Join queries from 4 tables

Similar Questions

• Migration of data from a table to another table

  have a table1 that includes the existing data in the format.
  
  ~@!%~X1~@!%~Y1 in three different coulmns
  
  creates a new empty table and the need to migrate the data above, which is present in 3 different columns in a column in the new table, as shown in the example below.
  
  table 1 existing data (| separator of columns for formatting)
  ------------------------------------------------------------------------------------------------------------------------------------
  ID name1 | Name2. Name3
  123 ~@!%~X1~@!%~Y1 | ~@!%~X2~@!%~Y2 | ~@!%~X3~@!%~Y3
  234 ~@!%~X4~@!%~Y4 | ~@!%~X5~@!%~Y5 | ~@!%~X6~@!%~Y6
  456 ~@!%~X7~@!%~Y7 | ~@!%~X8~@!%~Y8 |     ~@!%~X9~@!%~Y9
  
  
  Table 2, which will initially be empty and after migration, it should look as follows.
  
  ID name1
  ----------------------------------------------------------------------------------------------------------------------------
  123 ~@!%~X1~@!%~Y1 & & ~@!%~X2~@!%~Y2 & & ~@!%~X2~@!%~Y2
  234 ~@!%~X4~@!%~Y4 & & ~@!%~X5~@!%~Y5 & & ~@!%~X6~@!%~Y6
  456 ~@!%~X7~@!%~Y7 & & ~@!%~X7~@!%~Y7 & & ~@!%~X7~@!%~Y7
  
  as shown in the example above
  
  Name1 column has ~@!%~X1~@!%~Y1
  Column name2 has ~@!%~X2~@!%~Y2
  Name3 column has ~@!%~X3~@!%~Y3
  
  Once the data is migrating from table 1 for id - 123 looks like below, before joining data from 3 tables, I need apopend & & for each for the token I read of the tabl1 of name1 to end ii should be added "& &" also when I read the name2 I add "& &" at the end of the string before the concatination.
  
  Here's the sample that deals with data for id - 123 with & & (only & & other symbols are part of the data)
  
  ~@!%~X1~@!%~Y1 & & ~@!%~X2~@!%~Y2 & & ~@!%~X2~@!%~Y2
  
  need help in writing a note of migate
  
  Published by: [email protected] on April 2, 2010 15:42

  Hello

  You are looking for something like this

  CREATE TABLE table_new
  AS
     (SELECT id, name1 || '&&' || name2 || '&&' || name3 name1
      FROM table1);
  

  or if you have the table ready

  INSERT INTO table_new
     (SELECT id, name1 || '&&' || name2 || '&&' || name3 name1
      FROM table1);
  

  Thank you

  Alen

• Storage from multiple tables in one MySQL query

  So I would like, with a single MySQL query, extract data from a table (say 15 rows of data) and then extract data from another table (lets say another 15 ranks) and who joined the first, so that I can view and paginate a single list (of 30 rows now).  I don't think that a join is what I'm looking for, because I want the second list should be annexed to the first and not mixed in.  I would just use different queries, but I already have the paging function implemented for a single set of records, and implementation with two (where the second begins when the first leaves) seems daunting. However, if you have any tips on one, they would really help.

  Thank you

  You need to use a UNION to join the two select statements. The number of columns and the data types of each selection must match.

• Recursive queries with external table anchor

  Hi, I had the tables A and B. In table B, reviews are a tree structure. For each record in A, I want to get the respective records in B, in the following way:
  
  SELECT
  a.x, b.y
  Of
  one
  CROSS JOIN
  (
  SELECT * FROM b
  START WITH id = a.x
  CONNECTION BY b.id = b.parentid PRIOR
  ) b
  
  And I get the error indicating that the column a.x is not known. Is it possible to use columns of the external table as anchor points for a recursive queries? Or y at - it another way to form such a query?
  
  Thank you.
  
  Published by: David Hoksza on 13.11.2008 17:07

  Hello

  If all lines in b "inherits" the help of all its descendants.
  In this case, do a subquery on b which associates each line with all its descendants (and itself).

  WITH     bt     AS
  (
       SELECT DISTINCT     aid
       ,          CONNECT_BY_ROOT id     AS root_id
       FROM          b
       CONNECT BY     parentid     = PRIOR id
  )
  SELECT     a.id          AS "a.id"
  ,     bt.root_id     AS "b.id"
  FROM     a
  JOIN     bt     ON     a.id     = bt.aid
  ;
  

  Note that there is no START WITH clause. We want each row of b is considered to be the root of its own subtree.

