The Regex engine currently being used seems to not recognise non-Latin characters. Thus, see "word Boundary" in the middle of words.
Example: napušavali the see the š as an non-word character and matches \bavail in napu\bš\bavali.
Feed Preview:

Feed:

Oddly regex101 show that the Rust Regex Crate shouldn't do that:

Javascript Regex does though:

The Regex engine currently being used seems to not recognise non-Latin characters. Thus, see "word Boundary" in the middle of words.
Example:
napušavalithe see thešas an non-word character and matches\bavailinnapu\bš\bavali.Feed Preview:
Feed:
Oddly regex101 show that the Rust Regex Crate shouldn't do that:
Javascript Regex does though: