Query dynamic SQL & Ref Cursor - debug errors

Similar Questions

• The use of bind variables in dynamic query created for Ref Cursor

  Hello
  
  I'm in a situation where there is a Ref cursor to which the query is built execution based on a loop. This is why the number of links would be known until the program runs.
  The application is currently using literals instead of bind variables.
  
  code snippet of the above is
  strSql: = "select * from emp where 1 = 1 and ().
  
  loop cursor1
  If cond is true then
  strSql = strSql | "ename = ' |" Cursor1.ColumnName;
  end loop;
  
  Open cursor2 for strSql;
  
  How to use links in the example above.

  sb92075 wrote:

  user13019948 wrote:
  Hello

  Here is the code I have my trying to change literal-based link to the base.

  What do you mean by "based bind?

  who, what, how determines the values to be 'bound '?

  He's referring to the coding style. He is currently using concatenated literal, and the goal is to change it to use the bindings.

  If I understand this it is known as method 4 dynamic SQL and requires DBMS_SQL. There are examples autour but they vary according to the type of statement being generated - SELECT statements require column lists to be parsed, unlike the INSERT/UPDATE/DELETE.

  This came up recently on my current project and I hit a demo. Here a table of names and values accepted procedure and had to build these in a single WHERE clause along the lines of

  AND t_names(i) = t_values(i)
  

  for an undetermined number of elements in the array. For this demonstration, I used a table that we called "attribute" (don't ask) which has columns including 'attribute_id' and 'name', and I need to build a query along the lines of

  select description from attribute where attribute_id = :b1 and name = :b2
  

  by the way '1012' and 'ISIN' respectively. (I use a table better and after a CREATE statement for her but I have to rush right now, sorry).

  declare
     k_sql_base        constant varchar2(500) := 'select description from attribute';
  
     t_names           constant varchar2_t := varchar2_t('attribute_id',  'name');
     t_values          constant varchar2_t := varchar2_t('1012',          'ISIN');
  
     l_sql             varchar2(500) := k_sql_base;
     l_rows_fetched    integer := 0;
     l_value           varchar2(4000);
  
     l_cursor_handle   integer;
  
  begin
     -- Construct the SQL statement with column names and bind variables e.g.
     -- 'select description from mars.attribute where attribute_id = :b1 and name = :b2'
     for i in t_names.first .. t_names.last loop
        l_sql := l_sql ||
           case i
              when t_names.first then ' where ' else ' and '
           end ||
           t_names(i) || ' = :b' || i;
     end loop;
  
     dbms_output.put_line('SQL statment = ' || l_sql); 
  
     -- Parse the statement we built above (the remaining steps require a parsed cursor):
     l_cursor_handle := dbms_sql.open_cursor;
     dbms_sql.parse(l_cursor_handle, l_sql, dbms_sql.native);
  
     -- Associate the 1st column of output with variable l_value - required for SELECT statements:
     -- (actually the 3rd param here 'column' seems to be only used to get a datatype, in this case we want a string -
     -- dbms_sql.column_value actually extracts the value into a specified variable, which can be different.
     -- All examples in the documentation pass a local variable without further comment, so not entirely clear what this does other than set the output datatype.)
     dbms_sql.define_column(l_cursor_handle, 1, l_value, 4000);
  
     -- Now go through values array binding actual values to :bn variables in the cursor (similar to USING clause of EXECUTE IMMEDIATE)
     for i in t_values.first .. t_values.last loop
        dbms_sql.bind_variable(l_cursor_handle, ':b'||i, t_values(i));
        dbms_output.put_line('Bound :b'||i || ' as ' || t_values(i));
     end loop;
  
     -- Open the cursor and fetch the result (no loop here because we are expecting a single-row result):
     l_rows_fetched := dbms_sql.execute_and_fetch(l_cursor_handle);
  
     -- 'Returns value of the cursor element for a given position in a cursor'
     -- Copy the value of column 1 to variable l_value (has to match
     -- dbms_sql.column_value(l_cursor_handle, 1, l_value);
     dbms_sql.column_value(l_cursor_handle, 1, l_value);
  
     dbms_output.put_line('Result = ''' || l_value || '''');
  
     dbms_sql.close_cursor(l_cursor_handle);
  end;
  

  Hope that helps...

