1 ==============================
2 The syndication feed framework
3 ==============================
5 Django comes with a high-level syndication-feed-generating framework that makes
6 creating RSS_ and Atom_ feeds easy.
8 To create any syndication feed, all you have to do is write a short Python
9 class. You can create as many feeds as you want.
11 Django also comes with a lower-level feed-generating API. Use this if you want
12 to generate feeds outside of a Web context, or in some other lower-level way.
14 .. _RSS: http://www.whatisrss.com/
15 .. _Atom: http://www.atomenabled.org/
17 The high-level framework
18 ========================
23 The high-level feed-generating framework is a view that's hooked to ``/feeds/``
24 by default. Django uses the remainder of the URL (everything after ``/feeds/``)
25 to determine which feed to output.
27 To create a feed, just write a ``Feed`` class and point to it in your URLconf_.
29 .. _URLconf: ../url_dispatch/
34 If you're not using the latest Django development version, you'll need to make
35 sure Django's sites framework is installed -- including its database table.
36 (See the `sites framework documentation`_ for more information.) This has
37 changed in the Django development version; the syndication feed framework no
38 longer requires the sites framework.
40 To activate syndication feeds on your Django site, add this line to your
43 (r'^feeds/(?P<url>.*)/$', 'django.contrib.syndication.views.feed', {'feed_dict': feeds}),
45 This tells Django to use the RSS framework to handle all URLs starting with
46 ``"feeds/"``. (You can change that ``"feeds/"`` prefix to fit your own needs.)
48 This URLconf line has an extra argument: ``{'feed_dict': feeds}``. Use this
49 extra argument to pass the syndication framework the feeds that should be
50 published under that URL.
52 Specifically, ``feed_dict`` should be a dictionary that maps a feed's slug
53 (short URL label) to its ``Feed`` class.
55 You can define the ``feed_dict`` in the URLconf itself. Here's a full example
58 from django.conf.urls.defaults import *
59 from myproject.feeds import LatestEntries, LatestEntriesByCategory
62 'latest': LatestEntries,
63 'categories': LatestEntriesByCategory,
66 urlpatterns = patterns('',
68 (r'^feeds/(?P<url>.*)/$', 'django.contrib.syndication.views.feed',
69 {'feed_dict': feeds}),
73 The above example registers two feeds:
75 * The feed represented by ``LatestEntries`` will live at ``feeds/latest/``.
76 * The feed represented by ``LatestEntriesByCategory`` will live at
77 ``feeds/categories/``.
79 Once that's set up, you just need to define the ``Feed`` classes themselves.
81 .. _sites framework documentation: ../sites/
82 .. _URLconf: ../url_dispatch/
83 .. _settings file: ../settings/
88 A ``Feed`` class is a simple Python class that represents a syndication feed.
89 A feed can be simple (e.g., a "site news" feed, or a basic feed displaying
90 the latest entries of a blog) or more complex (e.g., a feed displaying all the
91 blog entries in a particular category, where the category is variable).
93 ``Feed`` classes must subclass ``django.contrib.syndication.feeds.Feed``. They
94 can live anywhere in your codebase.
99 This simple example, taken from `chicagocrime.org`_, describes a feed of the
100 latest five news items::
102 from django.contrib.syndication.feeds import Feed
103 from chicagocrime.models import NewsItem
105 class LatestEntries(Feed):
106 title = "Chicagocrime.org site news"
108 description = "Updates on changes and additions to chicagocrime.org."
111 return NewsItem.objects.order_by('-pub_date')[:5]
115 * The class subclasses ``django.contrib.syndication.feeds.Feed``.
116 * ``title``, ``link`` and ``description`` correspond to the standard
117 RSS ``<title>``, ``<link>`` and ``<description>`` elements, respectively.
118 * ``items()`` is, simply, a method that returns a list of objects that
119 should be included in the feed as ``<item>`` elements. Although this
120 example returns ``NewsItem`` objects using Django's
121 `object-relational mapper`_, ``items()`` doesn't have to return model
122 instances. Although you get a few bits of functionality "for free" by
123 using Django models, ``items()`` can return any type of object you want.
124 * If you're creating an Atom feed, rather than an RSS feed, set the
125 ``subtitle`` attribute instead of the ``description`` attribute. See
126 `Publishing Atom and RSS feeds in tandem`_, later, for an example.
128 One thing's left to do. In an RSS feed, each ``<item>`` has a ``<title>``,
129 ``<link>`` and ``<description>``. We need to tell the framework what data to
130 put into those elements.
