grep to only match whole words with non-english characters

Question

I want to only grep for a whole word. The problem is a file contains non-english characters, so grep -w doesn't work (f.e. matches "aąbcć" when searching for "bc"). I can't write any working regex with lookaround either. Can anybody help me?

Answer 1

Try to use word boundaries in grep:

grep "\<bc\>" file

Answer 2

Requiring GNU grep: grep -P '(^|\s)\Kbc(?=$|\s)' file

Using awk, I wonder if this would work:

awk -v word="bc" '{for (i=1; i<=NF; i++) if ($i == word) {print; break}}' file