Manipulation of PL/SQL, an overload of functions and NULL

Similar Questions

• SQL loader by using functions and position

  Hi all, I need help, loading data in my table using sql loader. study the following

  CREATE TABLE er
  (                     
  a1 NUMBER,
  a2 number,
  a3 VARCHAR2(100),
  a4 VARCHAR2(100),
  a5 VARCHAR2(100),
  a6  VARCHAR2(100),
  a7  VARCHAR2(100),
  a8  VARCHAR2(100)
  );
  
  OPTIONS (BINDSIZE=20548000, READSIZE=20548000, STREAMSIZE=20548000, DATE_CACHE=25000,  SKIP=0)
  LOAD DATA
  INTO TABLE er
  APPEND     
  TRAILING NULLCOLS
  (
    a1            POSITION(0001:0021)               ,
    a2            POSITION(0022:0042)       "DECODE(SUBSTR(:a2,1,3),'***',NULL,:a2)"      ,
    a3            POSITION(0043:0053)       ,
    a4            POSITION(0054:0064)          ,
    a5            POSITION(0065:0075)           ,
    a6            POSITION(0076:0086)       ,
    a7            POSITION(0087:0093)      "DECODE(SUBSTR(:a7,1,3),'***',NULL,:a7)"   
  )
  BEGIN
                   0.00 ******************** X          X          X          *X          ****
  END;

  If you look at the data, some fields have a lot of * and some af a bit such as *. I want to load data into a table, and when a field contains all * as the value, I want to set to null. If a field contains a * and alphanumeric, so this value should be as it is.
  
  in the above example, * must be set to null and * should also be defined with a null value. Note that there is a field with x as this area contain alpha numeric, it must be loaded into the table, as is.  the only time in field must be set to null, it's when the value contains all .
  
  someone in this forum suggest using to decode, but looks like it doesn't work and I get error when it reads the second field and try to insert into the column number a2.
  
  is it possible to use regular expressions to find out if a field contains all *. I also want to cut each field as they may contain spaces.
  can be a help for this, use the ctl sqlloader and above?

  You can include regular expressions in you SQL * file the charger control.

  An example can be found here:
  http://www.morganslibrary.org/reference/sqlloader.html
  Demos of 7 and 8 by using the UPPER and DECODE functions to illustrate how do.

• What is the best way to optimize a SQL query: call a function or doing a join?

  Hi, I want to know what is the best way to optimize a SQL query, call a function inside the SELECT statement, or make a simple join?

  It depends on.  Could be a.  Could be the other.  Could be no difference.  You would need to compare with your tables in your environment with your settings.

  If you put a gun to my head, I was given no other information and required that I answered the question, I would tend to wait that the join would be more effective.  In general, if you can do something in pure SQL, it will be more effective than if you call PL/SQL.

  Justin

• Expert SQL 1Z0-047: all functions are examined or just groups?

  Review of press of Oracle book says that all functions can be included in the review (Chapter 6 use of the box shows a row functions p.234 review)
  and reffer the manual reference
  
  And as I see the book and review group (Aggregate), specific einreihig, date on which the ordinary functions of Expressions and page are reviewed.
  

  With the help of line features unique to customize the output
  
  Describe the different types of functions that are available in SQL
  Describe the use of the conversion functions
  Use of character, number, date functions in SELECT statements
  -Describe the use of functions of Group
  Identify the Group functions available
  Use the group for aggregate functions
  -Use datetime functions
  -Regular expressions

  of the foregoing, I guess that the bellows are out of the review
  XML
  Data mining
  
  So what really are the functions examined gtoups?

  So what really are the functions examined gtoups?

  Nowhere (or at least anywhere outside the Oracle education) is a list of all of the specific functions that will be under review. Besides, "will be" more properly would be written "could be" since all takers single review never sees all possible questions. A given function may be on the test for some people and not others.

  When deciding on what features to include, what I've done (and I suppose that Steve did) was to look through the different functions of a specific type (i.e. conversion functions) and the weight of each based on how the function is common in my experience. I have included these functions which I consider to be seen more often and so that no matter what SQL expert expected to know on top of their head. That said, the review Expert SQL takes pleasure in the dark. I've seen this most often in variations of syntax, but it is possible that review manufacturers choose a function to include simply because it is rarely used. While I don't remember all these appearing on the exam that I took, I can't exclude that possibility and so my guide at the end of the functions section States it much more listed there in the SQL Reference Guide which peuvent appear in the review.

