[SOLVED] How to input egrep result as a variable to awk

Issue

I have a dataset that looks like the below:

col1,col2,col3,col4
read1,chr1,Unassigned_NoFeatures,
read2,chr2,Assigned,
read3,chr3,,Assigned

What I want to do is to perform egrep on the "Assigned" and "Unassigned_NoFeatures" strings and then append them to the read names in the first column. The problem I face is that these strings are not always in the same column but I need to grep them and append them to the first column (per entry).

I tried the approach below to input egrep results for each entry in my file and input it as a variable to awk but it did not work.

cat my_file.csv | variable=$(egrep -o "Assigned|Unassigned_NoFeatures") | awk -F, '{print $1"_"variable,$2}'

The desired output should be as:

read1_Unassigned_NoFeatures,chr1
read2_Assigned,chr2
read3_Assigned,chr3

How can I make the script work? Thanks

Solution

You never need grep when you're using awk. Given your input all you need to do is print the string from the 3rd or 4th field, whichever is not empty, e.g. using any awk:

$ awk 'BEGIN{FS=OFS=","} NR>1{print $1 "_" ($3?$3:$4), $2}' my_file.csv
read1_Unassigned_NoFeatures,chr1
read2_Assigned,chr2
read3_Assigned,chr3

or even just, as @jhnc suggested in a comment:

awk 'BEGIN{FS=OFS=","} NR>1{print $1 "_" $3$4, $2}' my_file.csv

but if you actually did need to find strings that match a regexp across a whole line it'd be:

awk 'BEGIN{FS=OFS=","} match($0,/Assigned|Unassigned_NoFeatures/){print $1 "_" substr($0,RSTART,RLENGTH), $2}' my_file.csv

Answered By - Ed Morton

Answer Checked By - Senaida (WPSolving Volunteer)