The built-in page module

FeinCMS is primarily a system to work with lists of content blocks which you can assign to arbitrary other objects. You do not necessarily have to use it with a hierarchical page structure, but that’s the most common use case of course. Being able to put content together in small manageable pieces is interesting for other uses too, e.g. for weblog entries where you have rich text content interspersed with images, videos or maybe even galleries.

Activating the page module and creating content types

To activate the page module, you need to follow the instructions in Installation instructions and afterwards add to your INSTALLED_APPS.

Before proceeding with syncdb, it might be a good idea to take a look at Page extension modules – the page module does have the minimum of features in the default configuration and you will probably want to enable several extensions.

You need to create some content models too. No models are created by default, because there is no possibility to unregister models. A sane default might be to create MediaFileContent and RichTextContent models; you can do this by adding the following lines somewhere into your project, for example in a file that will be processed anyway:

from django.utils.translation import ugettext_lazy as _

from import Page
from feincms.content.richtext.models import RichTextContent
from feincms.content.medialibrary.models import MediaFileContent

Page.register_extensions('datepublisher', 'translations') # Example set of extensions

    'title': _('Standard template'),
    'path': 'base.html',
    'regions': (
        ('main', _('Main content area')),
        ('sidebar', _('Sidebar'), 'inherited'),

Page.create_content_type(MediaFileContent, TYPE_CHOICES=(
    ('default', _('default')),
    ('lightbox', _('lightbox')),

It will be a good idea most of the time to register the RichTextContent first, because it’s the most used content type for many applications. The content type dropdown will contain content types in the same order as they were registered.

Please note that you should put these statements into a file of an app contained in INSTALLED_APPS. That file is executed at Django startup time.

Setting up the admin interface

The customized admin interface code is contained inside the ModelAdmin subclass, so you do not need to do anything special here.

If you use the RichTextContent, you need to download TinyMCE and configure FeinCMS’ richtext support:

    'TINYMCE_JS_URL': STATIC_URL + 'your_custom_path/tiny_mce.js',

Wiring up the views

Just add the following lines to your to get a catch-all URL pattern:

urlpatterns += patterns('',
    url(r'', include('feincms.urls')),

If you want to define a page as home page for the whole site, you can give it an override_url value of '/'.

More information can be found in Integrating 3rd party apps into your site

Adding another content type

Imagine you’ve got a third-party gallery application and you’d like to include excerpts of galleries inside your content. You’d need to write a GalleryContent base class and let FeinCMS create a model class for you with some important attributes added.

from django.db import models
from django.template.loader import render_to_string
from import Page
from gallery.models import Gallery

class GalleryContent(models.Model):
    gallery = models.ForeignKey(Gallery)

    class Meta:
        abstract = True # Required by FeinCMS, content types must be abstract

    def render(self, **kwargs):
        return render_to_string('gallery/gallerycontent.html', {
            'content': self, # Not required but a convention followed by
                             # all of FeinCMS' bundled content types


The newly created GalleryContent for Page will live in the database table page_page_gallerycontent.


FeinCMS requires your content type model to be abstract.

More information about content types is available in Content types - what your page content is built of.

Page extension modules

Extensions are a way to put often-used functionality easily accessible without cluttering up the core page model for those who do not need them. The extensions are standard python modules with a register() method which will be called upon registering the extension. The register() method receives the Page class itself and the model admin class PageAdmin as arguments. The extensions can be activated as follows:

Page.register_extensions('navigation', 'titles', 'translations')

The following extensions are available currently:

  • changedate — Creation and modification dates

    Adds automatically maintained creation and modification date fields to the page.

  • ct_tracker — Content type cache

    Helps reduce database queries if you have three or more content types.

  • datepublisher — Date-based publishing

    Adds publication date and end date fields to the page, thereby enabling the administrator to define a date range where a page will be available to website visitors.

  • excerpt — Page summary

    Add a brief excerpt summarizing the content of this page.

  • featured — Simple featured flag for a page

    Lets administrators set a featured flag that lets you treat that page special.

  • navigation — Navigation extensions

    Adds navigation extensions to the page model. You can define subclasses of NavigationExtension, which provide submenus to the navigation generation mechanism. See Letting 3rd party apps define navigation entries for more information on how to use this extension.

  • relatedpages — Links related content

    Add a many-to-many relationship field to relate this page to other pages.

  • seo — Search engine optimization

    Adds fields to the page relevant for search engine optimization (SEO), currently only meta keywords and description.

  • sites — Limit pages to sites

    Allows to limit a page to a certain site and not display it on other sites.

  • symlinks — Symlinked content extension

    Sometimes you want to reuse all content from a page in another place. This extension lets you do that.

