training android graphics simulation

Make a parser to accept input as follows:
[ {
"key": "move",
"target": {
"targetId": 0
}
}, {
"key": "speak",
"target": {
"dialog": ""
}
}
]

Make a scene descriptor that returns the description of the scene as formatted: {
"instruction": "Do something",
"description": [ {
"id": 0,
"position": {},
"rotation": {},
"scale": {},
"boundingBox": {}.
"animations": [
"idle",
"walk"
]
}
]
}