CODEWITHSUNDEEP
pythonjson
Visualisation App
Visualisation App

Reputation: 788

Printing parent data along with child data of nested JSON

I have a JSON of the following sort. In another table, I have JSONPaths, which tells me to get certain values. If I am to get the values of child elements, I need to get all the corresponding parent attribute values and store all of these values in a dataframe as a single row.

{
  "Parent": {
    "Name": "Bob",
    "Age": "80",
    "Children": [
      {
        "Name": "Michael",
        "Gender":"M",
        "Children": [
          {
            "Name": "Ezee",
            "Gender": "M",
            "Age": 20
          },
          {
            "Name": "Ezee",
            "Gender": "M",
            "Age": 28,
            "Children": [
              {
                "Name": "Dre",
                "Age": 1
              },
              {
                "Name": "George",
                "Age": 2
              }
            ]
          }
        ],
        "Age": 50,
        "MiddleName": "Jay"
      },
      {
        "Name": "Justin",
        "Gender": "M",
        "Children": [
          {
            "Name": "Emily",
            "Age": 18,
            "Gender": "F"
          }
        ],
        "Age": 45
      }
    ]
  }
}

Say I need to get the values for the JSON path: Parent/Children/0/Children/0/Name, I need to get the attribute values of name, age, etc. for the corresponding parent (Parent/Children/0/[Name or Age or etc.]) and store all the above values as a single row.

Currently, I am able to get the parent values separately by passing the related path and children values separately by passing that path.

def findValue(path, json_data):
    paths = path.split("/")    
    data = json_data
    for i in range(0,len(paths)):
        if isinstance(data, list):
            paths[i]=int(paths[i])
            data = data[paths[i]]
        else:
            data = data.get(paths[i])
    return data

How could I achieve this?

Upvotes: 0

Views: 1349

Answers (2)

Joe Iddon
Joe Iddon

Reputation: 20424

To get the last parent, you need to traverse your path up to the last time you follow a Children list.

I.e. Given your path: 'Parent/Children/0/Children/0/Name' you want to return the data for the parent at 'Parent/Children/0'.

This is easy enough to do in Python, just slice the path string up to the last occurrence of the substring /Children:

path[:path.rfind('/Children')]

You can then use similar code to what you currently have to get the parent's data:

parent = json_data
path = path[:path.rfind('/Children')]
for attr in path.split('/'):
    parent = parent[int(attr) if isinstance(parent, list) else attr]

which, for this example, would give us parent as:

{
  "Name": "Michael",
  "Gender": "M",
  "Children": [
    {
      "Name": "Ezee",
      "Gender": "M",
      "Age": 20
    },
    {
      "Name": "Ezee",
      "Gender": "M",
      "Age": 28,
      "Children": [
        {
          "Name": "Dre",
          "Age": 1
        },
        {
          "Name": "George",
          "Age": 2
        }
      ]
    }
  ],
  "Age": 50,
  "MiddleName": "Jay"
}

And to complete your question fully, if you wanted this person's attributes (without their list of Children) as a single row, you would have to decide to store them in a set way (such as alphabetically) and then you could use the .items() method of a dict to extract these into the right format:

[v for k,v in sorted(t for t in parent.items() if t[0] != 'Children')]

giving, for our example:

[50, 'M', 'Jay', 'Michael']
#Age, Gender, Middle Name, Name

Oh, and the whole of the first code can be compressed to a one-liner if you wanted:

__import__('functools').reduce(lambda d,a:d[int(a) if isintance(d,list) else a], path[:path.rfind('/Children')].split('/'), json_data)

Upvotes: 1

Thibault D.
Thibault D.

Reputation: 1637

If I understand well, all you want is given a path like Parent/Children/0/Children/0/Name get the path for the same property of its parent. In this case it would be Parent/Children/0/Name

Here is my try on a python interpreter, hope it helps you :

>>> path = "Parent/Children/0/Children/0/Name"
>>> path_l = path.split('/')
>>> rev = path_l[::-1]
>>> rev
['Name', '0', 'Children', '0', 'Children', 'Parent']
>>> rev.index('Children')
2
>>> rev = rev[rev.index('Children')+1:]
>>> rev
['0', 'Children', 'Parent']
>>> final = rev[::-1] + [path_l[-1]]
>>> final
['Parent', 'Children', '0', 'Name']
>>> parent_path = '/'.join(final)
>>> parent_path
'Parent/Children/0/Name'

Then using your function you can add your two value to a dataframe

>>> df = pandas.DataFrame({'Parent': [], 'Children':[]})
>>> df.append([parent], [children])

Upvotes: 1

Related Questions