How to print only words that doesn't match between two files?

Question

FILE1:

cat
dog
house
tree

FILE2:

dog
cat
tree

I need to be printed only: house

Answer 1

$ cat file1
cat
dog
house
tree

$ cat file2
dog
cat
tree

$ grep -vF -f file2 file1
house

The -v flag only shows non-matches, -f is for a filename to use as a filter, and -F is for exact matches (doesn't slow it down with any pattern matching).

Answer 2

If you want just the words, you can sort the files, diff them, then use sed to filter out diff's symbols:

diff <(sort file1) <(sort file2) | sed -n '/^</s/^< //p'

Answer 3

Using awk

awk 'FNR==NR{arr[$0]=1; next} !($0 in arr)' FILE2 FILE1

First build an associative array with words from FILE2 and than loop over FILE1 and only print those.

Using comm

comm -2 -3 <(sort FILE1) <(sort FILE2)

-2 suppresses lines unique to FILE2 and -3 suppresses lines found in both.

Answer 4

Awk is an option here:

awk 'NR==FNR { arr[$1]="1" } NR != FNR { if (arr[$1] == "") { print $0 } } ' file2 file1

Create an array called arr, using the contents of file2 as indexes. Then with file1, look at each entry and check to see if an entry in the array arr exists. If it doesn't, print.