My current recommendation is to handle translations using DocBook. Most teams I know of using AsciiDoc do it this way. Translation text gets extracted from the document source, so it doesn't matter as much at that point what language the source document is in. There are plenty of enterprises offering translation for DocBook.

Once you have the translations for DocBook, you could either publish using DocBook or convert back the translated document back to AsciiDoc to publish using Asciidoctor tooling. (using something like docbookrx - https://github.com/opendevise/docbookrx)

Honestly, I don't really see why it would be a big deal to translate AsciiDoc if you can find a group that does it for Markdown (or even if they don't do Markdown). The process will be very similar. It's just plain text.

We do plan to add gettext extraction capabilities in the Asciidoctor ecosystem. I just haven't been able to spend time on it yet. All the work we are doing to parse the AST is clearly a prerequisite for being able to generate po files directly from AsciiDoc source with perfect accuracy...then subsequently weave the translated text back in when it's available.

The idea is to be able to leverage the existing translation tools while still working with AsciiDoc. The only thing the translator would need to be aware of is perhaps the inline markup. My understanding is that they are already well aware of that concept. Having said that, there's still a lot to explore and understand, I realize.

Cheers,