JSON to Spark-JSON produces non well-formed JSON

[jeromesimeon]

In some cases, the JSON to SparkJSON data conversion produces irregular JSON. The main issue is with non-integer numbers, which get printed without digits after the decimal point. e.g.:

{"$data": "{"balance": 0., "customer": {"left": "10"}, "id": "0", "lastCheckTime": {"right": null}, "status": "Good"}", "$type": ["creditcard.Account"]}

Note the '0.' for the balance field.