Re: Translation design patterns or approaches

[Bento Borges Schirmer]

Hey Oliver,

Em sex., 8 de nov. de 2024 às 19:05, Oliver Corff via GNU roff
typesetting system discussion <groff@gnu.org> escreveu:
> Check, e.g. the macro files trans.tmac, de.tmac, sv.tmac, cs.tmac,
> fr.tmac etc.in /usr/share/groff/1.23.0/tmac
>
> They demonstrate the mechanism how you can use string macros with good
> mnemonic names that issue the names of months, days etc. according to
> selected language, including a lot of phrases necessary for writing letters.

I glanced at the files. I saw nothing for portuguese (groff 1.23.0),
but I always come across man pages in portuguese! [1]
In the future I could work to provide some pt.tmac and a hyphen.pt,
but maybe the people from Debian [2] already did it? I don't know.
Anyway, the month names mnemonics are useful!

I translated a little more yesterday. I noticed some patterns.

1. More common words like "e" and "and" could benefit from mnemonics
2. Loanwords are widespread in portuguese. The most common are not
italicized, but the remainder are. Some mechanism would be useful to
automatically italicize.
3. Borrowing from C's #ifdef #else #endif conditional compilation, it
would be useful to define regions that disappear and reappear
depending on language.

I'm just thinking. Except regarding "e" and "and", I don't think I'll
put any additional effort for now haha.

Best regards,
Bento

[1] https://man.archlinux.org/listing/extra/man-pages-pt_br/
[2] https://salsa.debian.org/manpages-l10n-team/manpages-l10n