  Sorry, I don't have Express Edition to test this. It works in Oracle Enterprise Edition 10.
  If CONNECT_BY_ROOT does not work (it was not available in Oracle 9), you can get the same information from SYS_CONNECT_BY_PATH.

• ORA-01445: cannot select ROWID, or sample, a view of join without key - preserved table form w/report and LOV

  I create a page "Form on a Table with Report" (APEX 5). Table I has 3 columns: cat_key, item_value, item_description. The cat_key is a foreign key in a table with columns cat_key, cat_value. I created the report and everything works fine.  The page of the report shows my domain cat_key in number, so I want it instead to display the cat_value. I change the field cat_key to a text based on LOV, select my LOV and now when I view the page of report I get the "'ORA-01445: cannot select ROWID from where sample, a join without key preserved table view ' error." The report uses the ROWID as a link to the edit page field.

  If I use the same tables and even shared LOV in a report page interactive, it works fine.

  What I am doing wrong?

  This is a limitation of the IR Apex will rewrite your query to a query that includes the lookup table and the columns. So now you have actually created a query on an inline view, which causes the error. You can see the query apex created when running in debug mode. APEX will join the table of choice with your primary table. For example from one of my IR:

  Select 'ID '...

  "MSL". "" HIS_COUNTRIES "=> my primary table

  ) ) b,

  (select rv_meaning d, rv_low_value r

  of msl_ref_codes_table

  where rv_domain = "JOHN".

  ) L1 = > query My LOV Lookup
  

  ...

  So, if possible, base your reports and forms on the real primary key, no rowid.

  You can also write the IR query with the search yourself:

  Select T1.rowid...

  of primary_table T1

  look_table L1

  ...

• To create an interactive report in the apex by selecting from multiple tables

  Hi, I am creating an interactive report by selecting from multiple tables.

  SELECT w.FIRST_NAME as name, w.SURNAME as name, i.ROAD Road, i.DATE_OF_INC as DATE_OF_INC, S.STATEMENT as a STATEMENT OF Declaration

  JOIN THE

  WITNESS w

  ON w.witness_id = s.FK1_WITNESS_ID

  JOIN THE

  Incident I have

  WE

  i.incident_no = w.FK1_INCIDENT_NO

  JOIN THE

  user_station ps

  ON ps.station_id = i.nearest_station_id

  JOIN THE

  the user in.

  WE

  in. STATION_ID = ps.station_id

  WHERE po.officer_id = 1

  by I continue to encounter this error "the report query requires a unique key to identify each row. The supplied key cannot be used for this query. Please change the report attributes to define a unique key column. "ORA-01445: cannot select ROWID from where sample, a join without key preserved table view '

  So I googled around and found that in the attibutres tables report, I need to change the "LINK" COLUMN so first, I changed "Link to custom target", but the report is so I changed it to "Exclude link Clolumn" again, the report didn't report and I STILL got a blank page with only the tabs.

  I wonder you can not create a report by selecting from multiple tables?

  If you can please I need your help.

  Thank you

  You can, but in this case, it might be easier to build your report tables that are joined Oracle views and then build your report out of the newly built sight...

  Or wrap a selection around your selection with joins, and then make the where clause on the external selection...

  Thank you

  Tony Miller
  Software LuvMuffin
  Ruckersville, WILL

• Join conditions, update a table

  Hello

  I created a table of choice in the apex, so my Dispatcher column, moved to the new table. (first time doing so)

  I created again the dispatcher of the column in the first table.

  I have the following code:

  SELECT R.DISPATCHER, B.DISPATCHER

  OF GCAT_V3 R JOIN GCAT_DISPATCHER_LOOKUP B

  ON R.DISPATCHER_ID = B.DISPATCHER_ID

  DISPATCHER	DISPATCHER
  -	SU. [email protected]
  -	SU. [email protected]
  -	SU. [email protected]
  -	SU. [email protected]

  I need to copy the data back.

  Basically, what I have in the column that b.Dispatcher must be copied to another column dispatcher.

  Any ideas?

  Kind regards

  Leonard

  Hi, Leonard,

  Whenever you have a question, please post a small example of data (CREATE TABLE and only relevant columns, INSERT statements) for all of the tables involved and the accurate results you want from this data, so that people who want to help you can recreate the problem and test their ideas.

