

  declare
  v_tname varchar2(200);
  out_cursor sys_refcursor;
  begin
  forms_text_dyn_sql_test(' 1 = 1 ', out_cursor );
  LOOP
  fetch out_cursor INTO v_tname;
  exit when out_cursor%NOTFOUND;
  DBMS_OUTPUT.PUT_LINE(v_tname);
  END LOOP;
  end;
  /
  

  I get the error:

  [1]: (Error): ORA-00920: invalid relational operator ORA-06512: at "ABS.FORMS_TEXT_DYN_SQL_TEST", line 6 ORA-06512: at line 5
  

  Looks like you can only set column_name =: z, column_name =: values of type y. You can not it seems to replace it with no WHERE CLAUSE?

  A bind variable, such as: it, represents a single value.
  If: is the VARCHAR2 '1 = 1', then

  SELECT tname FROM tab_test WHERE tname LIKE '%ABS_V4%' AND :y
  

  takes the value

  SELECT tname FROM tab_test WHERE tname LIKE '%ABS_V4%' AND '1 = 1'
  

  I think you want something like this:

  CREATE OR REPLACE PROCEDURE FORMS_TEXT_DYN_SQL_TEST
  (     p_where_clause      IN      VARCHAR2
  ,      p_out_cursor      OUT      SYS_REFCURSOR
  ) IS
    v_stmt VARCHAR2(1000);
  BEGIN
    v_stmt := 'SELECT ename FROM scott.emp WHERE ename LIKE ''%A%'' AND '
            || p_where_clause;
  
    OPEN p_out_cursor FOR v_stmt;
  
  END;
  /
  show errors
  
  SET  SERVEROUTPUT  ON
  
  declare
    v_tname varchar2(200);
    out_cursor sys_refcursor;
  begin
    forms_text_dyn_sql_test(' 1 = 1 ', out_cursor );
    LOOP
      fetch out_cursor INTO v_tname;
      exit when out_cursor%NOTFOUND;
      DBMS_OUTPUT.PUT_LINE(v_tname);
    END LOOP;
  end;
  / 
  

  Output:

  ALLEN
  WARD
  MARTIN
  BLAKE
  CLARK
  ADAMS
  JAMES
  

• outer join on query with the GOLD clause

  Hi all, I have a problem outerjoining a clause or a request with
  Here's my data
  

  WITH table1 AS
  (
   SELECT  'test' txt1, 'pak' txt2, 'ced' txt3, 'su' txt4 FROM dual UNION ALL
   SELECT  null txt1, 'pak' txt2, 'ced2' txt3, 'su2' txt4 FROM dual UNION ALL
   SELECT  null txt1, NULL txt2, 'ced3' txt3, 'su3' txt4 FROM dual UNION ALL
   SELECT  null txt1, NULL txt2, null txt3, 'su3' txt4 FROM dual UNION ALL
   SELECT  'text5' txt1, NULL txt2, null txt3, 'su3' txt4 FROM dual UNION ALL
   SELECT  null txt1, NULL txt2, null txt3, null txt4 FROM dual 
  
  
  )
  ,table2 AS
  (
   SELECT 111 pid, 'test' txt1, 'pak4' txt2, 'ced' txt3, 'su' txt4 FROM dual UNION ALL
   SELECT 222 pid, 'test1' txt1, 'pak' txt2, 'ced2' txt3, 'su2' txt4 FROM dual UNION ALL
   SELECT 333 pid, 'test2' txt1, 'pak3' txt2, 'ced3' txt3, 'su4' txt4 FROM dual UNION ALL
    SELECT 444 pid, 'test2' txt1, 'pak3' txt2, 'ced4' txt3, 'su3' txt4 FROM dual 
  
  )
  SELECT b.pid, a.*
  from table1 a, table2 b
  WHERE (a.txt1 = b.txt1 OR
         a.txt1 IS NULL AND a.txt2=b.txt2 OR
         Nvl(a.txt2, a.txt1) IS NULL AND a.txt3 = b.txt3 OR
         Nvl(a.txt2, a.txt1) IS NULL  AND a.txt3 IS NULL AND a.txt4 = b.txt4
         ) 

  as you can see I am attaching the table 1 and table 2. I join with txt1, if txt1 is null, then join by txt2, if null, then reach by txt3 and so on.
  the above produces this output code
  

  PID     TXT1     TXT2     TXT3     TXT4
  ===     ====     ===   ==== ====
  111     test     pak      ced     su
  222             pak      ced2     su2
  333                     ced3     su3
  444                          su3

  This output is partially correct. only 4 lines were in display and two has been left out

   SELECT  'text5' txt1, NULL txt2, null txt3, 'su3' txt4 FROM dual UNION ALL
   SELECT  null txt1, NULL txt2, null txt3, null txt4 FROM dual 

  I tried to use the outer join, but oracle will complain that I can't use outerjoin with the GOLD clause.
  
  can someone modify my query to display the output below?
  

