Question

I am trying to use the templatetag described in SO answer: https://stackoverflow.com/a/6217194/493211 in a project using Django 1.4.3 (with Python 2.7.2).

I adapted it like this:

from django import template


register = template.Library()

@register.filter
def template_exists(template_name):
    try:
        template.loader.get_template(template_name)
        return True
    except template.TemplateDoesNotExist:
        return False

So that I could use it like this in another template:

{% if 'profile/header.html'|template_exists %}        
  {% include 'profile/header.html' %}
{% else %}
  {% include 'common/header.html' %}
{% endif %}

This way, I could have avoided using solutions such as changing the order of my apps in INSTALLED_APPS.

However, it does not work. If the template does not exist, then the exception is raised within the stack/console but it is not propagated up to get_template(..) (from inside this statement), and thus not to my foolish API. Hence, this blows up in my face during the rendering. I uploaded the stacktrace to pastebin

Is this a wanted behavior from Django?

I ended up stop doing foolish things as is. But my question would remain.

Answer 1

What about a custom tag? This doesn’t provide the full functionality of include but seems to meet the needs in the question.:

@register.simple_tag(takes_context=True)
def include_fallback(context, *template_choices):
    t = django.template.loader.select_template(template_choices)
    return t.render(context)

Then in your template:

{% include_fallback "profile/header.html" "common/header.html" %}