Django

goatseo-django provides request-scoped SEO state, template tags, a context processor, view helpers and an optional middleware.

pip install "goatseo[django]"

Settings

from goatseo import Metadata

INSTALLED_APPS = [
    # ...
    "goatseo.django",
]

TEMPLATES = [
    {
        "BACKEND": "django.template.backends.django.DjangoTemplates",
        "APP_DIRS": True,
        "OPTIONS": {
            "context_processors": [
                "django.template.context_processors.request",
                "goatseo.django.context_processors.seo",
            ],
        },
    },
]

# Optional: mirror robots directives in the X-Robots-Tag header.
MIDDLEWARE = [
    # ...
    "goatseo.django.middleware.GoatSEOMiddleware",
]

GOATSEO_DEFAULTS = Metadata(site_name="Example", title_template="{title} | Example")

GOATSEO_DEFAULTS must be a Metadata instance: manage.py check reports anything else as goatseo.E001, and get_seo raises ImproperlyConfigured. It is read once and re-read when a test overrides it with override_settings.

Views

get_seo(request) returns the SEO object of the current request, creating it from GOATSEO_DEFAULTS on first use.

from django.shortcuts import get_object_or_404, render
from goatseo.django import get_seo
from goatseo.schema import Article as ArticleSchema, Person


def article_detail(request, slug):
    article = get_object_or_404(Article, slug=slug)
    seo = get_seo(request)
    seo.from_object(article)
    seo.schema(ArticleSchema(headline=article.title, author=Person(name=article.author_name)))
    return render(request, "article.html", {"article": article})

Function-based views

seo_defaults sets the view layer; the view body can still override it:

from goatseo.django import seo_defaults


@seo_defaults(title="Blog", description="Every article of the blog.")
def article_list(request): ...

Class-based views

SEOMixin adds seo_metadata (or get_seo_metadata()) as the view layer and, when the view has an object implementing seo_metadata(), uses it as the model layer. It also puts seo in the template context.

from django.views.generic import DetailView
from goatseo import Metadata
from goatseo.django import SEOMixin


class ArticleDetail(SEOMixin, DetailView):
    model = Article
    seo_metadata = Metadata(publisher="Example Media")

from goatseo.django import model_metadata


class Article(models.Model):
    title = models.CharField(max_length=200)
    summary = models.TextField(blank=True)

    def seo_metadata(self) -> Metadata:
        return model_metadata(title=self.title, description=self.summary)

model_metadata builds a Metadata where blank strings, which Django stores for empty fields, mean "not set": an empty summary falls back to a lower layer instead of failing validation.

Templates

{% load goatseo %}
<!doctype html>
<html lang="{{ seo.metadata.language|default:'en' }}">
<head>
  <meta charset="utf-8">
  {% goatseo_head %}
</head>

Tag	Output
{% goatseo_head %}	Every element: metadata, Open Graph, Twitter, JSON-LD
{% goatseo_jsonld %}	Only the JSON-LD script

Both tags need the request context processor. They use the seo context variable when it holds an SEO object, otherwise get_seo(request). Relative URLs are resolved with request.build_absolute_uri().

Middleware

GoatSEOMiddleware is optional. It copies the resolved robots directives to the X-Robots-Tag header, unless the response already has one. It does nothing for requests that never touched get_seo, but the context processor calls it on every template render: with robots in GOATSEO_DEFAULTS, every templated response gets the header, the admin included.