  PID     TXT1     TXT2     TXT3     TXT4
  ===    ====      ===   ====  ====
  111     test     pak      ced     su
  222             pak      ced2     su2
  333                     ced3     su3
  444                          su3
  NULL  NULL   NULL    NULL   NULL
       test5 

  Not sure you can do it with the style of the Oracle of outer joins, but open ANSI style joins is simple, in fact exactly as you had it.

  SQL> set null null;
  SQL> WITH table1 AS (
    2   SELECT  'test' txt1, 'pak' txt2, 'ced' txt3, 'su' txt4 FROM dual UNION ALL
    3   SELECT  null txt1, 'pak' txt2, 'ced2' txt3, 'su2' txt4 FROM dual UNION ALL
    4   SELECT  null txt1, NULL txt2, 'ced3' txt3, 'su3' txt4 FROM dual UNION ALL
    5   SELECT  null txt1, NULL txt2, null txt3, 'su3' txt4 FROM dual UNION ALL
    6   SELECT  'text5' txt1, NULL txt2, null txt3, 'su3' txt4 FROM dual UNION ALL
    7   SELECT  null txt1, NULL txt2, null txt3, null txt4 FROM dual),
    8  table2 AS (
    9   SELECT 111 pid, 'test' txt1, 'pak4' txt2, 'ced' txt3, 'su' txt4 FROM dual UNION ALL
   10   SELECT 222 pid, 'test1' txt1, 'pak' txt2, 'ced2' txt3, 'su2' txt4 FROM dual UNION ALL
   11   SELECT 333 pid, 'test2' txt1, 'pak3' txt2, 'ced3' txt3, 'su4' txt4 FROM dual UNION ALL
   12   SELECT 444 pid, 'test2' txt1, 'pak3' txt2, 'ced4' txt3, 'su3' txt4 FROM dual)
   13  SELECT b.pid, a.*
   14  from table1 a
   15     LEFT JOIN table2 b
   16        ON (a.txt1 = b.txt1 OR
   17            a.txt1 IS NULL AND a.txt2=b.txt2 OR
   18            Nvl(a.txt2, a.txt1) IS NULL AND a.txt3 = b.txt3 OR
   19            Nvl(a.txt2, a.txt1) IS NULL  AND a.txt3 IS NULL AND a.txt4 = b.txt4);
  
         PID TXT1   TXT2   TXT3   TXT4
  ---------- ------ ------ ------ ------
         111 test   pak    ced    su
         222 null   pak    ced2   su2
         333 null   null   ced3   su3
         444 null   null   null   su3
  null       text5  null   null   su3
  null       null   null   null   null
  

  John

• How to filter the result with a where clause clause?

  Hello
  
  I have a nice sql query. I would like to include a where clause with a joker clause.
  
  If the user selects for example a "AGENT_ID", then where clause would include the AGENT_ID, otherwise, every thing would appear.
  

  SELECT 
     "FICHE"."FICHE_ID" "FICHE_ID", 
     "FICHE"."AGENT_ID" "AGENT_ID", 
     "FICHE"."EQUIPE_ID" "EQUIPE_ID", 
     "FICHE"."DATE_ACTIVITE" "DATE_ACTIVITE", 
     "FICHE"."DATE_CREATION" "DATE_CREATION", 
     "FICHE"."TYPE_FICHE" "TYPE_FICHE", 
     "FICHE"."FICHE_CREE_PAR" "FICHE_CREE_PAR", 
     "FICHE"."TOTAL_HEURES" "TOTAL_HEURES", 
     "FICHE"."TOTAL_HEURES_EXCEPTIONNELLES" "TOTAL_HEURES_EXCEPTIONNELLES", 
     "FICHE"."DATE_MODIFICATION" "DATE_MODIFICATION"
  FROM 
     "FICHE"

  The WHERE clause:
  WHERE AGENT_ID = ' * ' if nothing is selected.
  
  WHERE AGENT_ID = "4290" If an agent ID is selected.
  
  Hope I'm clear that English is not my mother tongue.
  
  Thanks for you help.
  
  Christian

  Hello

  Try this

  WHERE "FICHE".AGENT_ID like NVL(:val, '%')
  

  see you soon

  VT