Pretty Print Relative Dates in Python
In this post, we will learn how to “pretty print” relative dates using the dateutil library in Python. Specifically, given a relativedelta object, we will print a string that represents the number of years, months, and days that have passed between two dates.
Introduction
I am working on a project called ManagerManager that helps me keep track of the people I work for. As my team has grown, it has become important to me to have a quick reminder to celebrate milestones such as work anniversaries.
PostgreSQL age Function
PostgreSQL has a useful built-in age function that allows you to get a human readable relative date from a date or datetime column.
For example, If I was in a database somewhere, I could use the age() function to find out how old I am.
select name, birthday, age(birthday) from person;| name | birthday | age |
|---|---|---|
| Lev | 1988-06-09 | 31 years 11 mons 6 days |
Abstracting Away the SQL
In the first iteration of this project, I relied on “raw” SQL for all database operations. This made it easy to take advantage of the age() function in PostgreSQL. As the project became more complex, working in SQL got out of hand.
I recently switched to SQLAlchemy to make working with the database simpler. Although SQLAlchemy allows you to call out to native database functions, I didn’t want to have a hard dependency on PostgreSQL for this application, so I started to search for a quick and dirty way to do the same thing in Python.
Exploring Python datetime
My initial investigation led me to the familiar datetime module in Python. This built-in module is great for manipulating date and time objects. It includes a timedelta class which allows you to compare two datetimes at up to microseconds of resolution. This is generally useful, but unfortunately the maximum interval that this class provides is days.
import datetime
dob = datetime.date.fromisoformat('1988-06-09')
today = datetime.date.today()
today - dob
>>> datetime.timedelta(days=11662)It turns out converting days to years, months, and weeks is much more complicated than I initially thought since you need to account for leap years and other nuances.
The dateutil Library
Luckily, the dateutil library provides several enhancements to the built-in datetime module. Most importantly for this use case, it has the relativedelta class which breaks the timedelta into years, months, days, all the way down to microseconds.
import datetime
from dateutil.relativedelta import relativedelta
relativedelta(today, dob)
dob = datetime.date.fromisoformat('1988-06-09')
today = datetime.date.today()
>>> relativedelta(years=+31, months=+11, days=+6)The only thing missing from dateutil.relativedelta is a nice way to “pretty print” the result. For example, what if years, months, or days is equal to 0? I don’t want a string that says “0 years, 4 months, and 1 day”.
Ternary Operators to the Rescue
This was the first time that I’ve used a ternary operator in python and it actually feels like a good use case here.
def tenure(self):
rd = relativedelta(datetime.now(), self.start_date)
years = f'{rd.years} years, ' if rd.years > 0 else ''
months = f'{rd.months} months, ' if rd.months > 0 else ''
days = f'{rd.days} days' if rd.days > 0 else ''
return f'{years}{months}{days}'A ternary operator is a shortcut to write an if/else statement. In my opinion, it is useful for simple situations like this. Let’s break this down a little bit.
In the tenure() function described above, the code for determining years is equivalent to the following:
if years > 0:
years = f'{rd.years} years, '
else:
years = ''Or, you could also write it like this:
years = ''
if years > 0:
years = f'{rd.years} years, 'However, I think the simple one-liner ternary operator makes the entire function a bit cleaner.
Going back to the birthday example, the final output of this function unsurprisingly looks like this:
31 years, 11 months, 6 days
Conclusion
To sum things up, we can use the dateutil library coupled with your choice of if/else syntax to provide a human readable way to show the difference between two dates.
Thank you for reading! Share your thoughts with me on bluesky, mastodon, or via email.
Check out some more stuff to read down below.
Most popular posts this month
- SQLite DB Migrations with PRAGMA user_version
- My Custom Miniflux CSS Theme
- Convert Markdown to PDF in Sublime Text
- Making cgit Pretty
- Using cgit
Recent Favorite Blog Posts
This is a collection of the last 8 posts that I bookmarked.
- My Running Tips from Kevin Bell's Blog
- tweet from Derek Sivers blog
- Rewrote my blog with Zine from Drew DeVault's blog
- A eulogy for Vim from Drew DeVault's blog
- Pluralistic: AI "journalists" prove that media bosses don't give a shit (11 Mar 2026) from Pluralistic: Daily links from Cory Doctorow
- Offline 23 hours a day from Derek Sivers blog
- Pluralistic: California can stop Larry Ellison from buying Warners (28 Feb 2026) from Pluralistic: Daily links from Cory Doctorow
- On Alliances from Smashing Frames
Articles from blogs I follow around the net
i’m calling it ‘wil wheatcon’ until i can think of something better
In an average year, I travel to around 5 or 6 cities for conventions. Almost every time I announce an appearance, the most common response is some version of “that’s great! When are you coming to [my town]?” I’m not coming to your town, but I am coming to...
via WIL WHEATON dot NET May 20, 2026On people writing about their use of AI
I find the trend of people posting about the way they use generative AI to be fascinating at an anthropological level. I do not remember the last time a piece of technology pushed so many different people into writing about the way they use it, or not use...
via Manuel Moreale — Everything Feed May 20, 2026Exporting Vinted Sold Data
A little javascript snippet to grab Vinted sales data from the website
via Robb Knight • Posts • Atom Feed May 20, 2026Generated by openring