Skip to content

Latest commit

 

History

History
271 lines (187 loc) · 8.29 KB

README.rst

File metadata and controls

271 lines (187 loc) · 8.29 KB
Raw

django-xinclude

PyPI version

Render a template using htmx with the current context.

hx-get is often used to delegate potentially computationally expensive template fragments to htmx. Achieving this sometimes requires more views, each of which needs to inherit from mixins that provide access to the same context.
django-xinclude provides a template tag that aims to make this easier by leveraging the cache.

Please note that this package is in a very early, exploratory stage.

Requirements

  • Python 3.10 to 3.12 supported.
  • Django 4.2 to 5.0 supported.
  • htmx

Setup

  • Install from pip:
python -m pip install django-xinclude
  • Add it to your installed apps:
INSTALLED_APPS = [
    ...,
    "django_xinclude",
    ...,
]
  • Include the app URLs in your root URLconf:
from django.urls import include, path

urlpatterns = [
    ...,
    path("__xinclude__/", include("django_xinclude.urls")),
]

You can use a different prefix if required.

Usage

Once installed, load the xinclude library and use the tag passing the template that you want to include:

{% load xinclude %}

{% xinclude "footer.html" %}{% endxinclude %}

Every feature of the regular include tag is supported, including the use of with and only.

You can use the following htmx-specific arguments:

  • hx-trigger: corresponds to the hx-trigger htmx attribute. Defaults to load once.
  • swap-time: corresponds to the swap timing of the hx-swap htmx attribute.
  • settle-time: corresponds to the settle timing of the hx-swap htmx attribute.

"Primary nodes" may be passed along to render initial content prior to htmx swapping. For example:

{% xinclude "footer.html" %}
    <div>Loading...</div>
{% endxinclude %}

django-xinclude plays well with the excellent django-template-partials package, to select specific partials on the target template.

Advanced usage

Below is a more complete example making use of the htmx transition classes. Note the intersect once trigger, which will fire the request once when the element intersects the viewport.

<style>
.htmx-swapping > #loading {
    opacity: 0;
    transition: opacity 1s ease-out;
}
</style>

{% xinclude "magic.html" with wand="🪄" hx-trigger="intersect once" swap-time="1s" settle-time="1s" %}
    <div id="loading">
        Loading...
    </div>
{% endxinclude %}

magic.html:

<style>
#items.htmx-added {
    opacity: 1;
    animation: appear ease-in 500ms;
}
</style>

<div id="items">
    🔮 {{ wand }}
</div>

You can preload the xinclude libary in every template by appending to your TEMPLATES builtins setting. This way you don't need to repeat the {% load xinclude %} in every template that you need the tag:

TEMPLATES = [
    {
        "BACKEND": "django.template.backends.django.DjangoTemplates",
        # ...,
        "OPTIONS": {
            "builtins": [
                "django_xinclude.templatetags.xinclude",
            ],
        },
    },
]

How It Works

django-xinclude first checks if it needs to render the target template synchronously; see the Section below for cases where this might be useful. If this is not the case, it stores the current context and the target template to the cache and constructs a url with a fragment_id that targets an internal view. It then renders a parent div element containing all the necessary htmx attributes. Once the htmx request fires, the view fetches the cache context and template that match the passed fragment_id and uses that context to render the template.

Cache

django-xinclude uses either the cache that corresponds to the XINCLUDE_CACHE_ALIAS setting, if specified, or CACHES["default"]. When setting a new cache key, it finds unpicklable values and discards them. If you want to see which keys get discarded, update your settings.LOGGERS to include "django_xinclude" with "level": "DEBUG".

All official Django cache backends should work, under one important condition:
Your cache should be accessible from all your app instances. If you are using multi-processing for your Django application, or multiple servers clusters, make sure that your django-xinclude cache is accessible from all the instances, otherwise your requests will result in 404s.

Authorization

The request user is expected to be the one that initially accessed the original view (and added to cache), or AnonymousUser in both cases; otherwise django-xinclude will return 404 for the htmx requests. If request.user is not available, for instance when django.contrib.auth is not in the INSTALLED_APPS, then django-xinclude assumes that the end user can access the data.

Rendering synchronously

There are cases where you might want to conditionally render fragments synchronously (i.e. use the regular include). For example, you could render synchronously for SEO purposes, when robots are crawling your pages, but still make use of the htmx functionality for regular users. django-xinclude supports this, it checks for a xinclude_sync attribute on the request and renders synchronously if that evaluates to True. So you can add a custom middleware that sets the xinclude_sync attribute upon your individual conditions.

See also Configuration below for the XINCLUDE_SYNC_REQUEST_ATTR setting.

Configuration

XINCLUDE_CACHE_ALIAS: str

The cache alias that django-xinclude will use, it defaults to CACHES["default"].

XINCLUDE_CACHE_TIMEOUT: int

The number of seconds that contexts will remain in cache. If the setting is not present, Django will use the default timeout argument of the appropriate backend in the CACHES setting.

XINCLUDE_SYNC_REQUEST_ATTR: str

The request attribute that django-xinclude will check on to determine if it needs to render synchronously. It defaults to xinclude_sync.

Running the tests

Fork, then clone the repo:

git clone [email protected]:your-username/django-xinclude.git

Set up a venv:

python -m venv .venv
source .venv/bin/activate
python -m pip install -e '.[tests,dev]'

Set up the pre-commit hooks:

pre-commit install

Then you can run the tests with the just command runner:

just test

Or with coverage:

just coverage

If you don't have just installed, you can look in the justfile for the commands that are run.


Complementary packages