Add "json_path" config option for nested data

[laurentS]

I was trying to load a JSON file that looks like this (the actual file is here):

{
  "3166-1": [
    {
      "alpha_2": "AW",
      "alpha_3": "ABW",
      "name": "Aruba",
      "numeric": "533"
    },
    {
      "alpha_2": "AF",
      "alpha_3": "AFG",
      "name": "Afghanistan",
      "numeric": "004",
      "official_name": "Islamic Republic of Afghanistan"
    }
]}

which led to an error because the list of useful objects is nested under the 3166-1 key.
I was able to load the data by adding the following json_array = json_array['3166-1']just after

tap-spreadsheets-anywhere/tap_spreadsheets_anywhere/json_handler.py

Line 26 in 84f6b8e

json_array = json.load(reader)

which tells me it should be possible to add a config option for the JSON handler (similar to delimiter for CSV handler for instance) that tells it where to look for the data. At a minimum, a string defining the key under which to get a list of items (so in my example, it could be like data_path: "3166-1"), but I imagine it should be possible to declare a full path, something like data_path: "section.subheading.one_more_level" so that it can grab items from arbitrarily deep in the tree.

I might give it a try in a PR if it makes sense to add something like this (I think so, I have a couple of use cases for it already 😉 )