Python - transforming a parent-child dict-CodePudding

I have a python dict like so that represents a parent-child relationship.

{'children': [{'child': {'name': 'chi1', 'parents': 'par', 'foo1': 'bar1'}}, {'child': {'name': 'chi2', 'parents': 'par', 'foo7': 'bar7'}}, {'child': {'name': 'chi1_chi1', 'parents': 'chi1', 'foo2': 'bar2', 'children': [{'child': {'name': 'chi1_chi1_chil1', 'foo3': 'bar3'}}, {'child': {'name': 'chi1_chi1_chi2', 'foo4': 'bar4'}}]}}, {'child': {'name': 'par', 'foo5': 'bar5'}}, {'child': {'name': 'chi1_chi2', 'parents': 'chi1, chi2', 'foo6': 'bar6'}}]}

Input Dict (show in YAML format):

children:
  - child:
      name: chi1
      parents: par
      foo1: bar1

  - child:
      name: chi2
      parents: par
      foo7: bar7
  - child:
      name: chi1_chi1
      parents: chi1
      foo2: bar2
      children:
        - child:
            name: chi1_chi1_chil1
            foo3: bar3
        - child:
            name: chi1_chi1_chi2
            foo4: bar4
  - child:
      name: par
      foo5: bar5
  - child:
      name: chi1_chi2
      parents: chi1, chi2                 # yay, I have two parents
      foo6: bar6

How do I transform it into a tree-like dict which shows the proper parent-child relationship such that I can get all the children of a parent given the parent's name.

{'children': [{'child': {'name': 'par', 'foo5': 'bar5', 'children': [{'child': {'name': 'chi2', 'foo7': 'bar7', 'parents': 'par', 'children': [{'child': {'name': 'chi1_ch2', 'foo6': 'bar6', 'parents': 'chi1, chi2'}}]}}, {'child': {'name': 'chi1', 'foo1': 'bar1', 'parents': 'par', 'children': [{'child': {'name': 'chi1_chi1', 'foo2': 'bar2', 'parents': 'chi1', 'children': [{'child': {'name': 'chi1_chi1_chi1', 'foo3': 'bar3', 'parents': 'chi1_chi1'}}, {'child': {'name': 'chi1_chi1_chi2', 'foo4': 'bar4', 'parents': 'chi1_chi1'}}]}}, {'child': {'name': 'chi1_chi2', 'parents': 'chi1', 'foo6': 'bar6'}}]}}]}}]}.

Expected result dict (show in YAML format):

children:
  - child:
      name: par
      foo5: bar5
      children:
        - child:
            name: chi2
            foo7: bar7
            parents: par
            children:
              - child:
                  name: chi1_ch2
                  foo6: bar6
                  parents: chi1, chi2
        - child:
            name: chi1
            foo1: bar1
            parents: par
            children:
              - child:
                  name: chi1_chi1
                  foo2: bar2
                  parents: chi1
                  children:
                    - child:
                        name: chi1_chi1_chi1
                        foo3: bar3
                        parents: chi1_chi1
                    - child:
                        name: chi1_chi1_chi2
                        foo4: bar4
                        parents: chi1_chi1
              - child:
                  name: chi1_chi2
                  parents: chi1
                  foo6: bar6

I've been struggling by creating with this for a couple of days to no avail. I'm not sure how to do this programmatically if there are so many levels. Please help. Thanks in advance!

CodePudding user response：

You can use recursion:

import yaml
data = {'children': [{'child': {'name': 'chi1', 'parents': 'par', 'foo1': 'bar1'}}, {'child': {'name': 'chi2', 'parents': 'par', 'foo7': 'bar7'}}, {'child': {'name': 'chi1_chi1', 'parents': 'chi1', 'foo2': 'bar2', 'children': [{'child': {'name': 'chi1_chi1_chil1', 'foo3': 'bar3'}}, {'child': {'name': 'chi1_chi1_chi2', 'foo4': 'bar4'}}]}}, {'child': {'name': 'par', 'foo5': 'bar5'}}, {'child': {'name': 'chi1_chi2', 'parents': 'chi1, chi2', 'foo6': 'bar6'}}]}
def check_children(node):
   if not node['children']:
      del node['children']
   return node

def build_tree(nodes, parent = ''):
   return {'children':[{'child':check_children({**i, 'children':i.get('children', []) build_tree(nodes, i['name'])['children']})} 
           for i in nodes if parent in i.get('parents', '').split(', ')]}

result = build_tree([i['child'] for i in data['children']])
print(yaml.dump(result))

Output

children:
- child:
    children:
    - child:
        children:
        - child:
            children:
            - child:
                foo3: bar3
                name: chi1_chi1_chil1
            - child:
                foo4: bar4
                name: chi1_chi1_chi2
            foo2: bar2
            name: chi1_chi1
            parents: chi1
        - child:
            foo6: bar6
            name: chi1_chi2
            parents: chi1, chi2
        foo1: bar1
        name: chi1
        parents: par
    - child:
        children:
        - child:
            foo6: bar6
            name: chi1_chi2
            parents: chi1, chi2
        foo7: bar7
        name: chi2
        parents: par
    foo5: bar5
    name: par