well, it's your baby i guess. but i won't like it so much if it isn't something that I as a computer programmer requires as much thinking as writing valid JSON.
Discussion
'g/isn't/is'
Most humans would stick to stuff like
`book:: 1 John 2:7-12 | KJV`
I recognize and render Bible books, Quran, and some others without hyphens or quotations. A number preceding letters is concatenated to a title. Anything that is " number:number-number" can be assumed to be chapter-number:versebeginning-verseending. And there are typical keywords like "preface" or "appendix", that can be identified. I also work with plural or singular and stuff like "Ch" "Ch." "Chapter" and Roman numerals. I can add that to the spec, as a hint to later parser-writers.
Anything more complex is probably auto-generated or written by an expert.
Another easy one is
`book:: Tale of Two Cities 8: Intro-3`
It's obvious that they mean
`book::tale-of-two-cities 8:introduction-3`