132 * To specify the contents of ``<title>`` and ``<description>``, create
133 `Django templates`_ called ``feeds/latest_title.html`` and
134 ``feeds/latest_description.html``, where ``latest`` is the ``slug``
135 specified in the URLconf for the given feed. Note the ``.html`` extension
136 is required. The RSS system renders that template for each item, passing
137 it two template context variables:
139 * ``{{ obj }}`` -- The current object (one of whichever objects you
140 returned in ``items()``).
141 * ``{{ site }}`` -- A ``django.contrib.sites.models.Site`` object
142 representing the current site. This is useful for
143 ``{{ site.domain }}`` or ``{{ site.name }}``. Note that if you're
144 using the latest Django development version and do *not* have the
145 Django sites framework installed, this will be set to a
146 ``django.contrib.sites.models.RequestSite`` object. See the
147 `RequestSite section of the sites framework documentation`_ for
150 If you don't create a template for either the title or description, the
151 framework will use the template ``"{{ obj }}"`` by default -- that is,
152 the normal string representation of the object. You can also change the
153 names of these two templates by specifying ``title_template`` and
154 ``description_template`` as attributes of your ``Feed`` class.
155 * To specify the contents of ``<link>``, you have two options. For each
156 item in ``items()``, Django first tries executing a
157 ``get_absolute_url()`` method on that object. If that method doesn't
158 exist, it tries calling a method ``item_link()`` in the ``Feed`` class,
159 passing it a single parameter, ``item``, which is the object itself.
160 Both ``get_absolute_url()`` and ``item_link()`` should return the item's
161 URL as a normal Python string. As with ``get_absolute_url()``, the
162 result of ``item_link()`` will be included directly in the URL, so you
163 are responsible for doing all necessary URL quoting and conversion to
164 ASCII inside the method itself.
166 * For the LatestEntries example above, we could have very simple feed templates:
168 * latest_title.html::
172 * latest_description.html::
174 {{ obj.description }}
176 .. _chicagocrime.org: http://www.chicagocrime.org/
177 .. _object-relational mapper: ../db-api/
178 .. _Django templates: ../templates/
179 .. _RequestSite section of the sites framework documentation: ../sites/#requestsite-objects
184 The framework also supports more complex feeds, via parameters.
186 For example, `chicagocrime.org`_ offers an RSS feed of recent crimes for every
187 police beat in Chicago. It'd be silly to create a separate ``Feed`` class for
188 each police beat; that would violate the `DRY principle`_ and would couple data
189 to programming logic. Instead, the syndication framework lets you make generic
190 feeds that output items based on information in the feed's URL.
192 On chicagocrime.org, the police-beat feeds are accessible via URLs like this:
194 * ``/rss/beats/0613/`` -- Returns recent crimes for beat 0613.
195 * ``/rss/beats/1424/`` -- Returns recent crimes for beat 1424.
197 The slug here is ``"beats"``. The syndication framework sees the extra URL bits
198 after the slug -- ``0613`` and ``1424`` -- and gives you a hook to tell it what
199 those URL bits mean, and how they should influence which items get published in
202 An example makes this clear. Here's the code for these beat-specific feeds::
204 class BeatFeed(Feed):
205 def get_object(self, bits):
206 # In case of "/rss/beats/0613/foo/bar/baz/", or other such clutter,
207 # check that bits has only one member.
209 raise ObjectDoesNotExist
210 return Beat.objects.get(beat__exact=bits[0])
212 def title(self, obj):
213 return "Chicagocrime.org: Crimes for beat %s" % obj.beat
216 return obj.get_absolute_url()
218 def description(self, obj):
219 return "Crimes recently reported in police beat %s" % obj.beat
221 def items(self, obj):
222 return Crime.objects.filter(beat__id__exact=obj.id).order_by('-crime_date')[:30]
224 Here's the basic algorithm the RSS framework follows, given this class and a
225 request to the URL ``/rss/beats/0613/``:
227 * The framework gets the URL ``/rss/beats/0613/`` and notices there's
228 an extra bit of URL after the slug. It splits that remaining string by
229 the slash character (``"/"``) and calls the ``Feed`` class'
230 ``get_object()`` method, passing it the bits. In this case, bits is
231 ``['0613']``. For a request to ``/rss/beats/0613/foo/bar/``, bits would
232 be ``['0613', 'foo', 'bar']``.