  If you ask about a DML operation, such as UPDATE, then INSERT statements, you post should show what looks like the tables before the DML, and the results will be the content of the table changed after the DML.

  Explain, using specific examples, how you get these results from these data.

  Always say what version of Oracle you are using (for example, 11.2.0.2.0).

  See the FAQ forum: Re: 2. How can I ask a question on the forums?

  You can use the UPDATE or MERGE to copy data from a table to the existing lines in another.  For example:

  UPDATE gcat_v3 r

  SET dispactcher =)

  SELECT b.dispatcher

  OF gcat_dispatcher_lookup b

  WHERE b.dispatcher_id = r.dispatcher_id

  )

  ;

  The above statement assumes that dispatcher_id is unique in gcat_dispatcher_lookup; But there is no need to be unique in gcat_v3.

• Extract single line transformed from two tables

  I have two tables T1 and T2 with the same set of columns. The columns are C1, C2, C3, COND1. I need to issue a query that returns a line transformed in order to respect the following rules:

  1. the returned line may contain values from both tables based on the values of column in table T1.

  2 return the column C1 in table T1; If it is NULL, then return to any value is present in the column C1 of T2.

  3. article 2 above apply to all columns like C1, C2 etc.

  I published the following query. The problem is that if a subquery does not reach the entire query fails. Someone help me please. Probably there is another simple method.

  SELECT NVL (T1.c1, T2.c1) c1, c2 (T1.c2, T2.c2) NVL NVL (T1.c3, T2.c3) c3

  FROM (SELECT c1, c2, c3

  FROM T1

  WHERE cond1 = 'T10') T1

  (SELECT c1, c2, c3

  THE T2

  WHERE cond1 = "T200") T2;

  Hello

  user4274403 wrote:

  I have two tables T1 and T2 with the same set of columns. The columns are C1, C2, C3, COND1. I need to issue a query that returns a line transformed in order to respect the following rules:

  1. the returned line may contain values from both tables based on the values of column in table T1.

  2 return the column C1 in table T1; If it is NULL, then return to any value is present in the column C1 of T2.

  3. article 2 above apply to all columns like C1, C2 etc.

  I published the following query. The problem is that if a subquery does not reach the entire query fails. Someone help me please. Probably there is another simple method.

  SELECT NVL (T1.c1, T2.c1) c1, c2 (T1.c2, T2.c2) NVL NVL (T1.c3, T2.c3) c3

  FROM (SELECT c1, c2, c3

  FROM T1

  WHERE cond1 = 'T10') T1

  (SELECT c1, c2, c3

  THE T2

  WHERE cond1 = "T200") T2;

  Are you saying that if there is no row in t1 that satisfy the condition cond1 = "T10", then you want to keep seeing the lines (and values) of the t2?

  And even, if there is no row in t2 that satisfy the cond1 = "T200" condition, then you want to keep seeing the lines (and values) from t1?

  If so, this looks like a job for a full outer join.

  You can change the join in a full outer join as follows:

  SELECT NVL (T1.c1, T2.c1) AS c1

  , NVL (T1.c2, T2.c2) C2

  NVL (T1.c3, T2.c3) AS c3

  FROM T1

  FULL OUTER JOIN T2 ON t1.cond1 = 'T10 '.

  AND t2.cond1 = 'T200.

  WHERE t1.cond1 = 'T10 '.

  OR t2.cond1 = "T200"

  ;

  If you would care to post CREATE TABLE and INSERT instructions for some examples of data, then I could test this.

  What happens if no table has all the lines that have good values in cond1?

• Update multiple columns from multiple tables in a single UPDATE request

  Hello

  I'm trying to figure if I'm heading in the right direction.

  I want to update multiple columns from multiple tables in a single UPDATE request. Also, I would like to update multiple columns in a table from the tables.

  Scenario 1

  UPDATE Table2, Table 3
  SET T2.Column1 = T1.Column1 
  ,T2.Column2 = T1.Column2
  ,T3.Column2 = T1.Column2
  FROM Table1 T1, Table2 T2, Table3 T3
  WHERE T1.id = T2.id
  and T1.id = T3.id
  
  

  Scenario 2

  UPDATE Table3
  SET T3.Column1 = T1.Column1 
  T3.Column2 = T1.Column2
  ,T3.Column3 = T2.Column3
  ,T3.Column4 = T2.Column4
  FROM Table1 T1, Table2 T2, Table3 T3
  WHERE T3.id = T1.id
  and T3.id = T2.id
  
  

  Hello

  For scenario 1, you must write separate instructions UPDATE table2 and table3.

