如何从 JSON 文件中提取的日期值中删除时间？

Question

I am working in python and using a JSON file and pulling info from it and sending to a csv file. The code I am using is as follows:

import csv
import json

csv_kwargs = {
    'dialect': 'excel',
    'doublequote': True,
    'quoting': csv.QUOTE_MINIMAL
}

inpfile = open('checkin.json', 'r', encoding='utf-8')
outfile = open('checkin.csv', 'w', encoding='utf-8')

writer = csv.writer(outfile, **csv_kwargs, lineterminator="\n")

for line in inpfile:
    d = json.loads(line)
    writer.writerow([d['business_id'],d['date']])

inpfile.close()
outfile.close()

checkin.json key values of business_id and date . The date values are in the form of 'MM:DD:YYYY HH:MM:SS' where it shows the date and then the time. Each business_id includes multiple dates associated with it. I included a line of the JSON file to show how each 'business_id' works and the dates associated with it. A line from the JSON is shown below:

{"business_id":"--1UhMGODdWsrMastO9DZw","date":"2016-04-26 19:49:16, 2016-08-30 18:36:57, 2016-10-15 02:45:18, 2016-11-18 01:54:50, 2017-04-20 18:39:06, 2017-05-03 17:58:02"}

My question is how do you code this to keep the date, but not the time being that they are in the same key value.

Answer 1

You can parse the date in your JSON as a timestamp and then truncate it to date using Python's built-in datetime module.

Import the module:

from datetime import datetime

Parse the date while writing:

for line in inpfile:
    d = json.loads(line)

    dates  = map(lambda dt: datetime.strptime(dt.strip(), '%Y-%m-%d %H:%M:%S').strftime('%Y-%m-%d'), d['dates'].split(' '))
    for date in dates:
       writer.writerow([d['business_id'], date])

Answer 2

If you're strictly using this program to convert the json file to csv, you can simply use string slices:

date, time = d['date'][:12], d['date'][12:] 

If you want to store it as a datetime object to do something else

dt = time.strptime(d['date'], "'%m:%d:%Y''%H:%M:%S'")
# Other stuff
dt_string = dt.strftime("'%m:%d:%Y'")

Answer 3

The formatting for date values described in you question isn't consistent, first you say it's MM:DD:YYYY , however in the sample line from the json input file it appears to be YYYY-MM-DD , and while such details may matter, that particular one doesn't to the revised code below. What did matter was the fact that there can be more than one, which is why I'm updating my answer.

import csv
import json

csv_kwargs = {
    'dialect': 'excel',
    'doublequote': True,
    'quoting': csv.QUOTE_MINIMAL,
}

with open('checkin.json', 'r', encoding='utf-8') as inpfile, \
     open('checkin.csv', 'w', encoding='utf-8', newline='') as outfile:

    writer = csv.writer(outfile, **csv_kwargs)

    for line in inpfile:
        d = json.loads(line)
        # Convert date value string into list of dates with the times removed.
        dates = [date.strip().split(' ')[0] for date in d['date'].split(',')]
        writer.writerow([d['business_id']] + dates)