In some cases, Jinja templates become too complicated. Lots of deeply nested if statements, clunky ways of working with variables, macros and many other things that hurt the eyes.

Jinja logo

What might be worth knowing is the fact that you can pass a Python function into your Jinja templates. Doing this can greatly improve the readability of your template as well as allow you to handle more complicated scenario’s.

Using a Python function in a Jinja template

The following Python will pass 2 functions to the Jinja template.

First we define 2 functions: ‘hello_world’ and ‘multiply’. These functions are placed in a dictionary.

After this, the render function is created. Inside this function, we use jinja_template.globals.update(func_dict) to pass the previously created func_dict and expose it during the Jinja rendering phase.

The function ends up rendering the template and returning the resulting string:

# !/usr/bin/env python3
from jinja2 import Environment, FileSystemLoader

def hello_world():
    return "hello world from within the function"

def multiply(x, y):
    return str(x * y)

func_dict = {
    "hello_world": hello_world,
    "multiply": multiply,

def render(template):
    env = Environment(loader=FileSystemLoader("/srv/templates/"))
    jinja_template = env.get_template(template)
    template_string = jinja_template.render()
    return template_string

if __name__ == "__main__":

In the following example Jinja, /srv/templates/test.j2, we use the functions that our previous Python passes into the template:

Calling the 'hello_world' function:
{{ hello_world() }}

Calling the 'multiply' function:
{{ multiply(2, 2) }}

When we run the Python script, we get the following result:

Calling the 'hello_world' function:
hello world from within the function

Calling the 'multiply' function:

Being able to use Python functions like this inside a Jinja template has been very useful to me. It increased the readability of some templates as I was able to replace large parts with a straightforward Python functions.

It also allowed me to handle more complex logic in a Python function. This allowed me to introduce more complicated configurations as well make templates behave based on the state of other systems.

It is not something that is required all the time, but it is nice to know that it is a possibility.