Hi Erich, > > > $ soelim foo | preconv | groff -Tutf8 | grep . ... > > > $ groff -V -Tutf8 -sk foo > > > preconv foo | soelim | troff -Tutf8 | grotty ... > >preconv -eutf8 example.mom | soelim | troff -Tutf8 | grotty > > Output: > > > Abschnitt.DOCTYPE DEFAULT UNICODE: Ä‐Ö‐Ü Ò‐Ó‐Ô‐Õ‐Ŏ‐Ō‐ > > Ç ä‐ö‐ü‐ ß ò‐ó‐ô‐õ‐ŏ‐ō‐ ç so’eD FILE BEGINS: UNICODE: Ã‐Ã‐à Ã‐Ã‐ > > Ã‐Ã‐Å‐Å‐ à ä‐ö‐ü‐ à ò‐ó‐ô‐õ‐Å‐Å‐ ç
And that's faulty output since you didn't request `²', for example. > But I want pdfmom and change that to > > preconv -eutf8 example.mom | soelim | pdfmom > example.preconv.pdf And you're carrying the fault over. My email that Peter replied to above shows how groff gets the pipeline order wrong for your needs, and thus -V's output is wrong too. > As it seems, the preconv-pipe manages latin1, but not the "true" > utf-8. preconv is only processing example.mom, converting it to input groff can read. When groff does that, it sees the .so and goes off and reads a second file, one that preconv hasn't converted. That's you problem. -- Cheers, Ralph. https://plus.google.com/+RalphCorderoy
