convert python string date to mysql datetime

Question

I have a string field witch is a date scraped from internet and reads like this:

 "Lun Ene 27, 2014 9:52 am", "Mar Feb 11, 2014 5:38 pm",...

Lun= day of week (lunes/monday) Ene = month (enero/january)

I need to enter them in a mysql table, in a datetime field.

'YYYY-MM-DD HH:MM:SS'

I imagine it is a very common issue and was wondering if someone already have a script to do it or could point out where I could look for...

Thanks in advance!

Answer 1

month_of_the_year = ['Ene', 'Feb', 'Mar', 'Abr', 'May', 'Jun', 'Jul', 'Ago', 'Sep', 'Oct', 'Nov', 'Dec']

def convert_to_mysql_format(string):
  explode = string.split()
  day_of_the_month = explode[2][:-1]
  if int(explode[2][:-1]) < 10:
    day_of_the_month = "%02d" % (int(explode[2][:-1]),)

  if explode[5] == 'am':
    time_split = explode[4].split(':')
    if time_split[0] == '12':
      time_split[0] = '00'
    elif int(time_split[0]) < 10:
      time_split[0] = "%02d" % int(time_split[0])

  else:
    time_split = explode[4].split(':')
    if int(time_split[0]) in range(1, 12):
      time_split[0] = str(int(time_split[0]) + 12)


  if month_of_the_year.index(explode[1]) < 12:
    explode[1] = "%02d" % (month_of_the_year.index(explode[1])+1)

  return explode[3]+'-'+explode[1]+'-'+day_of_the_month+' '+time_split[0]+':'+time_split[1]+':00'

print convert_to_mysql_format("Lun Ene 27, 2014 9:52 am")
print convert_to_mysql_format("Lun Ene 27, 2014 9:52 pm")

2014-01-27 09:52:00
2014-01-27 21:52:00

Answer 2

By default, Python runs using C locale:

>>> from datetime import datetime
>>> datetime.strptime("Tue Feb 11, 2014 5:38 PM", "%a %b %d, %Y %I:%M %p")
datetime.datetime(2014, 2, 11, 17, 38)
>>> import locale
>>> locale.nl_langinfo(locale.T_FMT_AMPM)
'%I:%M:%S %p'

Changing locale partially helps on my system:

>>> locale.setlocale(locale.LC_TIME, 'es_ES.UTF-8')
'es_ES.UTF-8'
>>> datetime.strptime("Lun Ene 27, 2014 9:52 am"[:-2], "%a %b %d, %Y %I:%M %p")
datetime.datetime(2014, 1, 27, 9, 52)
>>> locale.nl_langinfo(locale.T_FMT_AMPM)
''

T_FMT_AMPM is not set for some reason on my system for es_ES.UTF-8. To fix it, you could manually add 12 hours if the time string ends with 'pm'.

The strftime() and time behaviour is the same.

Note: the locale name may be different on other systems.