  That said, learn the ins and outs of more functions that you might really need to review is not a bad thing. Because you indicated that you are interested in the real expertise rather than the skills of paper, limit you just to the focus of the review is not intuitive. I guarantee you there are a lot of things I know about SQL for years to work in the language that are not on the subjects of examination.

• Any difference between the separate function and aggregation in sql query cost?

  Hello
  I've run many models sql stmts - such as:
  one) using a single table
  (b) using the two tables, using simple joins or outer joins
  
  but I didn't notice a difference in sql stmts in cost and implementation plan...
  In any case, my colleague insists on only using the aggregate function is less expensive compared to
  separate... (something I've not confirmed, that is why I believe that they are exactly the same...)
  
  For the situation described above under 1 sql... We could for example use
  

  select distinct deptno
  
   from emp

  select count(*), deptno
  
   from emp
  
   group by deptno

  select distinct owner, object_type from all_objects

  select count(*), owner, object_type from all_objects
  
  group by owner, object_type

  Have you ever found no difference between the two...?
  
  Note: I use Ora DB 10g v2.
  
  Thank you
  SIM

  Alexandre Gelin says:
  function separate and aggregation are for different uses and can give the same result, but if you use the aggregate function to get the separate records, it will be expensive...

  really? I find them almost identical to 11g. Actually GROUP BY is almost always more fast for me in the old days.

  article about this askTom Re: any difference between the separate function and aggregation in sql query cost?
  This applies mainly to the older oracle versions http://www.oracle.com/technology/products/rdb/pdf/distinct_derivedtables_groupby_sort_tech_article.pdf

• Problem inserting new line in a VO with SQL functions and subqueries

  Hello
  
  I use JDeveloper 10.1.3.4 and and stuck with insertion of a new line when the values to insert into some fields of the new line are returns of SQL functions and subqueries.
  
  At the sqlplus prompt, the following sql statement is tested and developed successfully:
  

  insert into VALIDATIONS values (
     '111223333', 'JANE', 'DOE',
     SYSDATE,
     (select to_char(sysdate, 'yymmdd') from dual)||
     (select to_char(count(TRANSACTION_DATE)+1,'fm0999') from VALIDATIONS
        where trunc(TRANSACTION_DATE) = to_char(sysdate))
  );

  Note that the values for the fields of 4th and 5th are returns of subqueries and SQL functions.
  
  Ideally, it would be simple to assemble a SQL and executed him as he did to the old way of servlet:
  

  String sql = "insert into VALIDATIONS values ('" + id + "', '" + firstName + "', '" + lastName + "', SYSDATE, (select to_char( ......)))";
  statement.executeUpdate(sql);

  I'm not sure it can be done in JDeveloper, and it was not found in the developer's guide. What I found was rather:
  

  // 1. Find the ServiceRequests view object instance.
  ViewObject svcReqs = am.findViewObject("ServiceRequests");
  // 2. Create a new row and insert it into the row set
  Row newSvcReq = svcReqs.createRow();
  svcReqs.insertRow(newSvcReq);
  // 3. Show effect of entity object defaulting for Status attribute
  System.out.println("Status defaults to: "+newSvcReq.getAttribute("Status"));
  // 4. Set values for some of the required attributes
  newSvcReq.setAttribute("CreatedBy",308); // Nancy Greenberg (user)
  Date now = new Date(new Timestamp(System.currentTimeMillis()));
  newSvcReq.setAttribute("RequestDate",now);
  newSvcReq.setAttribute("ProdId",119); // Ice Maker
  newSvcReq.setAttribute("ProblemDescription","Cubes melt immediately");
  // 5. Commit the transaction
  am.getTransaction().commit();

  In my case, it would be OK for the first three fields as shown below (in the code of transactionsRecords is a display object). But what about the 4th and 5th fields?
  