  To guard against someone else change one of these tables while you act so you can copy all relevant data in a global temporary table and update this global temporary table table3.

  ENGAGE only when two tables have been changed.

  You can write a procedure or an INSTEAD OF trigger to do all this.

  For scenario 2, you can reference many tables that you need when new table3.  It might be more efficient and simpler to use the MERGER rather than UPDATED.  For example:

  MERGE INTO table3 dst

  WITH THE HELP OF)

  SELECT t1.id

  t1.column1

  t1.column2

  t2.column3

  t2.column4

  FROM table1 t1

  JOIN table2 t2 ON t1.id = t2.id

  )             src

  WE (dst.id = src_id

  WHEN MATCHED THEN UPDATE

  SET dst.column1 = src.column1

  dst.column2 = src.column2,

  dst.column3 = src.column3,

  dst.column4 = src.column4,

  ;

• How to join this per_rating_levels this table with query table.

  Dear all,
  
  Guide how 2 join me per_rating_levels this table with query because, I want 2 see the per_rating_levels.name against all employees.
  When I join this table with query it shows several recording/cortion against this record.
  
  Query:
  
  SELECT
  PAPF.full_name employee_name,
  papf1.full_name supervisor_name,
  WOMEN'S WEAR. Employee_number,
  
  hr_general.decode_job (PAAF.job_id) job_name,
  Department of hr_general.decode_organization (PAAF.organization_id),
  PC.Name, PCE.Comments EmployeeComments,
  (by selecting pce1.comments in per_competence_elements pce1
  where
  PCE.assessment_id = pce1.assessment_id
  AND pce.competence_id = pce1.competence_id
  AND pce1.object_id = pce.object_id) ManagerComments;
  
  --(sélectionnez rtl.name dans rtl où les pc.) RATING_SCALE_ID = rtl. Name RATING_SCALE_ID)
  
  
  OF per_all_people_f women's wear.
  per_all_people_f papf1,
  per_all_assignments_f ADP,
  PA per_appraisals,
  pat per_appraisal_templates,
  per_assessments not,
  per_competence_elements pce,
  per_competences pc
  
  
  WHERE papf.person_id = paaf.person_id
  AND paaf.supervisor_id = papf1.person_id
  AND paaf.primary_flag = 'Y '.
  AND pa.appraisee_person_id = papf.person_id
  AND pa.appraisal_template_id = pat.appraisal_template_id
  AND pa.appraisal_id = pas.appraisal_id
  AND pat.assessment_type_id = pas.assessment_type_id
  AND pas.assessment_id = pce.assessment_id
  AND pce.object_id = papf.person_id
  AND pce.competence_id = pc.competence_id
  AND trunc (sysdate) BETWEEN papf.effective_start_date AND papf.effective_end_date
  AND trunc (sysdate) BETWEEN papf1.effective_start_date AND papf1.effective_end_date
  AND trunc (sysdate) BETWEEN paaf.effective_start_date AND paaf.effective_end_date
  
  - AND papf.employee_number =: p_employee_number
  - AND pa.appraisal_date =: p_appraisal_date
  - AND papf.business_group_id =: p_bg_id
  
  order of papf.employee_number
  
  
  Concerning

  user10941925 wrote:
  Dear all,

  Guide how 2 join me per_rating_levels this table with query because, I want 2 see the per_rating_levels.name against all employees.
  When I join this table with query it shows several recording/cortion against this record.

  '2' in your question means "to"? If so please do not use text instant message in this forum.

  Now I suppose that PRE_RATING_LEVELS is a table in your application. And you are trying to include this table in an existing query. But in doing so, you have found the Cartesian product, correct?

  In fact, how do you think someone a public forum without any knowledge of your table and data structure could help you?

