UTL_FILE Overwrites Data

Question

Guys,
I am trying to write multiple files with multiple lines in each file. These files will be named with ID grabbed from cursor(Cursor 1 in code). This code does write multiple files and names the file name correctly. The only problem is in each file there are supposed to be multiple lines and it seems to overwrite all of these lines and it only input the header and the last line of data grabbed from cursor 2. Please help. I need a condition that says write all the lines first on the first file before moving to the next file. Data in each file are unique by the way. I am stuck please help
--SET FEEDBACK OFF
--SET TIME OFF
SET VERIFY OFF
SET SERVEROUTPUT ON SIZE 1000000
--SET ECHO ON;
--SET TERMOUT ON;
--SET FEEDBACK ON;
--SET PAGESIZE 24;
--SET LINESIZE 300;
--SET SPACE 2;
--SET HEADING ON;
DECLARE
companyID varchar2(15);
V_OUTPUT VARCHAR2(2000);
V_HEADER VARCHAR(2000);
V_PRODUCTION_PERIOD     DATE := '&&1';
v_taxhdl varchar2(1);
v_accounting_date number(6);
v_production_date number(6);
v_commodity varchar2(2);
p_filename varchar2(200);
fHandler UTL_FILE.FILE_TYPE;
--Getting the company codes into companyID variable
CURSOR C1 IS
SELECT DISTINCT company as companyID
FROM myTable
WHERE (commodity = 'CN'
AND accounting_date = TO_NUMBER(TO_CHAR(ADD_MONTHS(TO_DATE('&&1','DD-MON-YYYY'),-1),'yyyymm'))
AND production_date = TO_NUMBER(TO_CHAR(ADD_MONTHS(TO_DATE('&&1','DD-MON-YYYY'),-1),'yyyymm')))
OR (commodity != 'CN'
AND accounting_date = TO_NUMBER(TO_CHAR(TO_DATE('&&1','DD-MON-YYYY'),'yyyymm'))
AND production_date = TO_NUMBER(TO_CHAR(ADD_MONTHS(TO_DATE('&&1','DD-MON-YYYY'),-1),'yyyymm')))
ORDER BY company;
CURSOR C2(p_company number) IS
SELECT taxhdl, accounting_date, production_date, commodity
FROM myTable
WHERE (commodity = 'CN'
AND ACCOUNTING_DATE = to_number(to_char(ADD_MONTHS(TO_DATE('&&1','DD-MON-YYYY'),-1),'yyyymm'))
AND company = p_company )
OR (commodity != 'CN'
AND ACCOUNTING_DATE = to_number(to_char(TO_DATE('&&1','DD-MON-YYYY'),'yyyymm'))
AND company = p_company )
GROUP BY taxhdl, accounting_date,production_date,commodity
ORDER BY commodity, production_date;
--Setting the the Heading of the file
BEGIN
V_HEADER := ('COMPANY_ID'||','||'TAXHDL' || ',' || 'ACCOUNTING_DATE'|| ',' ||'PRODUCTION_DATE' || ',' ||'COMMODITY');
FOR REC IN C1 LOOP
companyID := REC.companyID;
p_filename := companyID||'.SUM'||'&&1'||'.csv';
FOR C2_REC IN C2(companyID) LOOP
v_taxhdl := C2_REC.taxhdl;
v_accounting_date := C2_REC.accounting_date;
v_production_date := C2_REC.production_date;
v_commodity := C2_REC.commodity;
V_OUTPUT := (companyID ||','||v_taxhdl|| ',' ||v_accounting_date||',' ||
v_production_date||','||v_commodity);
/*The part needs fixing*/
fHandler := UTL_FILE.FOPEN('myDir',p_filename,'w');
UTL_FILE.PUTF(fHandler,V_HEADER||'\n');
UTL_FILE.PUTF(fHandler,V_OUTPUT);
UTL_FILE.FCLOSE(fHandler);
END LOOP;
-- exception
-- when NO_DATA_FOUND then
-- utl_file.fclose(f);
END LOOP;
END;

Answer 1

When opening a file, a file pointer is created. This points to byte (or character) where any I/O operation on that file will start. Each I/O operation on the file, moves this file pointer accordingly.
In other words, when you do a read on that file, that is the point that the read will start reading data. When you write to the file, that is the point to which your buffer will be written to.
You can open a file into two way - reset or append.
When opening in reset mode, the file pointer is moved to the beginning of the file. This is what you typically want when opening a file and reading from it. You want to read from the very 1st byte in that file.
When opening in append mode, the file pointer is moved to the end of the file. This is what you typically want to do when you open a file and write to it. You want to start writing (appending) to the end of file.
What happens when you open a file in reset mode and start writing to it? You overwrite whatever was there at the beginning of the file.
How do you open the file in that loop? You do not use the append attribute - so no need to guess where the file pointer is when you write to it.
And as already commented - not a good idea to open and close the file in a loop. It is a huge overhead. Open the file once. Loop through the data and write it to file. Close the file.
If you want to make the changes written to file immediately visible to other processes, flush the I/O buffer in the loop after the I/O write statement.