The code doesn't try to find all possible acronyms. It specifically tries to match the list of known acronyms against the supplied text. It does this with a single regular expression rather than analyzing each line for a collection of upper case characters. It wouldn't be too hard to flip things round to extract all things that look like an acronym from the text and then compare with the acronym list, but it would need us to agree what an acronym is since some seem to have lower case characters so \b[A-Z][A-Z\&][A-Z]\b isn't sufficient.

ah ok .. I remember you telling me about the match expression , I had not twigged that would mean no warning. The original Gaia one looks for any UPPERCASE words and reported them - this could be a separate.

I've made some fixes. I now scan for non-standard acronyms (R&D, FoV, arcsec etc) as a regex, but then I scan the text for acronyms defined as 2 or more upper case characters and compare with the list of known acronyms. Undefined acronyms are now reported.

ok NOW i see loads of undefined acronyms !