That would be frustrating. Unless the matching is made case-sensitive, which would probably be a good idea when trying to filter out a proper name.

Why? You could just match on word boundaries '\bRust\b'

You would still get false positives with that but only regarding actual word semantics (i.e. articles mentioning Rust as oxidation process and not as programming language) but that's the tradeoff I guess

/([^a-z]|^)rust([^a-z]|$)/i