json_to_csv.py

import os
import sys
import csv
import json
from collections import defaultdict


# value_types = [str, float, int]


def read_file(file_to_read):
    data = []
    counter = 0
    with open(file_to_read, encoding='utf-16') as file:
        for line in file:
            if counter == 0 and not line.startswith("["):
                line = "[" + line
                counter += 1
            data.append(line)

    data.append("]") if not data[-1].endswith(']') else None

    return data


if len(sys.argv) == 1:
    sys.exit('\n \n No input file name given as parameter \n')

else:
    files = []

    if len(sys.argv) == 2 and sys.argv[1] == "all":
        contents = os.listdir(os.path.dirname(os.path.abspath(__file__)))
        for item in contents:
            if item.endswith(".json") or item.endswith(".txt"):
                files.append(item)

    elif len(sys.argv) >= 2:
        files = [item for item in sys.argv[1:] if
                 os.path.isfile(os.path.join(os.path.dirname(os.path.abspath(__file__)), item)) and (
                 item.endswith('.json') or item.endswith(".txt"))]

    sys.exit('No file present') if len(files) == 0 else None

    for input_file in files:
        print("\n Processing file : ", input_file)
        data = read_file(input_file)

        str_data = ""
        for i in data:
            if not i == "\n":
                str_data += i

        if "}\n{" in str_data:
            str_data = str_data.replace("}\n{", "},\n{")
        if "} {" in str_data:
            str_data = str_data.replace("} {", "},\n{")

        json_data = json.loads(str_data)

        dict_objects_list = []

        # currently only for one level down the top level eg: {"name": "john", "phone" :{"mobile": XXXX, "home": ZZZZ } }
        for item in json_data:
            if isinstance(item, dict):
                per_object_dict = defaultdict(str)
                for key, value in item.items():
                    if not isinstance(value, dict):
                        per_object_dict[key] = str(value)
                    elif isinstance(value, dict):
                        for inner_key, inner_value in value.items():
                            per_object_dict[key] = str(inner_value)

                dict_objects_list.append(per_object_dict)

        keys = set(key for key in dict_objects_list[0].keys())

        # print('\n \n  keys : \n', sorted(keys))

        data = [[obj_dict[key] for key in sorted(keys)] for obj_dict in dict_objects_list]

        with open(input_file.split(".")[0] + '_output.csv', 'w', encoding='utf-8', newline='') as output:
            writer = csv.writer(output, delimiter=',')
            header = [key for key in sorted(keys)]
            writer.writerow(header)
            for item in data:
                writer.writerow(item)