strptime() with varying input data

Question

Some stable genius before my time wrote their dates into the database as strings. Not only that, but the format varies too. I have at least

05.08.1993

and

31.08.1993 00:00:00

Now I want to convert those to datetime objects using strptime(), e.g.

my_date = datetime.datetime.strptime(my_value, '%d.%m.%y')

So now I have to deal with the different formats. I can think of varous clumsy ways to do this, invlvong ugly cascades of ifs or trys. Is there an elegant way to do this?

I would suggest using RegEx to determine which format you got at Hand. You could also try using `dateutil.parser.parse()` https://stackoverflow.com/a/23277356/3858121 — Japu_D_Cret, Apr 13 '18 at 11:59

Rakesh · Answer 1 · 2018-04-13T12:08:26.703

3

I would suggest you use python dateutil

Demo:

from dateutil.parser import parse

a = "05.08.1993"
b = "31.08.1993 00:00:00"

print(parse(a).strftime('%d.%m.%y'))
print(parse(b).strftime('%d.%m.%y'))

Output:

08.05.93
31.08.93

Using Regex:

import re
import datetime

a = "05.08.1993"
b = "31.08.1993 00:00:00"
d = re.search("\d{2}.\d{2}.\d{4}", a)
if d:
    print(datetime.datetime.strptime(d.group(),'%d.%m.%Y').strftime('%d.%m.%y'))

d = re.search("\d{2}.\d{2}.\d{4}", b)
if d:
    print(datetime.datetime.strptime(d.group(),'%d.%m.%Y').strftime('%d.%m.%y'))

Output:

05.08.93
31.08.93

edited Apr 13 '18 at 12:08

answered Apr 13 '18 at 11:59

Rakesh

81,458
17
76
113

Looks good. Unfortunately, that's not in the standard libray, so getting this installed here is going to take me days or even weeks. – uwain12345 Apr 13 '18 at 12:31
@uwain12345 the regex should help then. – Rakesh Apr 14 '18 at 00:13

score 1 · Answer 2 · edited Jun 20 '20 at 09:12

1

dateutil.parser:

This module offers a generic date/time string parser which is able to parse most known formats to represent a date and/or time.

This module attempts to be forgiving with regards to unlikely input formats, returning a datetime object even for dates which are ambiguous.

edited Jun 20 '20 at 09:12

Community

1
1

answered Apr 13 '18 at 11:59

Maxim Egorushkin

131,725
17
180
271

floatingpurr · Answer 3 · 2018-04-13T12:05:59.617

0

For the 2 datatypes you mentioned, it should be sufficient to harness slicing:

my_date = datetime.datetime.strptime(my_value[:10], '%d.%m.%Y')

for a more robust approach, you can use generic date parsers, as other answers mention.

edited Apr 13 '18 at 12:05

answered Apr 13 '18 at 12:00

floatingpurr

7,749
9
46
106

score 0 · Answer 4 · answered Apr 13 '18 at 12:09

If it are only 2 patterns, this can be done with re

dates = ('05.08.1993', '31.08.1993 00:00:00')
pattern = re.compile(r'\d{2}\.\d{2}\.\d{4} \d{2}:\d{2}:\d{2}')
strp_pattern = '%d.%m.%Y %H:%M:%S'
strp_pattern_fallback = '%d.%m.%Y'

def parse_dates(dates):
    for date in dates:
        if pattern.match(date):
            yield datetime.datetime.strptime(date, strp_pattern)
        else:
            yield datetime.datetime.strptime(date, strp_pattern_fallback)

list(parse_dates(dates))

[datetime.datetime(1993, 8, 5, 0, 0), datetime.datetime(1993, 8, 31, 0, 0)]

strptime() with varying input data

4 Answers4