[SOLVED] Unix shell - Find unique columns by comparing two files

Issue

I have two files where I want to compare the first column to find the unique ones using unix shell.

File 1:

yellow
green
red
white

File 2:

yellow,not_started
red,in_progress

I want the output to be

green
white

Solution

Easy to do with join:

$ join -t, -v1 <(sort file1.txt) <(sort file2.txt)
green
white

-v1 tells it to print out records in the first file that don't match up to a record in the second file. The sorts are because join requires its input files to be sorted based on the (default first) join column.

Alternative using awk:

$ awk -F, 'NR == FNR { seen[$1] = 1; next } !($0 in seen)' file2.txt file1.txt
green
white

Note the second file comes first in the arguments to awk.

Answered By - Shawn