I'm revising some PDFs which each look at a single font, using XeLaTeX, and I've hit a gotcha which I had not expected. I show all the letters of selected Latin alphabets, one of which is Welsh because of its wide range of accents, and I'm revising my treatment of Old Standard (OldStandard-Regular.otf).
The Welsh alphabet appears fine, but amongst the large amount of output are reports for missing codepoints. That in itself is normal because I finish the document with showing currency symbols and other symbols, whether or not they are present. But just before the missing currency symbols I noticed: Missing character: There is no Ỳ (U+1EF2) in font Old Standard Regular/OT:scrip t=latn;language=dflt;mapping=tex-text;! Missing character: There is no ỳ (U+1EF3) in font Old Standard Regular/OT:scrip t=latn;language=dflt;mapping=tex-text;! Which is true, my listing of the codepoints omits those. And yet the alphabet is displayed, including Ỳ and ỳ. I wonder how many similar issues I had overlooked. What I don't understand is why the letters were rendered - I am used to spaces, or indications that a codepoint is missing. Is there logic somewhere which says that (maybe only a limited set of) accents can be produced using combining accents even where the precomposed codepoint is used ? This is TL2024 and I'm also using polyglossia although I doubt that could be involved. TIA ĸen -- Voluntary positions are like garden bindweed: untended, they will fill any void you might have. -- James Max