• wanted to extract data from nested table pl/sql Ref Cursor getting an erro

  create or replace type 'DEPT12' as an object (dno number (2), dname varchar2 (30), varchar2 (50)) loc;
  
  create or replace type dept_tab in the table in "DEPT12".
  
  create or replace type 'LOC12' as an object (locno number, loc_name varchar2 (100))
  
  create or replace type loc_tab in the table of "LOC12.
  
  create or replace type dept_loc_rec1 as an object (dept_tab, eno number, loc_dt loc_tab dept_dt);
  
  Create type dept_loc_tb as table of the dept_loc_rec1
  
  create table dept_loc_tb_bk1 (dept_dt dept_tab, eno number, loc_dt loc_tab)
  NESTED TABLE dept_dt
  STORE AS dept_tab12,
  NESTED TABLE loc_dt
  STORE AS loc_tab12
  
  
  
  
  insert into dept_loc_tb_bk1 values (dept_tab (dept12(3,'ABD','LOC')
  dept12(4,'ABD','LOC')
  (, dept12(5,'ABD','LOC')), 3, loc_tab (loc12(21,'AAB'),
  loc12(22,'AAB'),
  loc12(23,'AAB')));
  
  When I try to extract data from Ref cursor to pl/sql table that I get an error ora-06504: pl/sql: return types of the result set of variables or request do not match.
  I created a table nested, as well as the pl/sql nested table object dept_loc_tb and I said the same dept_loc_tb lv_dept_loc_tb, but trying to get in this variable we get an error above.
  
  Please anyone can solve my problem.
  -----------------
  declare
  type cr is ref cursor;
  cr_obj cr;
  
  lv_dept_loc_tb dept_loc_tb;
  
  Start
  Open cr_obj to select dept_dt, eno, dept_loc_tb_bk1 loc_dt;
  collect the fetch cr_obj in bulk in lv_dept_loc_tb;
  close cr_obj;
  end;

  Your query selects 3 distinct columns requires so 3 collections of matching types. You want to treat these 3 columns as an object of type DEPT_LOC_REC1:

  SQL> declare
    2  type cr is ref cursor;
    3  cr_obj cr;
    4
    5  lv_dept_loc_tb dept_loc_tb;
    6
    7  begin
    8  open cr_obj for select dept_dt,eno,loc_dt from dept_loc_tb_bk1;
    9  fetch cr_obj bulk collect into lv_dept_loc_tb;
   10  close cr_obj;
   11  end;
   12  /
  declare
  *
  ERROR at line 1:
  ORA-06504: PL/SQL: Return types of Result Set variables or query do not match
  ORA-06512: at line 9
  
  SQL> declare
    2  type cr is ref cursor;
    3  cr_obj cr;
    4
    5  lv_dept_loc_tb dept_loc_tb;
    6
    7  begin
    8  open cr_obj for select DEPT_LOC_REC1(dept_dt,eno,loc_dt) from dept_loc_tb_bk1;
    9  fetch cr_obj bulk collect into lv_dept_loc_tb;
   10  close cr_obj;
   11  end;
   12  /
  
  PL/SQL procedure successfully completed.
  
  SQL> 
  

  SY.
  P.S. discover sys_refcursor.

• How to clear the Ref Cursor runtime error

  Hello everyone,
  the code as follows

  create or replace 
  procedure Country_sel(key in varchar2)
  as
  cc Res_RelcountryLan.countrycode%type;
  len Res_Language.langname_en%type;
  lid Res_Language.langid%type;
  ab Res_Language.Abrivation%type;
  type refcursorr is ref cursor;
  cur refcursorr;
  d_stmt varchar2(100);
  begin
  d_stmt := 'select RCL.countrycode,RL.langid,RL.langname_'||key||',
  RL.Abrivation from  Res_RelCountryLan RCL inner join Res_Language RL ON RCL.LangId = RL.LangId';
  open cur for d_stmt;
  loop
  fetch cur into cc,lid,len,ab;
  if cur%found then
  dbms_output.put_line(cc||lid||len||ab);
  else
  exit;
  end if;
  end loop;
  close cur;
  commit;
  end  Country_sel;

  When I run this code im getting

  ORA-06502: PL/SQL: numeric or value error: character string buffer too small
  ORA-06512: at "RASOOL.COUNTRY_SEL", line 11
  ORA-06512: at line 6

  can you please help me get rid of this problem.
  
  
  thanking you,
  Prakash

  
  d_stmt varchar2(100); 
  

  Increase the size of d_stmt. Your an a larger string assignment

  
  d_stmt := 'select RCL.countrycode,RL.langid,RL.langname_'||key||',RL.Abrivation from  Res_RelCountryLan RCL inner join Res_Language RL ON RCL.LangId = RL.LangId'; 
  

  The size of the string above is more than 100 characters.

• How to create control provided by PL/SQL ref-cursor

  Dear valuable experts,
  
  I'm migrating oracle forms10g for JDeveloper. Currently I need a guideline or a reference to create pages of the ADF using the PL/SQL-Refcursor rather then tables or views.
  
  I need help also how can I create using PL/SQL (package) - Refcursor datacontrol.
  
