Hi,
i am pretty new to Unix environment .....
Can i get some help from any of you guyz on writing Unix script.
my requirement is like reading a csv file, finding a specific pattern in the lines and repalce the string with new string and write it to another file.

My file is file ABC123.dat
Col1,Col2,Col3
102628,AB10 0,21
102628,AB10 1,21
102628,AB10 2,21
102628,AB23 5,21
102628,AB35 9,21
102628,AB53 8AA,21
102628,AB53 8AB,21
102628,AB53 8AD,21
102628,AB53 8AE,21
102628,AB53 8AG,21
102636,AB54 0,21
102636,AB54 1,21
102636,AB54 2,21
102636,AB54 3,21
102636,AB54 4,21
102636,AB54 5,21
102636,AB54 6,21
102636,AB54 7A,21
102636,AB54 7B,21
102636,AB54 7D,21
102636,AB54 7E,21
888888,A99 9,21

and i have written a small script which finds the the pattern and prints the line numbers too but i am unable to write to another file....

file=/export/home/prashant/ABC123.dat

String1="[,][A-Z][A-Z][0-9][0-9][ ][0-9][,]"
String2="[,][A-Z][0-9][0-9][ ][0-9][,]"
String3="[,][0][,]" # Replacement for the string 1 & 2 if found the pattern in the file.

if [ -f "file" ]; then
echo "File $file does not exist."
exit 1
fi
cmd=$`(grep -c "[$String1|$String2]" $file)`
line=$(grep -n "$String1" $file)
line1=$(grep -n "$String2" $file)

if [ "$cmd" != "0" ]; then

echo "the String exists what you are expecting in the Line Number"
echo "$line"
echo "$line1"
else
echo "Word does not exist."
fi

out put is something like this ....
the String exists what you are expecting in the Line Number
2:102628,AB10 0,85
3:102628,AB10 1,85
4:102628,AB10 2,85
5:102628,AB23 5,85
6:102628,AB35 9,85
12:102636,AB54 0,85
13:102636,AB54 1,85
14:102636,AB54 2,85
15:102636,AB54 3,85
16:102636,AB54 4,85
17:102636,AB54 5,85
18:102636,AB54 6,85
23:888888,A99 9,85

but i am unable to replace the string1 & string2 after finding in the file with the string3 variable to another file .....
thanks
prashant

Hi,

something like this:

command:

grep -n -v "AB[0-9]\{1,2\} [0-9]\{1,2\}[A-Z]\{1,2\}" file > newfile

1:Col1,Col2,Col3
2:102628,AB10 0,21
3:102628,AB10 1,21
4:102628,AB10 2,21
5:102628,AB23 5,21
6:102628,AB35 9,21
12:102636,AB54 0,21
13:102636,AB54 1,21
14:102636,AB54 2,21
15:102636,AB54 3,21
16:102636,AB54 4,21
17:102636,AB54 5,21
18:102636,AB54 6,21
23:888888,A99 9,21

Thank you Chris .... its working
now i have to compare the original file with the new file and eliminate the matching records and write it to another file ..... need your suggestion on this trying to with diff command ....
but does it work on huge file .csv file which have a couple of million records.

Thank you very much
Prashant

command:

grep -n "AB[0-9]\{1,2\} [0-9]\{1,2\}[A-Z]\{1,2\}" file

7:102628,AB53 8AA,21
8:102628,AB53 8AB,21
9:102628,AB53 8AD,21
10:102628,AB53 8AE,21
11:102628,AB53 8AG,21
19:102636,AB54 7A,21
20:102636,AB54 7B,21
21:102636,AB54 7D,21
22:102636,AB54 7E,21