  Row newTransaction = transactionsRecords.createRow();
  transactionsRecords.insertRow(newTransaction);
  newTransaction.setAttribute("StudentId", id);
  newTransaction.setAttribute("FirstName", firstName);
  newTransaction.setAttribute("LastName", lastName);
  
  newTransaction.setAttribute("TransactionDate", <how to put "SYSDATE" here?>);
  newTransaction.setAttribute("ConfirmNumber", <how to put subqueries here?>);

  A big thank you to those who know or have experience before!
  
  
  Newman

  Newman,
  You can do this via a procedure call. This technique is described here
  I guess you can't do this directly in the EO or VO.
  You can do it the old way to create a prepared directly query and call it as you described. To get the declaration use getDBTransaction () .createPreparedStatement (...) method available in the module of your application.
  Be aware of the implications using the given solution. The framework doesn't know anything of what you are doing. So to see the changes that you used to update the tables that you change.

  Timo

• Passage of tables through multiple functions and PL/SQL procedures

  I am now a great application of PL/SQL. There is a main procedure which is initially called who subsequently passes information to other PL/SQL functions and procedures. Ultimately an error code and the string is passed to PUT_LINE so it can be displayed. I want to be able to do is to have a table that stores an error code and a string for every error it then go through each of the procedures and functions. This would mean passing these codes and the strings of a function to a function within the pl/sql application. What would be the best way to implement this and is it possible to move heavy or folders to other PL/SQL functions? Thank you.

  Here is a simulation->

  satyaki>
  satyaki>select * from v$version;
  
  BANNER
  ----------------------------------------------------------------
  Oracle Database 10g Enterprise Edition Release 10.2.0.3.0 - Prod
  PL/SQL Release 10.2.0.3.0 - Production
  CORE    10.2.0.3.0      Production
  TNS for 32-bit Windows: Version 10.2.0.3.0 - Production
  NLSRTL Version 10.2.0.3.0 - Production
  
  Elapsed: 00:00:00.20
  satyaki>
  satyaki>
  satyaki>create or replace type n_array is table of number;
    2  /
  
  Type created.
  
  Elapsed: 00:00:07.10
  satyaki>
  satyaki>CREATE OR REPLACE PROCEDURE Get_Array(array_in IN n_array,
    2                                        array_out OUT n_array)
    3  IS
    4  BEGIN
    5    array_out := n_array();
    6    FOR i IN 1..array_in.count
    7    LOOP
    8      array_out.extend;
    9      array_out(i) := array_in(i) * 2;
   10    END LOOP;
   11  END Get_Array;
   12  /
  
  Procedure created.
  
  Elapsed: 00:00:00.89
  satyaki>
  satyaki>
  satyaki>Create or Replace Procedure Set_Array(myArray IN n_array)
    2  is
    3    i   number(10);
    4    rec emp%rowtype;
    5    w n_array:=n_array(1200,3200);
    6    bucket n_array := n_array();
    7  Begin
    8    Get_Array(w,bucket);
    9
   10    for i in 1..myArray.count
   11    loop
   12      select *
   13      into rec
   14      from emp
   15      where empno = myArray(i);
   16      dbms_output.put_line('Employee No:'||rec.empno||' Name:'||rec.ename);
   17      for j in 1..bucket.count
   18      loop
   19        dbms_output.put_line('Commission Sub Type: '||bucket(j));
   20      end loop;
   21    end loop;
   22  End Set_Array;
   23  /
  
  Procedure created.
  
  Elapsed: 00:00:01.33
  satyaki>
  satyaki>
  satyaki>select * from emp;
  
       EMPNO ENAME      JOB              MGR HIREDATE         SAL       COMM     DEPTNO
  ---------- ---------- --------- ---------- --------- ---------- ---------- ----------
        9999 SATYAKI    SLS             7698 02-NOV-08      55000       3455         10
        7777 SOURAV     SLS                  14-SEP-08      45000       3400         10
        7521 WARD       SALESMAN        7698 22-FEB-81       1250        500         30
        7566 JONES      MANAGER         7839 02-APR-81       2975                    20
        7654 MARTIN     SALESMAN        7698 28-SEP-81       1250       1400         30
        7698 BLAKE      MANAGER         7839 01-MAY-81       2850                    30
        7782 CLARK      MANAGER         7839 09-JUN-81       4450                    10
        7788 SCOTT      ANALYST         7566 19-APR-87       3000                    20
        7839 KING       PRESIDENT            17-NOV-81       7000                    10
        7844 TURNER     SALESMAN        7698 08-SEP-81       1500          0         30
        7876 ADAMS      CLERK           7788 23-MAY-87       1100                    20
  
