CODEWITHSUNDEEP
pythonpython-3.x
Sebastian
Sebastian

Reputation: 1101

How to add trailing zeroes to datetime as string?

What's an easy way to modify the following timestamps (as strings) like this:

2020-11-04T19:48:19.9 --> 2020-11-04T19:48:19.90
2020-11-04T19:48:20 --> 2020-11-04T19:48:20.00

A bit of background: I have timestamps coming from an API that typically look like this:

2020-11-04T19:48:19.87

I want to convert them to datetime, which can be done with the datetime module:

my_ts = datetime.datetime.strptime(my_ts, '%Y-%m-%dT%H:%M:%S.%f')

Now %f expects 6 digits, which is easily achieved by:

my_ts = my_ts + '0000'

However, for my data the fraction-bit may be missing a trailing zero, or missing altogether, like this:

2020-11-04T19:48:19.9
2020-11-04T19:48:20

What's an easy way to achieve this?

What I currently do is this:

teststring = "2020-11-04T19:48:19"

if '.' not in teststring:
    fraction_length = 0
else:
    fraction_length = len(teststring.split(".",1)[1])

if fraction_length == 0:
    new_teststring = teststring + '.00'
elif fraction_length == 1:
    new_teststring = teststring + '0'
else:
    new_teststring = teststring

This appears to be unnecessarily complicated. There has to be a better way, doesn't it?

Any hints are greatly appreciated.

Upvotes: 4

Views: 1084

Answers (3)

Pedro Rodrigues
Pedro Rodrigues

Reputation: 2658

Try this.

def f(string, suffix='.00', length=22):
    if len(string) == length:
        return string
    return string + suffix[-(length - len(string)):]

suite = {
    '2020-11-04T19:48:19.9': '2020-11-04T19:48:19.90',
      '2020-11-04T19:48:20': '2020-11-04T19:48:20.00',
   '2020-11-04T19:48:20.00': '2020-11-04T19:48:20.00',
}

for case, expected in suite.items():
    assert f(case) == expected

Check it at colab.google.com

Upvotes: 1

itprorh66
itprorh66

Reputation: 3288

try using the dateutil library

as an example:

from dateutil import parser
dt1 = ['2020-11-04T19:48:19.9', '2020-11-04T19:48:20', '2020-11-04T19:48', '2020-11-04T19:48:20.01']

for dt in dt1:
    print(parser.parse(dt))

yields:

2020-11-04 19:48:19.900000
2020-11-04 19:48:20
2020-11-04 19:48:00
2020-11-04 19:48:20.010000

Upvotes: 2

martineau
martineau

Reputation: 123423

The problem seems to boil down into needing to handle the cases where there's no decimal point or there is but there are no digits following it — otherwise strptime() can handle it.

The code below uses a regular expression to determine which of these two cases has occurred to modify the timestamps string accordingly when necessary.

import datetime
import re

tests = (
    '2020-11-04T19:48:19',
    '2020-11-04T19:48:19.',
    '2020-11-04T19:48:19.9',
    '2020-11-04T19:48:19.90',
    '2020-11-04T19:48:19.900',
    '2020-11-04T19:48:19.90000',
    '2020-11-04T19:48:19.900000',
)

regex = re.compile(r'(\.\d*)')

for test in tests:
    try:
        my_ts = datetime.datetime.strptime(test, '%Y-%m-%dT%H:%M:%S.%f')
    except ValueError:
        m = regex.search(test)
        test += '0' if m else '.0'
        my_ts = datetime.datetime.strptime(test, '%Y-%m-%dT%H:%M:%S.%f')

    print(my_ts)

Output:

2020-11-04 19:48:19
2020-11-04 19:48:19
2020-11-04 19:48:19.900000
2020-11-04 19:48:19.900000
2020-11-04 19:48:19.900000
2020-11-04 19:48:19.900000
2020-11-04 19:48:19.900000

Upvotes: 1

Related Questions