  Lets see, here's your query. I formatted.

  
  select papf.full_name                                       employee_name
       , papf1.full_name                                      supervisor_name
       , papf.employee_number                                 employee_number
       , hr_general.decode_job(paaf.job_id)                   job_name
       , hr_general.decode_organization(paaf.organization_id) department
       , pc.name                                              name
       , pce.comments                                         employeecomments
       , (
            select pce1.comments
              from per_competence_elements pce1
             where pce.assessment_id = pce1.assessment_id
               and pce.competence_id = pce1.competence_id
               and pce1.object_id = pce.object_id
         )                                                    managercomments
    from per_all_people_f        papf
       , per_all_people_f        papf1
       , per_all_assignments_f   paaf
       , per_appraisals          pa
       , per_appraisal_templates pat
       , per_assessments         pas
       , per_competence_elements pce
       , per_competences         pc
   where papf.person_id           = paaf.person_id
     and paaf.supervisor_id       = papf1.person_id
     and paaf.primary_flag        = 'Y'
     and pa.appraisee_person_id   = papf.person_id
     and pa.appraisal_template_id = pat.appraisal_template_id
     and pa.appraisal_id          = pas.appraisal_id
     and pat.assessment_type_id   = pas.assessment_type_id
     and pas.assessment_id        = pce.assessment_id
     and pce.object_id            = papf.person_id
     and pce.competence_id        = pc.competence_id
     and trunc(sysdate) between papf.effective_start_date  and papf.effective_end_date
     and trunc(sysdate) between papf1.effective_start_date and papf1.effective_end_date
     and trunc(sysdate) between paaf.effective_start_date  and paaf.effective_end_date
  order
      by papf.employee_number 
  

  Now, you want to add the PRE_RATING_LEVELS in the list so that you can use the column NAME.

  First thing you need to do is to determine the relationship between PRE_RATING_LEVELS and other tables. A relationship can be

  1. one on one
  2 one-to-many
  3. - to-several

  So when you tried to join, your state of health has resulted in 2nd or 3rd type of relationship. If you arrive with someone who knows the business and the data and find the table that could uniquely identify a line of PRE_RATING_LEVELS.

• Explicit cursor, column from several tables in a variable

  Here's the situation. This DB has tables of metadata. An example of table name would be g3e_attribute. Once you insert a record in a metadata table a trigger of executions insert and insert records in the g3e_attribute_0009 table (table of language, this is what I call it). It would insert in g3e_attribute_000C is the default language is the French. There is also a delete trigger so that if you delete a record in a table of metadata it deletes the corresponding language table. Well, somehow we have entered the language tables that are not in the metadata table. A problem with the system.
  
  There are 139 tables of language so I prefer not to do it manually.
  
  
  So I'm keen to make this code
  
  delete < table language > where < primary > not in (select < primary > in < metadata table >).
  
  So I need to loop through the table and and the corresponding primary keys.
  
  I have this code to get the table names
  Select table_name, substr (table_name, 0, (length (table_name)-5)) in all_tables where table_name like 'G3E_ % _000% ';
  
  This will display both the language metadata tale table names and.
  The table of language with either _0009 or _000C as a suffix, I take everything but the last 5 characters to get the table metadata.
  
  And I have the code to find the primary key of the table.
  SELECT column_name FROM ALL_CONS_COLUMNS A JOIN ALL_CONSTRAINTS C ON A.CONSTRAINT_NAME = C.CONSTRAINT_NAME WHERE 'G3E_LEGENDENTRY' AND C.CONSTRAINT_TYPE = C.TABLE_NAME = 'P ';
  
  
  So I just need to loop through all the tables. Put these pieces together, that's where I encountered a problem. I started to do a procedure, but I ran into a problem.
  
  When you loop, you move the record being the explicit cursor in a variable. Well, you can set the variable as a vertain column type (%), or type (%ROWTYPE) of the line. Although the sql I use to create the cursor takes columns from multiple tables and returns three columns. So I'm not sure how ot define the variable. The only work around I found was to make a table column_name that contains the table table,. Define the variable as rowType.
  
  Is there an easier way to do this?
  
  
  Thank you

  A couple of points.

  You should not really mix ansi joins and joins in the place where clause. Choose a style and stick to it.

