Matt Healy - PropTech CTO
Technology Leader, Solutions Architect and Business Analyst
Perth, Western Australia
 

I recently wrote an API endpoint for a project at VaultRealEstate which required me to generate a month-by-month commission summary breakdown for a sales agent. The API endpoint accepts an arbitrary start and end month and should return a JSON object showing all the distinct months in that date range and the commission performance for those months (for example, for displaying a bar chart).

I found that this was actually not a trivial problem, and built-in Python libraries like datetime can't really solve this problem nicely. There are third party libraries like dateutil which can solve the problem, and normally I would immediately go for a library like this. However, this project is hosted on AWS Lambda and I'm conscious about the deployment size of the project getting bigger with each dependency introduced, so I really like to only introduce a dependency when it's necessary.

I found the following StackOverflow answer which seemed to suit my needs:

https://stackoverflow.com/a/34898764/272193

Adapting the answer for my use case results in the below snippet.

from datetime import datetime, timedelta
from collections import OrderedDict

# Sample start and end dates
start = datetime(year=2017, month=10, day=1)
end = datetime(year=2018, month=3, day=1)

# Get list of months >= start and < end

months = OrderedDict(((start + timedelta(_)).strftime("%Y-%m-01"), 0) for _ in range((end - start).days)) 
# OrderedDict([('2017-10-01', 0), ('2017-11-01', 0), ('2017-12-01', 0), ('2018-01-01', 0), ('2018-02-01', 0)])