       EMPNO ENAME      JOB              MGR HIREDATE         SAL       COMM     DEPTNO
  ---------- ---------- --------- ---------- --------- ---------- ---------- ----------
        7900 JAMES      CLERK           7698 03-DEC-81        950                    30
        7902 FORD       ANALYST         7566 03-DEC-81       3000                    20
  
  13 rows selected.
  
  Elapsed: 00:00:00.28
  satyaki>
  satyaki>declare
    2    v n_array:=n_array(9999,7777);
    3  begin
    4    Set_Array(v);
    5  end;
    6  /
  Employee No:9999 Name:SATYAKI
  Commission Sub Type: 2400
  Commission Sub Type: 6400
  Employee No:7777 Name:SOURAV
  Commission Sub Type: 2400
  Commission Sub Type: 6400
  
  PL/SQL procedure successfully completed.
  
  Elapsed: 00:00:00.15
  satyaki>
  satyaki>
  

  Kind regards.

  LOULOU.

• How to move the functions and procedures of packages in a schema?

  Hello

  I have less than the needs for a work at home and my question is if anyone can point me in the right direction to find documentation that can help me solve below. Any information will be appreciated. Thank you.

  Write procedures and functions (included or not in packets) as .txt or .sql scripts. Once they have been launched in Developer SQL they should, in addition, the functions and procedures of packages in the current schema. If the current schema contains only 2 packages, pac1 (p11, p12 procedures and functions f11, f12, f13) and pac2 (contains the p21 following procedures, p22, p23 and f21 and functions f21 (case of overload), running the scripts produce the following effects:)

  • Procedures p11, p12, p21, p22, p23, and f11, f12 and f13 functions will be created in the current schema.
  • PAC2 will contain 2 instances of the function of overload-f21; they will be kept in the package, without be created in the current schema. If all overloaded procedures/functions will be kept in their original packaging
  • If pac1 contains variables, cursors, and public types, keep us only the details of the package (and remove the body); usually if the packages contain no overloaded functions or procedures the body will be removed and if no variable, cursors, types of public then we will remove the header.
  • If in the triggers, procedures and functions we will call procedures/functions packages (procedures and functions moved into the current schema) they reference will need to be updated (via editing the body or recompilation) for example if pa1.p12 will be replaced with p12.
  • Scripts must have numbers in the next series (01... n) and the characters that explain the content.

  My only advice would be to remember that these are public forums and if YOU can find them, therefore cannot your instructors.

  In addition, those are horrible procedure names and function

  This mission seems to be around to overload the pl/sql objects and... prioritize a research on Oracle documentation around overhead and legacy
  http://docs.Oracle.com/CD/B28359_01/AppDev.111/b28371/adobjplsql.htm#i21148

• ORA-00904 during execution of the functions and procedures (IDE version 4.1.0.19)

  Recently, I updated my Oracle SQL Developer to the latest version (see below) and now when I try to run the function or procedure of the IDE, I had an alert with the error: "ORA-00904:"OBJECT_ID": invalid identifier." Code provider 904 ". Previous version of IDE works pretty well. What is the problem?

  Oracle SQL Developer 4.1.0.19

  Version 4.1.0.19

  Build a HAND - 19.07

  It seems that OBJECT_ID added to ALL_PROCEDURES for the 10g DB version, but after the version you have (10.2.0.1.0).  I don't have a link to the documentation specific to hand Oracle, but see the last comment by Steven Feuerstein in the next blog...

  PL/SQL challenge: ALL_PROCEDURES changes over version (5980)

  Typically when generation SQL for a particular version of the DB with SQL Developer, only major (10) and (more rarely) secondary numbers (2) gets checked, if your problem is one of those borderline cases that are easy to miss and less likely to get fixed. Whenever a customer has a specific need to use an older version, Oracle expects usually only the last set of patch (for example, 10.2.0.5) would be used.

