JSON Array Structure Variations

Question

Below are 3 JSON Array structure formats...

The first one, the one outlined at JSON.org, is the one I am familiar with:

Format #1

{"People": [
  {
    "name": "Sally",
    "age": "10"
  },
  {
    "name": "Greg",
    "age": "10"
  }
]}

The second one is a slight variation that names the elements of the array. I personally don't care for it; you don't name elements of an array in code (they are accessed by index), why name them in JSON?

Format #2

{"People": [
  "Person1": {
    "name": "Sally",
    "age": "10"
  },
  "Person2": {
    "name": "Greg",
    "age": "10"
  }
]}

This last one is another variation, quite similar to Format #2, but I have a hunch this one is incorrect because it appears to have extra curly braces where they do not belong.

Format #3

{"People": [
  {
    "Person1": {
      "name": "Sally",
      "age": "10"
    }
  },
  {
    "Person2": {
      "name": "Greg",
      "age": "10"
    }
  }
]}

Again, I'm confident that Format #1 is valid as it is the JSON Array format outlined at JSON.org. However, what about Format #2 and Format #3? Are either of those considered valid JSON? If yes, where did those formats come from? I do not see them outlined at JSON.org or on Wikipedia.

Answer 1

Both #1 and #3 are (nearly - there are commas missing) valid JSON, but encode different structures:

• #1 gives you an Array of Objects, each with name and age String properties
• #3 gives you an Array of Objects, each with a single Object property, each with name and age String properties.

The #2 is invalid: Arrays (as defined by [ ... ]) may not contain property names.