How to infer type from datetime string in Python

Question

I have a table for which I want to infer custom types for each column. Given a string, I want to distinguish whether it is a date, a time or a timestamp. For instance, for the given list:

input_list = [
    "4th Sep 2013",
    "95-04-04 23:33:05",
    "May 5th 1999",
    "03:00:00",
    "10-01-22",
    "12-12-25 05:48:01.0010"]

I want an output that tells me (a list of strings)

output_types = [
    "date",
    "timestamp",
    "date",
    "time",
    "date",
    "timestamp"]

I have tried using datetime , but since I can't be sure of the input format (especially for dates) it is hard to infer automatically the input type.

Answer 1

The parsedatetime module seems to provide this functionality, in the parsedatetime.Calendar.parse() method. According to the documentation:

If the datetimeString is parsed and date/time value found then the second item of the returned tuple will be a flag to let you know what kind of struct_time value is being returned:

 0 = not parsed at all 1 = parsed as a date 2 = parsed as a time 3 = parsed as a datetime 

I don't know if it supports every date/time format that you've given in the examples.

Answer 2

Python supports posix from an OS level. You can use posix to distinguish between the different string types.

You can find more relevant details here : https://docs.python.org/2/library/posix.html