Python - convert csv file to JSON

Question

I need to convert a csv file into a hierarchical JSON object (preferably using Python). I thought that the script I have (below) does a correct job of converting to JSON, bu

Answer 1

Order of elements is not important for json.

The json you are producing is wrong, though. The example you gave that you are trying to match does not have a "subject" key at the top level. It shows just "name" => "flare", and a list of children. Your script is producing json with a "subject" key.

Instead of

subject_dict = {'name':subject,'type':'subject','children':branch_list}

try

subject_dict = {'name':subject,'children':branch_list}

I don't know about D3.js, so I can't say if it expects an object that follows one exact structure.

Tip: one good tool to check the structure of json objects is http://jsoneditor.appspot.com. Just paste your string in there and click on "Go to Treeview". It will hellp you find differences between what you have and what you want very easily.

Answer 2

import csv
import json
file1 = csv.DictReader(open('filename.csv', 'r'))
output =[]
for each in complent:
    row = {}
    row['Id'] = each['Id']
    row['Name'] = each['Name']
    row['Address'] = each['Address']
    row['Mobile'] = each['Mobile']
    row['LandLine'] = each['LandLine']
    row['Email'] = each['Email']
    output.append(row)

json.dump(output,open('new_file.json','w'),indent=4,sort_keys=False)

Answer 3

D3 expects the value in JSON to also be in double quotes. so e.g. "level": 3 should be "level" : "3". When you use this data in D3 for calculation it will need to be a number again though so where you use it you can just use +data.xxx.yyy.level instead of data.xxx.yyy.level in the d3 code to turn it into a number.

Answer 4

Wouldn't this work?

import json

f = open('path/to/file','r')

arr=[]
headers = []

for header in f.readline().split(','):
  headers.append(header)

for line in f.readlines():
  lineItems = {}
  for i,item in enumerate(line.split(',')):  
    lineItems[headers[i]] = item
  arr.append(lineItems)

f.close()

jsonText = json.dumps(arr)

print jsonText