  Please help me.
  
  (!)
  Thank you.
  Zed
  ====

  Zed,

  The Fusion developer's Guide has 35.9.4 article titled "How to create a view object on a REF CURSOR" which should help you.

  John

• Zero error of iteration - the treatment of dynamic sql statements in dbms_xmlgen

  Hello
  
  I have a procedure that creates a dynamic sql v_sql
  
  cursor v_curr is
  Select *.
  of btctl_msg_log;
  
  BEGIN
  Select count (*) in the v_cnt of btctl_msg_log;
  
  IF v_cnt > 0 THEN
  C1 in v_curr
  LOOP
  v_sql: = "' | ' SELECT * from '. C1.msg_rcrd_src_tbl_nm | |' where rowid = ' | " ' ||'' ' || C1.msg_rcrd_src_tbl_id | " ' ||'' ' ||'' ' ;
  Select DBMS_XMLGEN.getXMLtype (v_sql) in the double v_xml;
  
  gives me an error
  
  ORA-19202: an error has occurred in the processing of XML
  ORA-24333: zero number of iterations
  ORA-06512: at "SYS." DBMS_XMLGEN", line 288
  ORA-06512: at line 1
  
  I don't know why is this error happening.
  Any help much appreciated.

  881575 wrote:
  Hello

  I have a procedure that creates a dynamic sql v_sql

  cursor v_curr is
  Select *.
  of btctl_msg_log;

  BEGIN
  Select count (*) in the v_cnt of btctl_msg_log;

  IF v_cnt > 0 THEN
  C1 in v_curr
  LOOP
  v_sql: = "' | ' SELECT * from '. C1.msg_rcrd_src_tbl_nm | |' where rowid = ' | " ' ||'' ' || C1.msg_rcrd_src_tbl_id | " ' ||'' ' ||'' ' ;
  Select DBMS_XMLGEN.getXMLtype (v_sql) in the double v_xml;

  gives me an error

  ORA-19202: an error has occurred in the processing of XML
  ORA-24333: zero number of iterations
  ORA-06512: at "SYS." DBMS_XMLGEN", line 288
  ORA-06512: at line 1

  I don't know why is this error happening.
  Any help much appreciated.

  Standard when boards (ab) use of EXECUTE IMMEDIATE is to compose the SQL statement in a single VARCHAR2 variable
  Then print the variable before passing to EXECUTE IMMEDIATE.
  COPY the statement & PASTE in sqlplus to validate its correctness.

• the variable name of database query to SQL Server using the Oracle database link

  Hi all
  
  I have an ApEx 4.1 application running on x 64 (11.2.0.1) 11g on Windows Server 2008 x 64, and I have a few points of data integration with SQL (2005 and 2008) server that I need to create. I have configured the database with dg4odbc link and it works perfectly... I can run queries on the SQL Server database without any problem using the database link.
  
  However, there is a scenario where the SQL Server database name is dynamic, and I need to generate on the fly in a PL/SQL block and then use it in a query dynamic SQL (all this in the ApEx). It of wherever I meet problems... when I asked the default database that is based on the ODBC connection and I don't have to specify the name of the database, no problem. But when I need access to one of the several other databases by default, I received the error "invalid table.
  
  It works well:* (note that 'fv' is the name of my database link)
  
  v_query1: = "select 'Release Date' from dbo." Schedules@FV where dbo. Annexes. "" SchedID "=: calendar";
  EXECUTE IMMEDIATE v_query1 in rel_date using the grid.
  
  
  
  
  I then take this rel_date variable, convert a varchar2 (rel_date_char), then use it as the name of the database in the following query...
  
  
  _ It returns an error(error ORA-00903: invalid table name)
  
  v_query2: = "select"PARTNO": rel_date_char.dbo.ProdDetails@fv where 'SchedID' =: calendar and"UnitID"=: unit"
  and 'MasterKey' =: master and "ParentKey" =: parent';
  
  EXECUTE IMMEDIATE v_query2 in part_number using planning, master, parent unit;
  
  
  
  I also tried using all of the following conditions without result:
  
  "select"PARTNO"of" | : rel_date_char | '.dbo. ProdDetails@fv where 'SchedID '...
  "select"PARTNO"of" | rel_date_char | '.dbo. ProdDetails@fv where 'SchedID '...
  "select"PARTNO"of" | @rel_date_char | '.dbo. ProdDetails@fv where 'SchedID '...
  "select"PARTNO"in @rel_date_char.dbo.ProdDetails @fv where 'SchedID'..."
  
  
  Is it possible to do it in PL/SQL?
  
  Thanks for any help!
  -Ian C.
  