• function and packages

  Hi friends,
  
  I have a requirement where in I need decode my source table values while inserting in my target table. I usually do my task by performing functions (don't use case and decode). My first question is
  
  (1) is using the function to accomplish these tasks is good. I have several of these conversion requirements.
  (2) Clubbing subprogrammes togther in packages is a good habit. But what should be the number of these subprogrammes. I can club togther 100 of this togther functions (a function of the average is 50 lines)? If this is not the case, what should be the thoughts of segregation?
  (3) are the oracle integrated like sqrt, length etc. deterministic functions?
  
  This code represents a typical requirement of the mine where in I map my name of source country in the target
  ID of country when migrating my data from the source to the target.
  
  

  create table source 
  (
  emp_id number,
  empname varchar2(100),
  sal number,
  country varchar2(100)
  );
  
  
  insert into source values (1,'ALEX',1000,'US');
  Insert into source values (2,'RAHUL',2000,'IND');
  Insert into source values (3,'KARTIK',3000,'IND');
  Insert into source values (4,'SHADOW',5000,'US');
  insert into source values(5,'LOUI',1500,'JAP');
  
  
  Create table source_country_code
  (
  country varchar2(100),
  abbrv  varchar2(100),
  area number,
  climate varchar2(100)
  );
  
  insert into source_country_code values ('UNITED STATES','US',10000,'COLD');
  insert into source_country_code values ('INDIA','IND',100000,'HOT');
  insert into source_country_code values ('JAPAN','JAP',1000000,'MODERATE');
  
  
  
  
  Create table target_country_code
  (
  country_id number,
  descr  varchar2(100),
  density number,
  weather varchar2(100)
  );
  
  insert into target_country_code values (101,'this is descrition txt 1',1001,'COLD');
  insert into target_country_code values (102,'this is descrition txt 2',1001,'HOT');
  insert into target_country_code values (103,'this is descrition txt 3',1001,'MODERATE');
  
  
  
  create table target
  (
  emp_id number,
  empname varchar2(100),
  sal number,
  country_id number
  );

  I generally functions for this kind of mappings. I have several of these requirements as indicated above and the
  cardinality of data is high.

  create or replace function transform (in_src_country varchar2) return number deterministic
  as
  l_country_id number;
  l_climate varchar2(1000);
  begin
  
  select climate into l_climate from source_country_code
     where upper(abbrv)=upper(in_src_country);
     
  select country_id into l_country_id
  from target_country_code
  where upper(weather)=upper(l_climate);
  
  return(l_country_id);
  exception 
  when no_data_found then
  return null;
  end;

  last step,
  
  so, what I usually do is
  
  insert into target
  Select emp_id, empname, sal, transform (country) in the source;
  
  
  
  As always, thank you
  
  Rahul
  
  Published by: Rahul K on 5 June 2013 10:51

  No technique is universally good or bad universally. You must understand your needs, the compromise and what you're trying to optimize for.

  If you don't care anything else you are looking for the absolute fastest performance, reduce what you have to do in PL/SQL is a good idea. Most of the people and most of the systems, however, prefer to have code that is a bit slower if it makes the code easier to maintain and easier to write or minimizes the number of places that the code is duplicated. If you attempt to load a data warehouse of company with billions of new lines every night, you're as almost exclusively worry about raw performance. And, if you design correctly, there is that one way to load any particular item of data, you don't have to worry about code duplication because there will always be an INSERT statement that inserts data in a particular table. If you build an OLTP application, on the other hand, it is very likely that there will be many places where you want to do something like search for a value and you are much less concerned about raw performance you're on maintainability and ensure that you do not duplicate the code. In these cases, well-designed functions and procedures are well worth the return miners (so if you do a query of single line as most OLTP systems are generally easier to treat if you write DSS queries that return millions of processing lines billionr lines.

  Justin

• Same algorithm in function and procedure so that one will be better?

  Why is better to pl sql function calculates a value instead of the procedure?
  If I apply the same algorithm in function and procedure while one will perform better?

  Siddharth Singh says:
  Oracle documentation

  http://docs.Oracle.com/CD/B10500_01/AppDev.920/a96624/08_subs.htm

  This:

  Generally, you use a procedure to perform an action and a function to calculate a value.

  Does not mean that the function are better to calculate a value, but you usually use functions to calculate a value.

