I could allow the alternative of using quotations, instead of hyphens.
`book:: Bible | "Song of Solomon" 5:1 | KJV`
`book:: "2 Ezra" preface:2-21`
`book:: "Wuthering Heights" "Ch XI":Preamble-8`
Discussion
The books and chapters in the DRB have prefaces and preambles, not only verses.
that's not a big addition of complexity to parsing that field, seems reasonable.
yeah, not sure, not often appearing in titles but then you can escape them. familiar to json but not really exactly familiar to wikilinks. most wikilinks refer to URLs so they are without spaces but ... well, idk, it really depends. the point i'm making is that the [[]] and the | separators serve as anchors for segmenting and parsing the text, so spaces are bounded within it already. writing a scan through text for this is simple, first you look for [[ then once you find that you are looking for ]]. if you hit the end of the event } then nothing was found.
i really see no reason to not make it as human friendly as possible that what someone would write off the top of their head is likely in the tags as the "proper" version.
it needs a special parser. we aren't bound by the limitations of JSON. what's more important is we are bound by the lowest common denominator of the user interested in attempting to craft a valid reference token.