  Published by: 946532 on July 15, 2012 19:45

  Just did a test using passthrough:

  SQL > set serveroutput on
  SQL > declare
  2 val varchar2 (100);
  3 c whole;
  4 whole nr;
  5. start
  c: 6 = dbms_hs_passthrough.open_cursor@FREETDS_DG4ODBC_EMGTW_11_2_0_3;
  7 dbms_hs_passthrough.parse@FREETDS_DG4ODBC_EMGTW_11_2_0_3 (c, "select count (*) from EMP");
  8 LOOP
  9 nr: = DBMS_Hs_Passthrough.fetch_row@FREETDS_DG4ODBC_EMGTW_11_2_0_3 (c);
  10 output when nr = 0;
  11 dbms_hs_passthrough.get_value@FREETDS_DG4ODBC_EMGTW_11_2_0_3 (c, 1, val);
  12 dbms_output.put_line (val);
  13 end of loop;
  14 dbms_hs_passthrough.close_cursor@FREETDS_DG4ODBC_EMGTW_11_2_0_3 (c);
  15 end;
  16.
  24576

  PL/SQL procedure successfully completed.

  SQL > declare
  2 val varchar2 (100);
  3 c whole;
  4 whole nr;
  5. start
  c: 6 = dbms_hs_passthrough.open_cursor@FREETDS_DG4ODBC_EMGTW_11_2_0_3;
  7 dbms_hs_passthrough.parse@FREETDS_DG4ODBC_EMGTW_11_2_0_3 (c, ' select count (*) from dbo.) EMP');
  8 LOOP
  9 nr: = DBMS_Hs_Passthrough.fetch_row@FREETDS_DG4ODBC_EMGTW_11_2_0_3 (c);
  10 output when nr = 0;
  11 dbms_hs_passthrough.get_value@FREETDS_DG4ODBC_EMGTW_11_2_0_3 (c, 1, val);
  12 dbms_output.put_line (val);
  13 end of loop;
  14 dbms_hs_passthrough.close_cursor@FREETDS_DG4ODBC_EMGTW_11_2_0_3 (c);
  15 end;
  16.
  24576

  PL/SQL procedure successfully completed.

  So all 3 ways work for me.

  Published by: kgronau on July 23, 2012 10:08

  Now, using the variables to make the selection:

  SQL > declare
  2 val varchar2 (100);
  3 c whole;
  4 whole nr;
  5 tabname varchar2 (20): = 'EMP ';
  6 ownr varchar2 (20): = "dbo."
  7 dbname varchar2 (20): = "door";
  Start 8
  c: 9 = dbms_hs_passthrough.open_cursor@FREETDS_DG4ODBC_EMGTW_11_2_0_3;
  10 dbms_hs_passthrough.parse@FREETDS_DG4ODBC_EMGTW_11_2_0_3 (c, ' SELECT count (*) FROM ': dbname: '.) ' || ownr | '.'|| tabname | ") ;
  11 LOOP
  12 nr: = DBMS_Hs_Passthrough.fetch_row@FREETDS_DG4ODBC_EMGTW_11_2_0_3 (c);
  13 when the exit nr = 0;
  14 dbms_hs_passthrough.get_value@FREETDS_DG4ODBC_EMGTW_11_2_0_3 (c, 1, val);
  15 dbms_output.put_line (val);
  16 end loop;
  17 dbms_hs_passthrough.close_cursor@FREETDS_DG4ODBC_EMGTW_11_2_0_3 (c);
  18 end;
  19.
  24576

  PL/SQL procedure successfully completed.

  => instead of executing the statement using the "execute Immediate" we use the PASTHROUGH package to pass the statement to SQL Server.

  Published by: kgronau on July 23, 2012 10:10

• Pipelined table vs ref cursor in a function return

  Hi gurus,
  
  Everybody has discovered that a (subject) is faster on the other? Data will be primarily consumed from an external application (.net). What are the benefits? I can't decide if that use is
  
  Thank you very much.

  user12868294 wrote:
  Hi gurus,

  Everybody has discovered that a (subject) is faster on the other? Data will be primarily consumed from an external application (.net). What are the benefits? I can't decide if that use is

  Thank you very much.

  They are two different things.

  A pipeline table acts as an array, but you must always choose in it and so if your consumption that in .net, you would still use a Ref Cursor I guess to query this table in pipeline (I guess .net is not query the tables directly, but must use some sort of slider Ref?)

  Tables in pipeline can be fast, but it depends on what you need to. Is there a reason why you really need a feature in pipeline? If this is not the case, just use a normal query with a Ref Cursor, so your .net application only retrieves the data properly.