233 * ``get_object()`` is responsible for retrieving the given beat, from the
234 given ``bits``. In this case, it uses the Django database API to retrieve
235 the beat. Note that ``get_object()`` should raise
236 ``django.core.exceptions.ObjectDoesNotExist`` if given invalid
237 parameters. There's no ``try``/``except`` around the
238 ``Beat.objects.get()`` call, because it's not necessary; that function
239 raises ``Beat.DoesNotExist`` on failure, and ``Beat.DoesNotExist`` is a
240 subclass of ``ObjectDoesNotExist``. Raising ``ObjectDoesNotExist`` in
241 ``get_object()`` tells Django to produce a 404 error for that request.
242 * To generate the feed's ``<title>``, ``<link>`` and ``<description>``,
243 Django uses the ``title()``, ``link()`` and ``description()`` methods. In
244 the previous example, they were simple string class attributes, but this
245 example illustrates that they can be either strings *or* methods. For
246 each of ``title``, ``link`` and ``description``, Django follows this
249 * First, it tries to call a method, passing the ``obj`` argument, where
250 ``obj`` is the object returned by ``get_object()``.
251 * Failing that, it tries to call a method with no arguments.
252 * Failing that, it uses the class attribute.
254 * Finally, note that ``items()`` in this example also takes the ``obj``
255 argument. The algorithm for ``items`` is the same as described in the
256 previous step -- first, it tries ``items(obj)``, then ``items()``, then
257 finally an ``items`` class attribute (which should be a list).
259 The ``ExampleFeed`` class below gives full documentation on methods and
260 attributes of ``Feed`` classes.
262 .. _DRY principle: http://c2.com/cgi/wiki?DontRepeatYourself
264 Specifying the type of feed
265 ---------------------------
267 By default, feeds produced in this framework use RSS 2.0.
269 To change that, add a ``feed_type`` attribute to your ``Feed`` class, like so::
271 from django.utils.feedgenerator import Atom1Feed
274 feed_type = Atom1Feed
276 Note that you set ``feed_type`` to a class object, not an instance.
278 Currently available feed types are:
280 * ``django.utils.feedgenerator.Rss201rev2Feed`` (RSS 2.01. Default.)
281 * ``django.utils.feedgenerator.RssUserland091Feed`` (RSS 0.91.)
282 * ``django.utils.feedgenerator.Atom1Feed`` (Atom 1.0.)
287 To specify enclosures, such as those used in creating podcast feeds, use the
288 ``item_enclosure_url``, ``item_enclosure_length`` and
289 ``item_enclosure_mime_type`` hooks. See the ``ExampleFeed`` class below for
295 Feeds created by the syndication framework automatically include the
296 appropriate ``<language>`` tag (RSS 2.0) or ``xml:lang`` attribute (Atom). This
297 comes directly from your `LANGUAGE_CODE setting`_.
299 .. _LANGUAGE_CODE setting: ../settings/#language-code
304 The ``link`` method/attribute can return either an absolute URL (e.g.
305 ``"/blog/"``) or a URL with the fully-qualified domain and protocol (e.g.
306 ``"http://www.example.com/blog/"``). If ``link`` doesn't return the domain,
307 the syndication framework will insert the domain of the current site, according
308 to your `SITE_ID setting`_.
310 Atom feeds require a ``<link rel="self">`` that defines the feed's current
311 location. The syndication framework populates this automatically, using the
312 domain of the current site according to the SITE_ID setting.
314 .. _SITE_ID setting: ../settings/#site-id
316 Publishing Atom and RSS feeds in tandem
317 ---------------------------------------
319 Some developers like to make available both Atom *and* RSS versions of their
320 feeds. That's easy to do with Django: Just create a subclass of your ``Feed``
321 class and set the ``feed_type`` to something different. Then update your
322 URLconf to add the extra versions.
324 Here's a full example::
326 from django.contrib.syndication.feeds import Feed
327 from chicagocrime.models import NewsItem
328 from django.utils.feedgenerator import Atom1Feed
330 class RssSiteNewsFeed(Feed):
331 title = "Chicagocrime.org site news"
333 description = "Updates on changes and additions to chicagocrime.org."
336 return NewsItem.objects.order_by('-pub_date')[:5]
338 class AtomSiteNewsFeed(RssSiteNewsFeed):
339 feed_type = Atom1Feed
340 subtitle = RssSiteNewsFeed.description
343 In this example, the RSS feed uses a ``description`` while the Atom feed
344 uses a ``subtitle``. That's because Atom feeds don't provide for a
345 feed-level "description," but they *do* provide for a "subtitle."
