I am using this AWK command to produce a file for each value in the first column of input:
awk -F "," '{print > $1 ".csv" }' test.csv
test.csv content
1,Rahul,
2,Atul,
3,Sachin,
4,Reyansh,
1,Rahul,
3,Sachin
This produces output files named 1.csv, 2.csv, 3.csv and 4.csv.
However, my requirement is that I need to include a prefix to get sourcefile1.csv, sourcefile2.csv, sourcefile3.csv and sourcefile4.csv.
The filename (sourcefile in the example) is in a shell variable $fle_name.
I tried to include it in the AWK program this way:
awk -F "," '{print > $1 "`echo $fle_name`.csv" }' test.csv
But this produces $fle_name1.csv, $fle_name2.csv etc.
Shell variables are not expanded in single quotes so awk prints the
literal value of 1`echo $fle_name`.csv and so on. In most cases
though it's a good practice to enclose awk command between single
quotes so that it doesn't conflict with shell syntax. For this reason
we could pass the shell variable to awk using -v:
awk -F "," -v file="$fle_name" '{print > $1 file ".csv" }' test.csv
Example:
$ fle_name=MY_FILE
$ gawk -F "," -v file="$fle_name" '{print > $1 file ".csv" }' test.csv
$ ls -l
total 20
-rw-rw-r-- 1 user user 18 Mar 26 11:04 1MY_FILE.csv
-rw-rw-r-- 1 user user 8 Mar 26 11:04 2MY_FILE.csv
-rw-rw-r-- 1 user user 19 Mar 26 11:04 3MY_FILE.csv
-rw-rw-r-- 1 user user 11 Mar 26 11:04 4MY_FILE.csv
-rw-rw-r-- 1 user user 56 Mar 26 11:00 test.csv