Re: [web-annotation] Split model into vocab and model+json-ld

@azaroth42 I have a tool that renders [some 
HTML](https://livefyre.github.io/activity-vocabulary/) from a JSON-LD 
document describing [the underlying 
vocabulary](https://github.com/gobengo/livefyre-activity-vocabulary/blob/master/livefyre.jsonld)
 (borrowing some rdfs and hydra props).

The script is a little overfit for my use case, but perhaps a good 
starting point to something more generalizable. I just published it at
 
[gobengo/livefyre-activity-vocabulary](https://github.com/gobengo/livefyre-activity-vocabulary).

Basically
[This 
script](https://github.com/gobengo/livefyre-activity-vocabulary/blob/master/bin/make-public-data.js)
 reads the jsonld and uses node 'jsonld' lib to mash it into a more 
customer JSON structure using JSON-LD framing. It also renders some 
markdown files to html to use as descriptions.

The structure is like
```bash
⚡ bin/make-public-data.js | jq '[(keys - ["data"])[] as $key | { 
($key): .[$key] | keys }] | add'
{
  "classes": [
    "activities",
    "objects",
    "properties"
  ]
}
```

Then it uses that as template context to some HTML templates (using 
`[harp](https://www.npmjs.com/package/harp) compile`)

The 
[Makefile](https://github.com/gobengo/livefyre-activity-vocabulary/blob/master/Makefile#L38)
 makes it all explicit.

You may find it useful.

-- 
GitHub Notification of comment by gobengo
Please view or discuss this issue at 
https://github.com/w3c/web-annotation/issues/97#issuecomment-159189627
 using your GitHub account

Received on Tuesday, 24 November 2015 08:02:19 UTC