347 If you provide a ``description`` in your ``Feed`` class, Django will *not*
348 automatically put that into the ``subtitle`` element, because a subtitle
349 and description are not necessarily the same thing. Instead, you should
350 define a ``subtitle`` attribute.
352 In the above example, we simply set the Atom feed's ``subtitle`` to the
353 RSS feed's ``description``, because it's quite short already.
355 And the accompanying URLconf::
357 from django.conf.urls.defaults import *
358 from myproject.feeds import RssSiteNewsFeed, AtomSiteNewsFeed
361 'rss': RssSiteNewsFeed,
362 'atom': AtomSiteNewsFeed,
365 urlpatterns = patterns('',
367 (r'^feeds/(?P<url>.*)/$', 'django.contrib.syndication.views.feed',
368 {'feed_dict': feeds}),
375 This example illustrates all possible attributes and methods for a ``Feed`` class::
377 from django.contrib.syndication.feeds import Feed
378 from django.utils import feedgenerator
380 class ExampleFeed(Feed):
382 # FEED TYPE -- Optional. This should be a class that subclasses
383 # django.utils.feedgenerator.SyndicationFeed. This designates which
384 # type of feed this should be: RSS 2.0, Atom 1.0, etc.
385 # If you don't specify feed_type, your feed will be RSS 2.0.
386 # This should be a class, not an instance of the class.
388 feed_type = feedgenerator.Rss201rev2Feed
390 # TEMPLATE NAMES -- Optional. These should be strings representing
391 # names of Django templates that the system should use in rendering the
392 # title and description of your feed items. Both are optional.
393 # If you don't specify one, or either, Django will use the template
394 # 'feeds/SLUG_title.html' and 'feeds/SLUG_description.html', where SLUG
395 # is the slug you specify in the URL.
397 title_template = None
398 description_template = None
400 # TITLE -- One of the following three is required. The framework looks
401 # for them in this order.
403 def title(self, obj):
405 Takes the object returned by get_object() and returns the feed's
406 title as a normal Python string.
411 Returns the feed's title as a normal Python string.
414 title = 'foo' # Hard-coded title.
416 # LINK -- One of the following three is required. The framework looks
417 # for them in this order.
421 Takes the object returned by get_object() and returns the feed's
422 link as a normal Python string.
427 Returns the feed's link as a normal Python string.
430 link = '/foo/bar/' # Hard-coded link.
432 # GUID -- One of the following three is optional. The framework looks
433 # for them in this order. This property is only used for Atom feeds
434 # (where it is the feed-level ID element). If not provided, the feed
435 # link is used as the ID.
437 # (New in Django development version)
439 def feed_guid(self, obj):
441 Takes the object returned by get_object() and returns the globally
442 unique ID for the feed as a normal Python string.
447 Returns the feed's globally unique ID as a normal Python string.
450 feed_guid = '/foo/bar/1234' # Hard-coded guid.
452 # DESCRIPTION -- One of the following three is required. The framework
453 # looks for them in this order.
455 def description(self, obj):
457 Takes the object returned by get_object() and returns the feed's
458 description as a normal Python string.
461 def description(self):
463 Returns the feed's description as a normal Python string.
466 description = 'Foo bar baz.' # Hard-coded description.
468 # AUTHOR NAME --One of the following three is optional. The framework
469 # looks for them in this order.
471 def author_name(self, obj):
473 Takes the object returned by get_object() and returns the feed's
474 author's name as a normal Python string.
477 def author_name(self):
479 Returns the feed's author's name as a normal Python string.
482 author_name = 'Sally Smith' # Hard-coded author name.
484 # AUTHOR E-MAIL --One of the following three is optional. The framework
485 # looks for them in this order.
487 def author_email(self, obj):
489 Takes the object returned by get_object() and returns the feed's
490 author's e-mail as a normal Python string.
493 def author_email(self):
495 Returns the feed's author's e-mail as a normal Python string.
498 author_email = 'test@example.com' # Hard-coded author e-mail.
500 # AUTHOR LINK --One of the following three is optional. The framework
501 # looks for them in this order. In each case, the URL should include
502 # the "http://" and domain name.
504 def author_link(self, obj):
506 Takes the object returned by get_object() and returns the feed's
507 author's URL as a normal Python string.