• Dynamic SQL with a Ref Cursor

  Hello
  
  I have a package that returns a Ref Cursor, in this procedure, I have a dynamic sql code that is built according to certain values, and the query is a select query, is it possible that I can put that dynamic sql in the ref cursor and return of the procedure.
  
  Or y at - it no alternative better workaround.
  
  Thanks in advance.
  
  Naveen

  Yes you can.

  Try this...

  create or replace package test_pack is
  type ref_cur is ref cursor;
  procedure just_print(ref_var ref_cur);
  end;
  /
  
  create or replace package body test_pack is
  procedure just_print(ref_var ref_cur) is
  l_var emp%rowtype;
  begin
  loop
  fetch ref_var into l_var;
  exit when ref_var%notfound;
  dbms_output.put_line(l_var.ename);
  end loop;
  end;
  end;
  /
  
  declare
  cur_var test_pack.ref_cur;
  dsql varchar2(100);
  begin
  dsql := 'select * from emp where deptno=10';
  open cur_var for dsql;
  test_pack.just_print(cur_var);
  end;
  /
  CLARK
  KING
  MILLER
  
  PL/SQL procedure successfully completed.
  
  Elapsed: 00:00:00.00
  

  Kind regards
  Prazy

• REF cursor in sql dynamic help to run immediately

  Hello
  
  How can we get the Ref cursor out a dynamic sql statement by executing immediate proceedings
  for example, immediately run ' open CROR for select * from dynamicTable'
  
  in this area, CROR is a dynamic cursor and table name is dynamic (known at run time), and we can't write static sql statement.
  
  Thank you

  I don't know what exactly you are after but here is a sample of what can be done.

  SQL> VAR r REFCURSOR;
  SQL> BEGIN
    2          OPEN :r FOR 'SELECT * FROM DUAL';
    3  END;
    4  /
  
  PL/SQL procedure successfully completed.
  
  SQL> PRINT r
  
  D
  -
  X
  

  You can use the OPEN... FOR education with a dynamic string.

• 2 running, 1 inside and 1 sql external, dynamic queries using loops and ref Cursor

  Hi all
  
  I'm under Oracle 10.2.0.2. I'm currently building dynamic sql using external and internal queries ref Cursor. Queries make use of user selected lists of ID, which is stored in a table in memory at run time. My example is as follows. From a list of book_publication_id on a web page, the user selects several books. The same user selects then several maps from a list of map_publication_id on the same web page. A book contains several maps and some of these cards will appear in more than one book (aka many many relationships).
  
  This should then ask the following output to a new web page:

  Outer loop:  Display book details for book_publication_id 230
    Inner loop:  Display 1st map details for map_publication_id 340
    Inner loop:  Display 2nd map details for map_publication_id 346
    Inner loop:  Display 3rd map details for map_publication_id 350
  Outer loop:  Display book details for book_publication_id 240
    Inner loop:  Display 1st map details for map_publication_id 346
    Inner loop:  Display 2nd map details for map_publication_id 375
  Outer loop:  Display book details for id 255
  and so on.

  In the example above, the outer loop displays the details of the book for book_publication_id 230 and the inner loop displays all the users selected the maps in book_publication_id 230. Then, it moves on the details of the book for book_publication_id 240 and done the same thing again. Similar in some ways to how break would work in SQL * Plus, even if the table is built like an HTML, the book details must be on a separate line for the card details. However I don't know how I would want it to work.
  
  So far, using loops, I could not show all the details of card for each different book_publication_id, after the details of the book for only 1 book_publication_id are displayed each time. If a loop in a loop. I hope that makes sense. I think this is my internal request that it be built differently. My code for this part of the program, which is currently wrong, is as follows:

