News Classifier Datasets from json to csv format => https://www.kaggle.com/rmisra/news-category-dataset

news_jsont_to_csv.py

"""Convert Data from .json to .csv easily readable by Pandas"""
import os
import sys
import ast
from pathlib import Path
from typing import List, Dict

import pandas as pd

# Set directory to file directory so other paths are easily relative to it without error
filename = sys.argv[0]
dir_name = os.path.dirname(filename)
os.chdir(dir_name)

with open(Path('../data/News_Category_Dataset_v2.json', mode='r')) as f:
    data = f.readlines()
  
data_dict: Dict[str, List] = {
    'categories': [],
    'headlines': [],
    'authors': [],
    'links': [],
    'short_descriptions': [],
    'dates': []
}

for line in data:
    line = ast.literal_eval(line)
    data_dict['categories'].append(line.get('category', 'Null'))
    data_dict['headlines'].append(line.get('headline', 'Null'))
    data_dict['authors'].append(line.get('authors', 'Null'))
    data_dict['links'].append(line.get('link', 'Null'))
    data_dict['short_descriptions'].append(line.get('short_description', 'Null'))
    data_dict['dates'].append(line.get('date', 'Null'))
    
df = pd.DataFrame(data_dict)
# Insert None for error in values
df.to_csv(Path('../data/News Category.csv'), index=False, errors=None)