JSON like data to CSV file in python - not showing headers correctly

Question

I am transforming JSON like data to CSV and having a few issues.

The code is here:

import json
import csv

def parse_file(inputed_file):
    with open(input_file, 'r') as inputed_file:
        content = inputed_file.readlines()
        split_file = open('test.csv', 'w')

        for line in content:
            lines = line.split('\t')

            data = json.loads(lines[0])
            writer = csv.DictWriter(split_file, fieldnames = ["title", "firstname"], delimiter = ',')
            writer.writeheader()

The problem is this is adding a header on each row for the data, I want to only have the header displayed once. Then add this for the data to go below the headers:

writer.writerow(data)

I have looked at this and tried it but failed: How can I convert JSON to CSV?.

Barmar · Accepted Answer · 2016-12-08T11:27:31.983

0

Create the DictWriter outside the loop, and just call writer.writeheader() there. Then call writer.writerow() inside the loop.

def parse_file(inputed_file):
    with open(input_file, 'r') as inputed_file:
        content = inputed_file.readlines()
        split_file = open('test.csv', 'w')
        writer = csv.DictWriter(split_file, fieldnames = ["title", "firstname"], delimiter = ',')
        writer.writeheader()

        for line in content:
            lines = line.split('\t')
            data = json.loads(lines[0])
            writer.writerow(data)

edited Dec 08 '16 at 11:27

answered Dec 08 '16 at 11:20

Barmar

741,623
53
500
612

thanks but this still isn't adding the data below the headers :( Please help – Sam Dec 08 '16 at 11:24
I've changed to a simpler answer – Barmar Dec 08 '16 at 11:27
Thank you so much, this works, the only thing is it's putting an empty line between each row. Do you know how I can change this? – Sam Dec 08 '16 at 11:39
I don't think it does that automatically. Maybe your JSON data has newlines in it? – Barmar Dec 08 '16 at 11:42

JSON like data to CSV file in python - not showing headers correctly

1 Answers1