  --Global variable section contains:
  var_user_chosen_map_list_ids VARCHAR2(32767);
  var_details VARCHAR2(32767);
  ......
  PROCEDURE PROCMAPSEARCH (par_user_chosen_map_list_ids PKG_ARR_MAPS.ARR_MAP_LIST)
  IS
  BEGIN
  FOR rec_user_chosen_map_list_ids IN 1 .. par_user_chosen_map_list_ids.count
  LOOP
     var_user_chosen_map_list_ids := var_user_chosen_map_list_ids || 
     '''' || 
     par_user_chosen_map_list_ids(rec_user_chosen_map_list_ids) || 
     ''',' ;
  END LOOP;
   var_user_chosen_map_list_ids := substr(var_user_chosen_map_list_ids,
                                          1, 
                                          length(var_user_chosen_map_list_ids)-1);
  var_details := FUNCMAPDATAFIND (var_user_chosen_map_list_ids);
  htp.print(var_details);
  END PROCMAPSEARCH;
  FUNCTION FUNCMAPDETAILS (par_user_chosen_map_list_ids IN VARCHAR2(32767)
  RETURN VARCHAR2
  AS
  TYPE cur_type_map IS REF CURSOR;
  cur_book_search cur_type_map;
  var_book_date NUMBER(4);
  var_book_title VARCHAR2(32767);
  cur_map_search cur_type_map;
  var_map_date NUMBER(4);
  var_map_title VARCHAR2(32767);
  begin:
  OPEN cur_book_search FOR
  'SELECT BI.book_date,
          BT.book_title
   FROM   BOOK_INFO BI,
          BOOK_TITLE BT,
          TABLE (sys.dbms_debug_vc2coll(' || par_user_chosen_book_list_ids || ')) BL_1
   WHERE BI.book_title_id = BT.book_title_id
   AND BI.book_publication_id = BL_1.column_value';
  OPEN cur_map_search FOR
  'SELECT MI.map_date,
         MT.map_title
  FROM map_info MI,
       map_title MT,
       TABLE (sys.dbms_debug_vc2coll(' || par_user_chosen_book_list_ids || ')) BL_2
       TABLE (sys.dbms_debug_vc2coll(' || par_user_chosen_map_list_ids || ')) ML
  WHERE MI.map_title_id = MT.map_title_id
  AND BI.book_publication_id = BL_2.column_value
  AND BI.book_publication_id = MI.pub_publication_id
  AND MI.map_publication_id = ML.column_value';
  LOOP
  LOOP
  FETCH cur_map_compare INTO
  var_book_date,
  var_book_title;
  var_details
  var_details := var_details || 'Book date: '||
                      var_book_date ||
                      'Book title: ' ||
                      var_book_title;
  FETCH cur_map_compare INTO
  var_map_date,
  var_map_title;
  var_details := var_details || 'Map date: '||
                      var_map_date ||
                      'Map title: ' ||
                      var_map_title
  EXIT WHEN cur_book_compare%NOTFOUND;
  END LOOP;
  EXIT WHEN cur_map_compare%NOTFOUND;
  END LOOP;
  RETURN var_details;
  END FUNCMAPDETAILS;

  If anyone has any ideas or suggestions, I would be grateful. It is an extension of my previous code, I posted a question recently. As I am working and learning a step at a time, I left this idea in my previous question, that I had to make sure I knew that first.
  
  Kind regards
  
  Tim

  Using dynamic SQL

  declare
    dd sys_refcursor;
    ee sys_refcursor;
    d dept%rowtype ;
    e emp%rowtype ;
  begin
  open dd for 'select * from dept' ;
  loop
  fetch dd into d ;
  exit when dd%notfound ;
  dbms_output.put_line('Department:'||d.dname);
    open ee for 'select * from emp where deptno='||d.deptno ;
    loop
    fetch ee into e ;
    exit when ee%notfound ;
    dbms_output.put_line('..Employee:'||e.empno||':'||e.ename);
    end loop;
  end loop ;
  end ;
  /
  Department:ACCOUNTING
  ..Employee:7782:CLARK
  ..Employee:7839:KING
  ..Employee:7934:MILLER
  Department:RESEARCH
  ..Employee:7369:SMITH
  ..Employee:7566:JONES
  ..Employee:7788:SCOTT
  ..Employee:7876:ADAMS
  ..Employee:7902:FORD
  Department:SALES
  ..Employee:7499:ALLEN
  ..Employee:7521:WARD
  ..Employee:7654:MARTIN
  ..Employee:7698:BLAKE
  ..Employee:7844:TURNER
  ..Employee:7900:JAMES
  Department:OPERATIONS
  
  PL/SQL procedure successfully completed.
  

  HTH

  SS

• REF CURSOR ERROR

  I get the following error when you try to select using a ref cursor

  SQL error: ORA-06504: PL/SQL: return variables of the game results or the query types do not match

  Here are the steps that I have followeed. Can someone tell me what I did wrong.

  -CREATE AN OBJECT

  create or replace type rsn_rec as object
    (
   rsn_cd   varchar2(10),
   rsn_desc  varchar2(30)
  );
  

  -CREATES A TABLE OF OBJECT

  create or replace type rsn_tbl as table of rsn_rec;
  

  PACKAGE/CREATED FUNCTION

  create or replace package pkg_rept_test as
    type refcur is ref cursor;
    function get_rsn return rsn_tbl pipelined;
  end pkg_rept_test;
  create or replace package body pkg_rept_test as
    function get_rsn
      return rsn_tbl pipelined is
        o_cursor refcur;
        rec rsn%ROWTYPE;
      begin
        open o_cursor for 
         select 
          rsn_cd,         
          rsn_desc       
         from rsn;
        loop
          fetch o_cursor into rec;
          exit when (o_cursor%notfound);
          pipe row(rsn_rec(rec.rsn_cd,                                                                                                   rec.rsn_desc)
                           );
        end loop;
        return;
    end;
  end pkg_rept_test;
  

  -THE TABLE SELECTION

  select * from table(pkg_rept_test.get_rsn)
  

  -ERROR
  SQL error: ORA-06504: PL/SQL: return variables of the game results or the query types do not match

  Your code works for me just as you posted.

