Unicode input with plain LuaTeX

Question

grün
gr\"un
\bye

yields

grn
grün

Is it possible to have the Unicode ü correctly rendered with plain LuaTeX?

---

Would it also be possible with plain TeX and a METAFONT font? Or would a METAFONT font have to be converted into something else if one wanted to use it with Unicode input?

Answer 1

You might use METAFONT fonts:

\catcode`ü=\active
\protected\def ü{\"u}

grün
gr\"un
\fontname\font
\bye

You might study sections from 6.3 to 6.5 of the LuaTeX manual to see whether an approach with virtual fonts is feasible. However, this would require extensive work for defining virtual fonts from existing ones. Using European Modern fonts (that are T1 encoded, using LaTeX lingo) might be easier.

On the other hand, you can easily inject a font loader in LuaTeX and use OpenType fonts. For instance

\input luaotfload.sty

\font\tenrm="Latin Modern Roman"
\tenrm

grün
gr\"un
\fontname\font

\bye

Note that the second “grün” prints wrongly, because accents are not in the slots used by plain TeX and one needs to redefine the accent commands, for instance

\input luaotfload.sty

\def\"#1{#1^^^^0308}

\font\tenrm="Latin Modern Roman"
\tenrm

grün
gr\"un
\fontname\font

\bye

Redefining a few accents is obviously easier than many active characters.

Answer 2

I'm not an expert on this, so I cannot exactly explain how it works and why it works and stuff. But since nobody gave you satisfying answer on your second question sofar, I try to contribute myself. So, I guess I have a solution if your concern is using German. If you also want to use the Greek alphabet or math symbols and stuff, then I am afraid, this cannot work without LuaTeX or something similar that supports Unicode.

This problem is solved in the Czech-Slovak distribution of TeX, which, if I understand it corretly, has its own slightly modified fonts and internally it transforms the utf8 input into iso8859-2 coding, where all the Czech, as well as German, characters appear. See https://www.ctan.org/pkg/csplain

You can write, for example

Příliš žluťoučký kůň úpěl ďábelské ódy.

„Fix, Schwyz!“, quäkt Jürgen blöd vom Paß.

\bye

And after compiling using pdfcsplain instead of pdftex, you get the right result.