  Usually, a function is can be more useful if you plan to use in a query.
  that is to say:

  SELECT myfunc(col1) from mytable;
  

  That you can't do with a procedure. Functions are also usually limited to the returning a single value, while procedures may have several OUTPUT parameters.

  Kind regards.
  Al

  Published by: Alberto Faenza on 21 November 2012 15:31

• How to perform a function and return the result in a variable of liaison

  Hello
  
  I'm trying to calculate the sum of the salaries of all persons with a particular using a function JOB_ID TOTAL_INCOME (v_job_id).
  
  create or replace function total_income
  + (v_job_id in VARCHAR2) +.
  Number IS BACK
  v_total number (6);
  
  cursor get_sal is
  Select the pay of employees
  where job_id = v_job_id;
  BEGIN
  v_total: = 0;
  for emp in get_sal
  loop
  v_total: = v_total emp.salary; +.
  end loop;
  
  dbms_output.put_line (' Total salary ' | v_job_id |' is: ' | v_total);
  Return v_total;
  END;
  
  Now I woud like to perform this function and assign the value returned in a variable binding test_sal
  
  test_sal variable number (6)
  SELECT total_income ('AD_VP') in: test_sal FROM DUAL;
  dbms_output.put_line (' Sal Total :'||: test_sal);
  
  This returns the errors below:
  
  SELECT total_income ('AD_VP') in: test_sal FROM DUAL
  *+
  Error on line 0
  ORA-01036: illegal variable name/number
  
  dbms_output.put_line (' Sal Total :'||: test_sal);
  Error on line 3
  ORA-00900: invalid SQL statement
  
  Could someone help me what could be the problem? Thanks for your time...

  Hi Kiran and welcome to the forum,

  You mix SQL and PL/SQL

  It's the way in which SQL:

  SQL> SELECT total_income('AD_VP') FROM DUAL;
  
  TOTAL_INCOME('AD_VP')
  ---------------------
                    123
  1 row selected.
  

  It comes to PL/SQL in SQL * more

  SQL> variable test_sal number
  SQL> execute :test_sal := total_income('AD_VP')
  PL/SQL procedure successfully completed.
  SQL> exec dbms_output.put_line('Total Sal:'||:test_sal)
  Total Sal:123
  PL/SQL procedure successfully completed
  

  Concerning
  Peter

• Doubts in functions and procedures

  Hello world
  
  Can someone tell the exact differences between function and procedure?
  
  My doubts are
  1. If the procedure can return and is not necessary it must return a value (in other words, if we want to come back and is not as one can write procedure) then why do we need to function?
  
  2. I tried to call a function that updates a table in the code, I got an error saying that the function does not use DML.
  
  3. I want to know the combination that
  procedure a. can call a function and vice versa?
  b. function (which uses DML instructions) can be called from sql statement?
  and if all other combinations are there let me know with reasons.
  
  Thanks in advance,
  Vinay

  Have you looked at in the [reference Guide and the User Guide for PL/SQL | http://download.oracle.com/docs/cd/B19306_01/appdev.102/b14261/toc.htm] already?
  http://download.Oracle.com/docs/CD/B19306_01/AppDev.102/b14261/subprograms.htm#i4075
  http://download.Oracle.com/docs/CD/B19306_01/AppDev.102/b14261/subprograms.htm#i4074

  (1) for example: functions can be used in SQL, procedures cannot.
  (2) generally, the methods are used for DML.
  It depends on how you are performing the function:

  hoek&XE> create table t as select 1 col from dual;
  
  Tabel is aangemaakt.
  
  hoek&XE> create or replace function f
    2   return number
    3   as
    4    num number;
    5   begin
    6    update t
    7    set col = 2;
    8    num := sql%rowcount;
    9   return(num);
   10  end;
   11  /
  
  Functie is aangemaakt.
  
  hoek&XE> select f from dual;
  select f from dual
         *
  FOUT in regel 1:
  .ORA-14551: cannot perform a DML operation inside a query
  ORA-06512: at "HOEK.F", line 6
  
  hoek&XE> 
  
  hoek&XE> var num number
  hoek&XE> exec :num := f;
  
  PL/SQL-procedure is geslaagd.
  
  hoek&XE> print :num
  
         NUM
  ----------
           1
  
  hoek&XE> select * from t;
  
         COL
  ----------
           2
  

  Yes 3A)
  No. 3B), see 2.