  You can't do the DDL directly in pl/sql. You need to do this, use dynamic sql. something like:

  create or replace
  PROCEDURE SYNC_LANGMETATABLES AS
  
  CURSOR c_TABLE_PRIMARY IS
     select D.table_name, substr(D.table_name,0,(length(D.table_name)-5)),
            column_name
     from all_tables D,
        JOIN ALL_CONSTRAINTS C
           on C.TABLE_NAME = D.TABLE_NAME
        join ALL_CONS_COLUMNS A
           ON A.CONSTRAINT_NAME = C.CONSTRAINT_NAME
     where D.table_name like 'G3E_%_000%' AND
           C.CONSTRAINT_TYPE = 'P';
  
     v_LANGTABLE ALL_TABLES.TABLE_NAME%TYPE;
     v_METATABLE ALL_TABLES.TABLE_NAME%TYPE;
     v_PRIMKEY ALL_CONS_COLUMNS.COLUMN_NAME%TYPE;
  
  BEGIN
     OPEN c_TABLE_PRIMARY;
     LOOP
        FETCH c_TABLE_PRIMARY INTO v_LANGTABLE,v_METATABLE,v_PRIMKEY;
        EXIT WHEN c_TABLE_PRIMARY%NOTFOUND;
        execute immediate 'delete from '||v_langtable||' where '||v_PRIMKEY||
                          ' NOT IN(SELECT '||v_PRIMKEY||' FROM v_METATABLE)';
     END LOOP;
     CLOSE C_TABLE_PRIMARY;
  END;
  

  John

• help in registration of the records from two tables

  HI: I have two tables joined the first field. The field is the primary key in the first table. Need help listing records from both tables with each a line/record results.
  

  create table EVENTS (
  event_key varchar2(64) primary key,
  event_description varchar2(64),
  create_time int
  );
  
  
  create table EVENT_UPDATES (
  event_key varchar2(64) NOT NULL ,
  update_description varchar2(64),
  update_time int
  );
  
  
  insert into EVENTS values('Event1', 'This is event1', 1);
  insert into EVENT_UPDATES values('Event1', 'Ticket created', 3);
  insert into EVENT_UPDATES values('Event1', 'Event cleared', 10);
  insert into EVENTS values('Event2', 'This is event2', 4);
  insert into EVENT_UPDATES values('Event2', 'Ticket created', 6);
  insert into EVENT_UPDATES values('Event2', 'Event cleared', 8);

  I want to print each record in the table of EVENTS such as a line and the corresponding records in EVENT_UPDATES as a line like this record
  

  Event1   1     This is event1
              3     Ticket created
              10   Event cleared
  Event2   4     This is event2
              6     Ticket created
              8     Event cleared

  TIA
  Ravi

  select  case weight
            when 1 then event_key
          end key,
          time_val,
          description
    from  (
            select  event_key,
                    create_time time_val,
                    event_description description,
                    1 weight
              from  events
           union all
            select  event_key,
                    update_time,
                    update_description,
                    2 weight
              from  event_updates
          )
    order by event_key,
             weight
  /
  
  KEY          TIME_VAL DESCRIPTION
  ---------- ---------- -------------------------
  Event1              1 This is event1
                      3 Ticket created
                     10 Event cleared
  Event2              4 This is event2
                      6 Ticket created
                      8 Event cleared
  
  6 rows selected.
  
  SQL> 
  

  SY.

• Correspondence of the columns from one table to another

  I have a database column that contains the name of a person. I have another table that has a column that contains the names separated by a comma.

  I want to match records where all records of two columns have a corresponding name. I tried this:

  < cfquery name = "MyQuery" datasource = "myds" >

  SELECT *.

  OF TBL_A

  WHERE TBL_A.customername IN (#DifferentQuery.ListOfNames #)

  < / cfquery >

  Everything that I try results in only get a recording. Basically, I want to get all records from a table where one of its columns contains the name of a column in another table.

  It's a many-to-many relationship. That is to say an editorial can be linked to many ads, and vice versa.  He is best represented with three tables. Such as:

  Table: [Editorial] column: [EditorialID (PK), EditorialText,...]

  Table: Columns of the [advertiser]: [AdvertiserID (PK), AdvertiserName]

  Table:   [Editorial_Advertisers] columns: [EditorialID, AdvertiserID]

  The name of the advertiser throw himself into a column...

  my client can decide which ads get related to which editorials

  The third table stores the relationship between ads and editorials.  So instead of storing a list of advertisers, you need to insert record for each EditorialID + AdvertiserID combination. You can then use joins to retrieve all the advertisements for EditorialID 123.

  SELECT  a.AdvertiserName
  FROM     Editorial_Advertiser ea INNER JOIN Advertiser a ON ea.AdvertiserID = a.AdvertiserID
  WHERE   ea.EditorialID = 123
  

• data from 3 tables with later dates

  Hello
  Need help with the PL/SQL code, I need to write a code that will get the data from 3 tables with the most recent date.
  
