> No, it's not a bug in preconv, it's a bug in troff: `.hcode' and `.hw' > are limited to raw 8bit characters. Instead, they should accept any > characters entities like `\[u00AD]'. Similarly, `.hpfcode' should be > modified to allow entities, too.
I agree! That really would be the best solution. (.hcode then can't match a .tr consistent syntax without white space separators like ".tr abcd" but to address non-7-bit ASCII this is ok.) > Please file an issue so that it is not forgotten. Ok, I'll do that. > PS: A clean solution currently is to put `.hw' lines into a separate > file that uses the proper 8bit encoding, and that file gets then > included with `.so'. Since `preconv' gets called before `.soelim' > in groff's pipeline, different encodings won't clash. That's exactly what I did yesterday after the discussion with Ralph. I really do not like latin1 coded files put for a one-line one-purpose file it is ok :) Carsten
