36👍
It works rather intuitively. include
a urlconf that has yet another namespaced include
will result in nested namespaces.
## urls.py
nested2 = patterns('',
url(r'^index/$', 'index', name='index'),
)
nested1 = patterns('',
url(r'^nested2/', include(nested2, namespace="nested2"),
url(r'^index/$', 'index', name='index'),
)
urlpatterns = patterns('',
(r'^nested1/', include(nested1, namespace="nested1"),
)
reverse('nested1:nested2:index') # should output /nested1/nested2/index/
reverse('nested1:index') # should output /nested1/index/
It’s a great way to keep urls organized. I suppose the best advice I can give is to remember that include
can take a patterns
object directly (as in my example) which lets you use a single urls.py
and split views into useful namespaces without having to create multiple urls files.
11👍
UPDATE 2 (2019-10-09)
As eugene comments, UPDATE 1 no longer works for more recent versions of Django, which require an app_name
to be defined in a urls.py
when it’s being included.
On GitHub I’ve created a Django project (myproject
) with a couple of apps (products
and books
) to demonstrate how this is done to create nested namespaces. In summary, the various urls.py
look like this:
# myproject/urls.py
from django.urls import include, path
from products import urls as products_urls
from products import views
urlpatterns = [
path("", views.site_home, name="home"),
path("products/", include(products_urls, namespace="products"),)
]
# products/urls.py
from django.urls import include, path
from books import urls as books_urls
from . import views
app_name = "products"
urlpatterns = [
path("", views.index, name="product_index"),
path("books/", include(books_urls, namespace="books")),
]
# books/urls.py
from django.urls import path
from . import views
app_name = "books"
urlpatterns = [
path("", views.index, name="book_index"),
path("<slug:book_slug>/", views.detail, name="book_detail"),
]
So you can use these nested URL names like this:
reverse("products:books:book_index")
# '/products/books/'
reverse("products:books:book_detail", kwargs={"book_slug": "my-book"})
# '/products/books/my-book/'
UPDATE 1
Django 2.0 introduced two relevant changes. First, the urls()
function is now in django.urls
, so the first line of the urls.py
example above would be:
from django.urls import include, url
Second, it introduce the path()
function as a simpler alternative for paths that don’t require a regular expression. Using that, the example urls.py
would be like this:
from django.urls import include, path
nested2 = [
path('index/', 'index', name='index'),
]
nested1 = [
path('nested2/', include(nested2, namespace='nested2')),
path('index/', 'index', name='index'),
]
urlpatterns = [
path('nested1/', include(nested1, namespace='nested1')),
]
ORIGINAL ANSWER
While Yuji’s answer is correct, note that django.conf.urls.patterns
no longer exists (since Django 1.10) and plain lists are used instead.
The same example urls.py
should now be like this:
from django.conf.urls import include, url
nested2 = [
url(r'^index/$', 'index', name='index'),
]
nested1 = [
url(r'^nested2/', include(nested2, namespace='nested2')),
url(r'^index/$', 'index', name='index'),
]
urlpatterns = [
url(r'^nested1/', include(nested1, namespace='nested1')),
]
And still used like:
reverse('nested1:nested2:index') # should output /nested1/nested2/index/
reverse('nested1:index') # should output /nested1/index/
- [Django]-LEFT JOIN Django ORM
- [Django]-Django – Cascade deletion in ManyToManyRelation
- [Django]-Can a generic.GenericForeignKey() field be Null?