  For an individual ACT_CODE the output of the SQL query should display the data including the last dates back to 3 tables, if there is no
  Date of the table, it should show the remaining data (think that the left join will do here)
  
  Names of tables:
  Institution_UPDT aiu
  AC ASQ_CONTACT
  GR_AUTHORIZE gr
  
  All 3 tables have ACT_Code in common
  
  Column names
  
  INSTITUTION_UPDT IAU - IAU. ACT_CODE, AIU.project_id as proj, IAU. UPDT_TYPE_ID, IAU. User_id, IAU. UPDT_DATE
  
  ASQ_CONTACT ac - ac. ACT_CODE as contact_code, ac.project_id, ac.first_name, ac.middle_initial, ac.last_
  Name, AC.title, AC. Status, AC.status_date
  
  GR_AUTHORIZE gr - GR ACT_CODE as grad_code, gr.name, gr.title AS grad_title, gr.submit_date
  
  
  Are the names of the columns date
  AC.status_date,
  IAU. UPDT_DATE and
  Gr.submit_date
  
  Thanks to you all
  appreciate your help
  
  Jesh

  Hi, Ngoumba,

  If a given ACT_Code couldn't miss from any of the tables, then you will use better full outer joins, not a join left outer.

  Perhaps it would be more effective to make a UNION of the three tables, then rotate the results in three datecolumns.

  You can use the GROUP BY aggregation to get the last date for each ACT_Code in each table.
  If you need other columns in the row which is the last date, you can use the ROW_NUMBER analytic function, like this:

  SELECT  ACT_Code
  ,     updt_date
  ,     ROW_NUMBER () OVER ( PARTITION BY  ACT_Code
                            ORDER BY          updt_date     DESC
                    ) AS r_num
  FROM    institution_updt
  

  The lines with r_num = 1 are the most recent

  This is a technique of ot the UNION-PIVOT example:

  WITH     union_data     AS
  (
       SELECT    ACT_Code
       ,       MAX (updt_date)     AS last_date
       ,       1                  AS table_id
       FROM       institution_updt
       GROUP BY  ACT_Code
  UNION ALL
       SELECT    ACT_Code
       ,       MAX (status_date)     AS last_date
       ,       2                  AS table_id
       FROM       ASQ_Contact
       GROUP BY  ACT_Code
  UNION ALL
       SELECT    ACT_Code
       ,       MAX (submit_date)     AS last_date
       ,       3                  AS table_id
       FROM       GR_Authorize
       GROUP BY  ACT_Code
  )
  SELECT       ACT_Code
  ,       MAX (CASE WHEN table_id = 1 THEN last_date END)     AS aiu_updt_date
  ,       MAX (CASE WHEN table_id = 2 THEN last_date END)     AS ac_status_date
  ,       MAX (CASE WHEN table_id = 3 THEN last_date END)     AS gr_submit_date
  FROM       union_data
  GROUP BY  ACT_Code
  ORDER BY  ACT_Code
  ;
  

  Published by: Frank Kulash, on September 16, 2009 15:02
  Added UNION-pivot example

• Select random from the tables

  Another problem of mine, I hope we can clear the air:
  I have 4-5 tables with different products, these products have a price. In another table, I have just a column with a value (my money). What I have to do, he chooses one product from each table and the sum of all should not be greater than my money? A randomization but how?
  If you need to learn more, say please and I'll post an example, I need help thanks

  Hello

  You can create the source of the page as a SQL something like element:

  WITH X AS (
  SELECT ROWNUM RR, DRINKS.NAME || ' ' || DRINKS.PRICE || ' ' || F1.PRODUCT || ' ' || F1.PRICE || ' ' || F2.PRODUCT || ' ' || F2.PRICE || ' ' || DESERT.PRODUCT || ' ' || DESERT.PRICE OUTPUT
  FROM MONEY M
  CROSS JOIN DRINKS
  CROSS JOIN F1
  CROSS JOIN F2
  CROSS JOIN DESERT
  WHERE DRINKS.PRICE + F1.PRICE + F2.PRICE + DESERT.PRICE <= M.MONEY)
  SELECT X.OUTPUT
  FROM X
  WHERE X.RR = TRUNC(DBMS_RANDOM.VALUE(1, (SELECT COUNT(*) FROM X)))
  

  Andy