510 def author_link(self):
512 Returns the feed's author's URL as a normal Python string.
515 author_link = 'http://www.example.com/' # Hard-coded author URL.
517 # CATEGORIES -- One of the following three is optional. The framework
518 # looks for them in this order. In each case, the method/attribute
519 # should return an iterable object that returns strings.
521 def categories(self, obj):
523 Takes the object returned by get_object() and returns the feed's
524 categories as iterable over strings.
527 def categories(self):
529 Returns the feed's categories as iterable over strings.
532 categories = ("python", "django") # Hard-coded list of categories.
534 # COPYRIGHT NOTICE -- One of the following three is optional. The
535 # framework looks for them in this order.
537 def copyright(self, obj):
539 Takes the object returned by get_object() and returns the feed's
540 copyright notice as a normal Python string.
545 Returns the feed's copyright notice as a normal Python string.
548 copyright = 'Copyright (c) 2007, Sally Smith' # Hard-coded copyright notice.
550 # ITEMS -- One of the following three is required. The framework looks
551 # for them in this order.
553 def items(self, obj):
555 Takes the object returned by get_object() and returns a list of
556 items to publish in this feed.
561 Returns a list of items to publish in this feed.
564 items = ('Item 1', 'Item 2') # Hard-coded items.
566 # GET_OBJECT -- This is required for feeds that publish different data
567 # for different URL parameters. (See "A complex example" above.)
569 def get_object(self, bits):
571 Takes a list of strings gleaned from the URL and returns an object
572 represented by this feed. Raises
573 django.core.exceptions.ObjectDoesNotExist on error.
576 # ITEM LINK -- One of these three is required. The framework looks for
577 # them in this order.
579 # First, the framework tries the get_absolute_url() method on each item
580 # returned by items(). Failing that, it tries these two methods:
582 def item_link(self, item):
584 Takes an item, as returned by items(), and returns the item's URL.
589 Returns the URL for every item in the feed.
592 # ITEM_GUID -- The following method is optional. This property is
593 # only used for Atom feeds (it is the ID element for an item in an
594 # Atom feed). If not provided, the item's link is used by default.
596 # (New in Django development version)
598 def item_guid(self, obj):
600 Takes an item, as return by items(), and returns the item's ID.
603 # ITEM AUTHOR NAME -- One of the following three is optional. The
604 # framework looks for them in this order.
606 def item_author_name(self, item):
608 Takes an item, as returned by items(), and returns the item's
609 author's name as a normal Python string.
612 def item_author_name(self):
614 Returns the author name for every item in the feed.
617 item_author_name = 'Sally Smith' # Hard-coded author name.
619 # ITEM AUTHOR E-MAIL --One of the following three is optional. The
620 # framework looks for them in this order.
622 # If you specify this, you must specify item_author_name.
624 def item_author_email(self, obj):
626 Takes an item, as returned by items(), and returns the item's
627 author's e-mail as a normal Python string.
630 def item_author_email(self):
632 Returns the author e-mail for every item in the feed.
635 item_author_email = 'test@example.com' # Hard-coded author e-mail.
637 # ITEM AUTHOR LINK --One of the following three is optional. The
638 # framework looks for them in this order. In each case, the URL should
639 # include the "http://" and domain name.
641 # If you specify this, you must specify item_author_name.
643 def item_author_link(self, obj):
645 Takes an item, as returned by items(), and returns the item's
646 author's URL as a normal Python string.
649 def item_author_link(self):
651 Returns the author URL for every item in the feed.
654 item_author_link = 'http://www.example.com/' # Hard-coded author URL.
656 # ITEM ENCLOSURE URL -- One of these three is required if you're
657 # publishing enclosures. The framework looks for them in this order.
659 def item_enclosure_url(self, item):
661 Takes an item, as returned by items(), and returns the item's
665 def item_enclosure_url(self):
667 Returns the enclosure URL for every item in the feed.
670 item_enclosure_url = "/foo/bar.mp3" # Hard-coded enclosure link.
672 # ITEM ENCLOSURE LENGTH -- One of these three is required if you're
673 # publishing enclosures. The framework looks for them in this order.
674 # In each case, the returned value should be either an integer, or a
675 # string representation of the integer, in bytes.
677 def item_enclosure_length(self, item):
679 Takes an item, as returned by items(), and returns the item's
683 def item_enclosure_length(self):
685 Returns the enclosure length for every item in the feed.
688 item_enclosure_length = 32000 # Hard-coded enclosure length.