  • titles — Additional titles

    Adds additional title fields to the page model. You may not only define a single title for the page to be used in the navigation, the <title> tag and inside the content area, you are not only allowed to define different titles for the three uses but also enabled to define titles and subtitles for the content area.

  • translations — Page translations

    Adds a language field and a recursive translations many to many field to the page, so that you can define the language the page is in and assign translations. I am currently very unhappy with state of things concerning the definition of translations, so that extension might change somewhat too. This extension also adds new instructions to the setup_request method where the Django i18n tools are initialized with the language given on the page object.

    While it is not required by FeinCMS itself it’s still recommended to add django.middleware.locale.LocaleMiddleware to the MIDDLEWARE_CLASSES; otherwise you will see strange language switching behavior in non-FeinCMS managed views (such as third party apps not integrated using feincms.content.application.models.ApplicationContent or Django’s own administration tool). You need to have defined settings.LANGUAGES as well.


These extension modules add new fields to the Page class. If you add or remove page extensions after you’ve run syncdb for the first time you have to change the database schema yourself, or use Database migration support for FeinCMS with South.

Using page request processors

A request processor is a function that gets the currently selected page and the request as parameters and returns either None (or nothing) or a HttpResponse. All registered request processors are run before the page is actually rendered. If the request processor indeed returns a HttpResponse, further rendering of the page is cut short and this response is returned immediately to the client.

This allows for various actions dependent on page and request, for example a simple user access check can be implemented like this:

def authenticated_request_processor(page, request):
    if not request.user.is_authenticated():
        return HttpResponseForbidden()


register_request_processor has an optional second argument named key. If you register a request processor with the same key, the second processor replaces the first. This is especially handy to replace the standard request processors named path_active (which checks whether all ancestors of a given page are active too) and redirect (which issues HTTP-level redirects if the redirect_to page field is filled in).

Using page response processors

Analogous to a request processor, a response processor runs after a page has been rendered. It needs to accept the page, the request and the response as parameters and may change the response (or throw an exception, but try not to).

A response processor is the right place to tweak the returned http response for whatever purposes you have in mind.

def set_random_header_response_processor(page, request, response):
    response['X-Random-Number'] = 42


register_response_processor has an optional second argument named key, exactly like register_request_processor above. It behaves in the same way.


TinyMCE is configured by default to only allow for minimal formatting. This has proven to be the best compromise between letting the client format text without destroying the page design concept. You can customize the TinyMCE settings by creating your own init_richtext.html that inherits from admin/content/richtext/init_tinymce.html. You can even set your own CSS and linklist files like so:

        'TINYMCE_JS_URL': STATIC_URL + 'your_custom_path/tiny_mce.js',
        'TINYMCE_CONTENT_CSS_URL': None,  # add your css path here
        'TINYMCE_LINK_LIST_URL': None  # add your linklist.js path here

FeinCMS is set up to use TinyMCE but you can use CKEditor instead if you prefer that one. Change the following settings:

FEINCMS_RICHTEXT_INIT_TEMPLATE = 'admin/content/richtext/init_ckeditor.html'
        'CKEDITOR_JS_URL': STATIC_URL + 'path_to_your/ckeditor.js',

ETag handling

An ETag is a string that is associated with a page – it should change if (and only if) the page content itself has changed. Since a page’s content may depend on more than just the raw page data in the database (e.g. it might list its children or a navigation tree or an excerpt from some other place in the CMS alltogether), you are required to write an etag producing method for the page.

# Very stupid etag function, a page is supposed the unchanged as long
# as its id and slug do not change. You definitely want something more
# involved, like including last change dates or whatever.
def my_etag(page, request):
    return 'PAGE-%d-%s' % (, page.slug )
Page.etag = my_etag



To create a sitemap that is automatically populated with all pages in your Feincms site, add the following to your top-level

from import PageSitemap
sitemaps = {'pages' : PageSitemap}

urlpatterns += patterns('',
    url(r'^sitemap\.xml$', 'django.contrib.sitemaps.views.sitemap',
        {'sitemaps': sitemaps}),

This will produce a default sitemap at the /sitemap.xml url. A sitemap can be further customised by passing it appropriate parameters, like so:

sitemaps = {'pages': PageSitemap(max_depth=2)}

The following parameters can be used to modify the behaviour of the sitemap:

  • navigation_only – if set to True, only pages that are in_navigation will appear in the site map.
  • max_depth – if set to a non-negative integer, will limit the sitemap generated to this page hierarchy depth.
  • changefreq – should be a string or callable specifying the page update frequency, according to the sitemap protocol.
  • queryset – pass in a query set to restrict the Pages to include in the site map.
  • filter – pass in a callable that transforms a queryset to filter out the pages you want to include in the site map.
  • extended_navigation – if set to True, adds pages from any navigation extensions. If using PagePretender, make sure to include title, url, level, in_navigation and optionally modification_date.