python append dictionary to list

Question

According to this post, I need to use .copy() on a dictionary, if I want to reference a dictionary which gets updated in a loop (instead of always referencing the same dictionary). However, in my code example below this doesn't seem to work:

main.py:

import collections
import json

nodes_list = ['donald', 'daisy', 'mickey', 'minnie']
edges_list = [('donald', 'daisy', '3'), ('mickey', 'minnie', '3'), ('daisy', 'minnie', '2')]
node_dict, edge_dict = collections.defaultdict(dict), collections.defaultdict(dict)
ultimate_list = []


for n in nodes_list:
    node_dict["data"]["id"] = str(n)
    ultimate_list.append(node_dict.copy())

for e in edges_list:
    edge_dict["data"]["id"] = str(e[2])
    edge_dict["data"]["source"] = e[0]
    edge_dict["data"]["target"] = e[1]
    ultimate_list.append(edge_dict.copy())

print(json.dumps(ultimate_list, indent=2))

As a result here I get the following:

[
  {
    "data": {
      "id": "minnie"
    }
  },
  {
    "data": {
      "id": "minnie"
    }
  },
  {
    "data": {
      "id": "minnie"
    }
  },
  {
    "data": {
      "id": "minnie"
    }
  },
  {
    "data": {
      "target": "minnie",
      "id": "2",
      "source": "daysi"
    }
  },
  {
    "data": {
      "target": "minnie",
      "id": "2",
      "source": "daysi"
    }
  },
  {
    "data": {
      "target": "minnie",
      "id": "2",
      "source": "daysi"
    }
  }
]

Whereas I would actually expect to get this:

[
  {
    "data": {
      "id": "donald"
    }
  },
  {
    "data": {
      "id": "daisy"
    }
  },
  {
    "data": {
      "id": "mickey"
    }
  },
  {
    "data": {
      "id": "minnie"
    }
  },
  {
    "data": {
      "target": "donald",
      "id": "3",
      "source": "daysi"
    }
  },
  {
    "data": {
      "target": "mickey",
      "id": "3",
      "source": "minnie"
    }
  },
  {
    "data": {
      "target": "minnie",
      "id": "2",
      "source": "daysi"
    }
  }
]

Can anyone please tell me what I'm doing wrong here?

Answer 1

dict.copy only makes a shallow copy of the dict, the nested dictionaries are never copied, you need deep copies to have those copied over too.

However, you can simply define each new dict at each iteration of the loop and append the new dict at that iteration instead:

for n in nodes_list:
    node_dict = collections.defaultdict(dict) # create new instance of data structure
    node_dict["data"]["id"] = str(n)
    ultimate_list.append(node_dict)

Same applies to the edge_dict:

for e in edges_list:
    edge_dict = collections.defaultdict(dict)
    ...
    ultimate_list.append(edge_dict)

Answer 2

Use copy.deepcopy(your_dict): deepcopy.