Slideshow

Working with Time

The time module

Let's go straight in and see some examples of what the time module can do:

import time
time.time()

    1607080140.0588841

Scary! This was the epoch, also known as Unix time, at the moment the code was executed, and is the number of seconds since midnight on January 1st, 1970 in UTC / GMT timezone.

Friendlier time formats

billionth = time.localtime(1000000000)
billionth

    time.struct_time(tm_year=2001, tm_mon=9, tm_mday=9, tm_hour=2, tm_min=46, tm_sec=40, tm_wday=6, tm_yday=252, tm_isdst=1)
  • the localtime() function, for example, takes an epoch and returns a struct_time object which has various attributes likely to be of use
time.strftime("%a, %d %b %Y %H:%M:%S %Z", billionth)

    'Sun, 09 Sep 2001 02:46:40 BST'
  • the .strftime() or 'string formatted time method can be used to make time values more human-readable, with great flexibility over the formatting

Code walkthrough

Let's make use of the ability to record timestamps in our Spotter class definition, to automatically renew the access_token if it has expired.

Fork the repl if you'd like to try it out yourself, remembering to create your own new .env file.

Nothing new to see in main.py, but there is in spotter.py, where we have imported time and made changes to the get_token() and get_item() methods.

get_token()

...
self.token_request_timestamp = time.time()
...
  • before making the API call, we log the current time
...
self.token_expiry = self.token_request_timestamp + token_info['expires_in']
self.token_expiry_local = time.strftime("%a, %d %b %Y %H:%M:%S %Z", \
                                        time.localtime(self.token_expiry))
...
  • we fetch the 'expires_in' value from the response and calculate the time the token will expire, assigning the result to the .token_expiry attribute of the object
  • we also assign a more readable version to token_expiry_local

Recording the time before making the API call and then adding the expires_in value protects us from the possibility of any 'gap' due to latency between the server response and our program receiving it.

get_item()

if time.time() > self.token_expiry:
print('Getting new token...')
self.get_token()
  • if the current time when get_item() is called is later than the token_expiry, get_token() will be called again and the token and token_expiry will be updated

As a result, we can use the get_item() method of any instance of Spotter for as long as we like, without needing to manually renew the token or create a new object to make more API calls.