No way of adding custom snippets seems to function

[OndraX]

So first off, many thanks for the great work in bringing this awesome plugin into Vim as well.

However: I've tried to set up my own abbreviations/snippets using it (as per the readme; on a sidenote, the help provided there is really sparse), doing:

{
    "javascript": {
        "filters": "js",
        "profile": "js",
        "abbreviations": {
            "br": "<brzzz/>",
            "foo": "fullName: Ember.computed('firstName', 'lastName', {\n\tget() {\n\t\treturn `${this.get('firstName')} ${this.get('lastName')}`;\n\t}\n})"
        }
    }
}

in my ~/.vim/snippets.json (yes this is exactly the code krzkrzkrz had in his file in https://github.com/mattn/emmet-vim/issues/293; that is because he later reported it to work and I couldn't find an "official" example that was guaranteed to work -- the pointer to emmet's official guide still leaves the issue of my own potential JSON errors open; I can submit my own attempt at the file upon request) and

let g:user_emmet_settings = webapi#json#decode(join(readfile(expand('~/.vim/snippets.json')), "\n"))

in my ~/.vimrc of course also having installed the webapi plugin (via pathogen) However, in my sample index.js file, typing foo resulted in a

Another side note: How do I then add my own, vimscript-dictionary-esque other configuration settings, or should everything be in snippets.json?

Also, trying the vimscript way (again, copying the example code) and instead "manually" setting the user_emmets_settings global to this:

 let g:user_emmet_settings = {
 \  'indentation' : '  ',
 \  'perl' : {
 \    'aliases' : {
 \      'req' : 'require '
 \    },
 \    'snippets' : {
 \      'use' : "use strict\nuse warnings\n\n",
 \      'warn' : "warn \"|\";",
 \    }
 \  }
 \}

and typing req,, ; use,, ; or warn_,, , I simply saw all of them expand to an HTML-esque tag instead of the desired sequence of characters.

I must be missing something (JSON grammar, for example). But none of this seams to work for me. Any help (or readme expansion) would be greatly appreciated.