The Go version treats file data as byte arrays; might want to do the same and use Scheme bytevectors instead of strings. OTOH the docs say it's not supposed to be a general binary format...
Similar format: https://github.com/google/hrx
The Markdown of archives.
I have been using txtar to provide context to LLM's recently.
[flagged]
Simple format. Makes me want to write a parser for it using a shared buffered reader library Iām working on in Rust.