  Connected to:

  Oracle Database 11 g Enterprise Edition Release 11.2.0.1.0 - 64 bit Production

  With partitioning, OLAP, Data Mining and Real Application Testing options

  SQL > create table rsn (rsn_cd varchar2 (10), rsn_desc varchar2 (30));

  Table created.

  SQL > insert into rsn values ("cd test", "test description");

  1 line of creation.

  SQL > commit;

  Validation complete.

  SQL > create or replace the rsn_rec as an object type

  2        (

  3 rsn_cd varchar2 (10),

  rsn_desc 4 varchar2 (30)

  5      );

  6.

  Type of creation.

  SQL > create or replace type rsn_tbl in the rsn_rec table;

  2.

  Type of creation.

  SQL > create or replace package pkg_rept_test as

  2 type refcur is ref cursor;

  3 function get_rsn return rsn_tbl in pipeline;

  4 end pkg_rept_test;

  5.

  Package created.

  SQL > create or replace package body pkg_rept_test as

  2 function get_rsn

  3 return rsn_tbl pipeline is

  4 o_cursor refcur.

  5 rec rsn % ROWTYPE;

  6 start

  7. open o_cursor for

  8. Select

  rsn_cd 9,.

  10 rsn_desc

  11 of rsn;

  12 loop

  13 extract o_cursor in rec;

  When exit 14 (o_cursor % notfound);

  line 15 pipe (rsn_rec (rec.rsn_cd, rec.rsn_desc)

  16                               );

  17 end of loop;

  18 return;

  end 19;

  20 end pkg_rept_test;

  21.

  Package body created.

  SQL > select * from table (pkg_rept_test.get_rsn);

  RSN_CD RSN_DESC

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

  tests of cd test description

• Irregular data loss - function from PL/SQL returning data using Ref Cursor

  Database Version: 10.2.0.4.0 (node 2 CARS)
  
  The high-level process flow is as below:
  (1) insert records in a few tables & commit the same
  (2) call the pl/sql function to extract files (on certain conditions with joins with other tables) of the tables which are filled in step 1.
  -> It uses the ORDER BY clause to queries inline & line number 5000 records return for each call.
  Sense - if inline query is supposed to return 1,00,000 records then 20 calls to the same function. This, because the application cannot contain records beyond number.
  (3) the data returned by the ref cursor is then processed by application (Tibco BW) to generate the flat file.
  
  We are facing the problem of data loss in the file and there is no fixed model. It happens once between 200-300 calls process.
  Resolution: When the problem occurs, triggering the process and in almost every time re-outbreak of the process provides required data.
  
  Guidance on what could be the reason?
  
  * Examples of Code for the function:
  CREATE OR REPLACE FUNCTION FUNC_GET_HRCH_TOTAL_DATA)
  outinstrid in NUMBERS
  outinstrkey in NUMBERS
  rownumberstart in NUMBERS
  rownumbereend in NUMBERS
  err_code OUT VARCHAR2,
  err_msg OUT VARCHAR2)
  RETURN PACK_TYPES. HRCH_TOTAL_CURSOR
  IS
  REF_HRCH_TOTAL_CURSOR PACK_TYPES. HRCH_TOTAL_CURSOR;
  BEGIN
  
  OPEN FOR REF_HRCH_TOTAL_CURSOR
  SELECT *.
  FROM (SELECT A.HIERARCHY_KEY, B.KEY, B.VAL_KEY, A.KEY_NEW, C.ITEMID, B.VAL_TAG, B.sort_order, ROWNUM ROWNUMBER
  OF AOD_HRCH_ITEM A, AOD_HRCH_ATTR B, AOD_HRCH_ITEMS C
  WHERE A.outputid = B.outputid
  AND A.outputid = C.outputid AND A.outputkey = B.outputkey
  AND A.outputkey = C.outputkey AND A.outputid = outinstrid
  AND A.outputkey = outinstrkey AND A.ITEM_SEQ = B.ITEM_SEQ
  AND A.ITEM_SEQ = C.ITEM_SEQ AND A.HIERARCHY_LEVEL_ORDER = B.SORT_ORDER
  ORDER BY A.HIERARCHY_LEVEL_ORDER DESC)
  WHERE ROWNUMBER < rownumbereend
  AND ROWNUMBER > = rownumberstart;
  
