Export table to csv file by using procedure (csv name with timestamp)

Question

I want to export one table to each csv file and name the csv file with timestamp. For example, if I have a table t1 , after export, a csv file name t1.YYYYMMDDHHMISS.csv is generated. Here is my code:

create or replace procedure 
T_to_CSV(Tname varchar2,Directory varchar2)

BEGIN

set colsep ,     -- separate columns with a comma
set pagesize 0   -- No header rows
set trimspool on -- remove trailing blanks
set headsep off  -- this may or may not be useful...depends on your 

spool timestamp.csv --don't know how to write in directory

select *
  from Tname

end

Here are my questions:

(1) I don't know how to output the csv file into the requried directory, how to fix the spool code please?

Should I use spool D:\\path\\filename.csv please?

(2) In the spool line, how to change the name of the csv file using the timestamp now() please?

Answer 1

There are a few steps:

• create a directory using `CREATE DIRECTORY my_dir as 'C:\\dir';
• make sure that Oracle has read,write on the folder on the computer (best accomplished by creating it in the Oracle install folder)
• grant the user executing the procedure GRANT read,write on DIRECTORY my_dir to the_user;
• download and compile the handy procedure here

I have used this and it works really nicely.

Usage

data_dump ( 'Select emp_name from emp',
             CURRENT_TIMESTAMP||'filename.csv',
             my_dir);

(vastly simplified sample!)

After creating the directory verify your work by running this:

• Select * from ALL_DIRECTORIES;
• you should see your directory
• logon to the machine where the database is located and verify the folder path exists and the oracle user has permissions on it. Networked drives are only possible if the user running the oracle service has permissions on that folder

Answer 2

• Save and copy file in a directory {file need not be executable}
• Export ORACLE_HOME, PATH and SID of database instance
• Navigate to that directory and execute sqlplus

• Spool file will be created in the same directory as the .sql file

   SET COLSEP , SET HEADSEP OFF -- character used when defining a two-line column heading SET TRIMSPOOL ON -- trim trailing spaces from each line SET LINESIZE 32000 -- number of characters to be printed on one line SET WRAP OFF -- truncates lines longer than LINESIZE SET NUMWIDTH 5 -- default width while displaying numbers COLUMN tm new_value iso8601 noprint SELECT to_char(sysdate, 'YYYY-MM-DD') tm FROM dual; spool output_file_&iso8601..csv -- Append new data to spool file: "spool output_file_&iso8601..csv append" SELECT * FROM table_name spool OFF 

Answer 3

Thanks kevin for sharing the procedure and it was very useful for me. I have customized the code:

1. To Add the column names in the output csv file which was not working earlier.
2. When i was passing the delimiter as parameter then it was adding the comma in the starting of every row(,1,2,3) which has been corrected.

I am also sharing the customized code which might help others. Customized code can be downloaded here .

1. Customized Code to add column names

  FOR i IN t_describe.FIRST .. t_describe.LAST LOOP IF i <> t_describe.LAST THEN put('UTL_FILE.PUT(v_fh,'''||t_describe(i).col_name||'''||'''||v_delimiter||''');'); ELSE put(' UTL_FILE.PUT(v_fh,'''||t_describe(i).col_name||''');'); END IF; END LOOP; put(' UTL_FILE.NEW_LINE(v_fh);'); 

2. Customized Code for delimiter

   IF i <> t_describe.LAST THEN
   put(' UTL_FILE.PUT(v_fh,"'||t_describe(i).col_name||'"(i) ||'''||v_delimiter||''');');
   ELSE
   put(' UTL_FILE.PUT(v_fh,"'||t_describe(i).col_name||'"(i));');
   END IF;

And the correct way to call the procedure is to bind the variable with values

data_dump(query_in => 'Select 1 from dual',file_in => 'file.csv',directory_in => 'MY_DIR', delimiter_in => '|' );

Thanks

Naveen