Hi Bhagat,
U can use the following query to make connection to a oracle database and spool the reults
to a text file:--

sqlplus -s username@databse_string/password << EOF
whenever sqlerror exit sql.sqlcode;
whenever oserror exit FAILURE

set define off
set head off
set feedback off
set echo off
set pagesize 0
set pages 0
set linesize 200
set timing off

spool appropriate path/dummy.txt
UR QUERY GOES HERE
spool off

EOF

If u want to make only manipulations to databse without spooling the results to a any fimle,
dont put it with in SPOOL command.

Hope it works for u......

here is the script to query and put output in a file

and check this url where the same question was answered.this

Thanks both of you!!!
That was very helpful!!
Appreciate your guidance

Regards,
Bhagat

Hi Guys,
I'm looking for a slightly modified script this time...
ie.Query Oracle database table and populate the result set of the query in a text file.The column values should be seperat
I also want the count of records returned by the Oracle query.
Could you someone help me out with a sample code?

Thanks,
Bhagat

Hi Bhagat
There is small change and i guess will be pretty simple.Give path and filename in spool command where it will store the output.In select statement give column names you want to be in the output and here '|' is used as a delimiter you can use any char you want to.
and counting for no of rows you can chack with a command to spooled file
x=`wc -l /path/filename | awk '{print $1}'`
echo "no of records \c $x"

sqlplus -s user/pwd@host<<eof>urfile
set serveroutput on feedback off linesize 1000 pagesize 0;
select * from table where rownum<5;
EXIT;
eof