  
  RETURN REF_HRCH_TOTAL_CURSOR;
  EXCEPTION
  WHILE OTHERS
  THEN
  err_code: = x_progress | ' - ' || SQLCODE;
  err_msg: = SUBSTR (SQLERRM, 1, 500);
  
  END FUNC_GET_HRCH_TOTAL_DATA;
  /
  
  Published by: meet_sanc on February 16, 2013 10:42

  Your SELECT statement is almost certainly incorrect

  SELECT *
    FROM ( SELECT A.HIERARCHY_KEY, B.KEY, B.VAL_KEY, A.KEY_NEW, C.ITEMID, B.VAL_TAG, B.sort_order,ROWNUM ROWNUMBER
             FROM AOD_HRCH_ITEM A, AOD_HRCH_ATTR B, AOD_HRCH_ITEMS C
            WHERE A.outputid = B.outputid
              AND A.outputid = C.outputid AND A.outputkey = B.outputkey
              AND A.outputkey = C.outputkey AND A.outputid = outinstrid
              AND A.outputkey = outinstrkey AND A.ITEM_SEQ = B.ITEM_SEQ
              AND A.ITEM_SEQ = C.ITEM_SEQ AND A.HIERARCHY_LEVEL_ORDER = B.SORT_ORDER
            ORDER BY A.HIERARCHY_LEVEL_ORDER DESC)
   WHERE ROWNUMBER < rownumbereend
     AND ROWNUMBER >= rownumberstart;
  

  Since the ORDER BY is applied after the ROWNUM is assigned in this case, your query is requested for a period of 5000 lines any arbitrariness. It would be perfectly valid for a single line to return in each of your 200 different calls or for a line to return in any of them.

  You definitely want to do something in the sense of the canonical askTom wire

  select *
    from ( select a.*, rownum rnum
             from ( YOUR_QUERY_GOES_HERE -- including the order by ) a
            where rownum <= MAX_ROWS )
   where rnum >= MIN_ROWS
  

  That said, it seems inconceivable that Tibco is unable to manage a cursor that returns more than a certain number of lines. You do a ton of work to return the data pages that are certainly not necessary. Unless you're saying that you somehow paralyzed your installation of Tibco giving him a ridiculously small amount of memory to process, something doesn't look good. A slider is just a pointer - it holds that no data - so the number of lines that you can extract a slider should have no impact on the amount of memory on the client application needs.

  As others have already pointed out, your exception handler is almost certainly do more harm than good. Return the error codes and error messages as from the OUT parameters, instead of simply allowing the exception to propagate deletes a ton of useful information (such as the mistake of the stack) and makes your process much less robust.

  Justin

• How to perform procedures with REF CURSOR &amp; PLSQL settings table in SQL Dev?

  I have 3 SQL Developer. I create a procedure with a set of tables/PL/SQL as an input parameter and a SYS_REFCURSOR as output parameter.
  
  When I press the green arrow (Run button) block the dialog box run PL/SQL PL/SQL has no place at the PL/SQL table with records of entry and also to see the output of the SYS_REFCURSOR.
  
  I get the error ORA-06550 and many other errors when I run the present.
  
  How to do this please?

  In the debug code, lose the ABS. Prefix the SYS_REFCURSOR QC_SQL_DEVELOPER_TEST.
  Also, do not use same name of your settings on your bind variable (so e.g.: OUT_REF_CURSOR).
  REF CURSOR is supported output types, so you will get the output of their share in the output tab.
  However the output system does not recognize your own types as your PL/SQL table. You need to write an exit procedure yourself, indexing correctly, using the DBMS_OUTPUT.

  Have fun
  K.

• Error when you work with Ref Cursor

  Hi, I tried the following, but the err

  DECLARE
   TYPE ref_nm IS REF CURSOR;
   vref REF_NM;
   vemp emp%rowtype;
  BEGIN
   OPEN vref FOR SELECT ename ,sal FROM EMP;
   LOOP
    FETCH vref INTO vemp;
    EXIT WHEN vref%NOTFOUND;
      DBMS_OUTPUT.PUT_LINE ( vemp.ename ||','||vemp.sal ); 
   END LOOP;
    CLOSE vref;
  END;

  Error is
  

  ORA-06504: PL/SQL: Return types of Result Set variables or query do not match 

  Use you this structure as a buffer of extraction:
  PEMV emp % rowtype;

  This structure contains the whole line of EMP - all columns.

  That's what you're looking for cursor - 2 columns (not the whole line):
  Vref OPEN for SELECT ename, sal of EMP;

  You can not expect from PL/SQL to find out how to move the 2 column values in a structure that has more than just 2 columns. The error message is quite clear about this - read and think what is transport of the error.