• SQL Analysis: Extract column expression and SQL query tags

  Hi all
  
  I have a requirement and posted earlier reg. this.
  An SQL statement must be parsed and column names must be extracted separately.
  I can't use dbms_sql. I need to extract the names of columns through manipulating strings in PL/SQL only.
  I don't have to validate the SQL code.
  
  for example.
  SELECT EMPNO, Upper (ENAME), DEPTNO Emp_Name, DECODE (Deptno, 10, 'ACCOUNTING', 20, 'HR', 'OTHER') Dept_Name OF double
  
  Output should be,.
  
  Column names
  --------------------
  (1) EMPNO
  (2) Emp_Name Upper (ENAME)
  (3) DEPTNO
  (4) Dept_Name DECODE(Deptno,10,'ACCOUNTING', 20,'HR','OTHERS')
  
  How can I retrieve the names of columns using PL/SQL for the above format?
  Pls help.
  
  Kind regards
  Sam

  If the requirement is to have them come forward...

  SQL> ed
  Wrote file afiedt.buf
  
    1  declare
    2    v_sql VARCHAR2(4000) := q'[SELECT EMPNO, Upper(ENAME) Emp_Name, DEPTNO, DECODE(Deptno,10,'ACCOUNTING', 20,'HR','OTHERS') Dept_Name FROM Dual]';
    3    v_cols VARCHAR2(4000) := regexp_replace(v_sql, 'SELECT (.*) FROM .*', '\1');
    4    PROCEDURE parse_col(p_cols IN VARCHAR2, lvl NUMBER) IS
    5      v_str VARCHAR2(4000);
    6      v_br  NUMBER := 0;
    7      v_pos NUMBER := 1;
    8    begin
    9      LOOP
   10        EXIT WHEN v_pos > length(p_cols);
   11        CASE SUBSTR(p_cols,v_pos,1)
   12          WHEN '(' THEN
   13            v_br := v_br + 1;
   14          WHEN ')' THEN
   15            v_br := v_br - 1;
   16          WHEN ',' THEN
   17            IF v_br = 0 THEN
   18              DBMS_OUTPUT.PUT_LINE(to_char(lvl,'fm99')||': '||trim(v_str));
   19              parse_col(SUBSTR(p_cols,v_pos+1),lvl+1);
   20              v_str := NULL;
   21              EXIT;
   22            END IF;
   23          ELSE NULL;
   24        END CASE;
   25        v_str := v_str || SUBSTR(p_cols,v_pos,1);
   26        v_pos := v_pos + 1;
   27      END LOOP;
   28      IF v_str IS NOT NULL THEN
   29        DBMS_OUTPUT.PUT_LINE(to_char(lvl,'fm99')||': '||trim(v_str));
   30      END IF;
   31    end;
   32  begin
   33    parse_col(v_cols,1);
   34* end;
  SQL> /
  1: EMPNO
  2: Upper(ENAME) Emp_Name
  3: DEPTNO
  4: DECODE(Deptno,10,'ACCOUNTING', 20,'HR','OTHERS') Dept_Name
  
  PL/SQL procedure successfully completed.
  
  SQL>
  

• Resolution of delay function and the borough

  A small question about the delay () function.  The resolution is apparently 1ms according to the help file.  What happens if I pass a value with a fraction of a mlliseconds, say something like:

  Delay(0.0128);
  

  For 12.8ms.  Does round to the top and the 13th month of delay?  Or round to 12ms?  Undefined?  Something else?

  Thank you.

  Hey tstanley,

  I did some research on the function and it seems that the feature is for timer trying to get closer to the 12.8ms, but because the resolution is only 1 ms, it could still be +/-1ms to the value that you enter. So in some of our tests, it was waiting for 13.8ms, since it is in the accuracy of the clock 1ms.

  Because Windows is not a real-time operating system, there are several factors that could keep the timer delay for 13.8ms - other processes can run on the machine and anticipate the process timer, for example. If this level of accuracy is required, you can consider using the CVI RT module on a system in real-time.