690 # ITEM ENCLOSURE MIME TYPE -- One of these three is required if you're
691 # publishing enclosures. The framework looks for them in this order.
693 def item_enclosure_mime_type(self, item):
695 Takes an item, as returned by items(), and returns the item's
699 def item_enclosure_mime_type(self):
701 Returns the enclosure MIME type for every item in the feed.
704 item_enclosure_mime_type = "audio/mpeg" # Hard-coded enclosure MIME type.
706 # ITEM PUBDATE -- It's optional to use one of these three. This is a
707 # hook that specifies how to get the pubdate for a given item.
708 # In each case, the method/attribute should return a Python
709 # datetime.datetime object.
711 def item_pubdate(self, item):
713 Takes an item, as returned by items(), and returns the item's
717 def item_pubdate(self):
719 Returns the pubdate for every item in the feed.
722 item_pubdate = datetime.datetime(2005, 5, 3) # Hard-coded pubdate.
724 # ITEM CATEGORIES -- It's optional to use one of these three. This is
725 # a hook that specifies how to get the list of categories for a given
726 # item. In each case, the method/attribute should return an iterable
727 # object that returns strings.
729 def item_categories(self, item):
731 Takes an item, as returned by items(), and returns the item's
735 def item_categories(self):
737 Returns the categories for every item in the feed.
740 item_categories = ("python", "django") # Hard-coded categories.
742 # ITEM COPYRIGHT NOTICE (only applicable to Atom feeds) -- One of the
743 # following three is optional. The framework looks for them in this
746 def item_copyright(self, obj):
748 Takes an item, as returned by items(), and returns the item's
749 copyright notice as a normal Python string.
752 def item_copyright(self):
754 Returns the copyright notice for every item in the feed.
757 item_copyright = 'Copyright (c) 2007, Sally Smith' # Hard-coded copyright notice.
760 The low-level framework
761 =======================
763 Behind the scenes, the high-level RSS framework uses a lower-level framework
764 for generating feeds' XML. This framework lives in a single module:
765 `django/utils/feedgenerator.py`_.
767 Feel free to use this framework on your own, for lower-level tasks.
769 The ``feedgenerator`` module contains a base class ``SyndicationFeed`` and
772 * ``RssUserland091Feed``
776 Each of these three classes knows how to render a certain type of feed as XML.
777 They share this interface:
779 ``__init__(title, link, description, language=None, author_email=None,``
780 ``author_name=None, author_link=None, subtitle=None, categories=None,``
783 Initializes the feed with the given metadata, which applies to the entire feed
784 (i.e., not just to a specific item in the feed).
786 All parameters, if given, should be Unicode objects, except ``categories``,
787 which should be a sequence of Unicode objects.
789 ``add_item(title, link, description, author_email=None, author_name=None,``
790 ``pubdate=None, comments=None, unique_id=None, enclosure=None, categories=())``
792 Add an item to the feed with the given parameters. All parameters, if given,
793 should be Unicode objects, except:
795 * ``pubdate`` should be a `Python datetime object`_.
796 * ``enclosure`` should be an instance of ``feedgenerator.Enclosure``.
797 * ``categories`` should be a sequence of Unicode objects.
799 ``write(outfile, encoding)``
801 Outputs the feed in the given encoding to outfile, which is a file-like object.
803 ``writeString(encoding)``
805 Returns the feed as a string in the given encoding.
810 This example creates an Atom 1.0 feed and prints it to standard output::
812 >>> from django.utils import feedgenerator
813 >>> f = feedgenerator.Atom1Feed(
814 ... title=u"My Weblog",
815 ... link=u"http://www.example.com/",
816 ... description=u"In which I write about what I ate today.",
818 >>> f.add_item(title=u"Hot dog today",
819 ... link=u"http://www.example.com/entries/1/",
820 ... description=u"<p>Today I had a Vienna Beef hot dog. It was pink, plump and perfect.</p>")
821 >>> print f.writeString('utf8')
822 <?xml version="1.0" encoding="utf8"?>
823 <feed xmlns="http://www.w3.org/2005/Atom" xml:lang="en"><title>My Weblog</title>
824 <link href="http://www.example.com/"></link><id>http://www.example.com/</id>
825 <updated>Sat, 12 Nov 2005 00:28:43 -0000</updated><entry><title>Hot dog today</title>
826 <link>http://www.example.com/entries/1/</link><id>tag:www.example.com/entries/1/</id>
827 <summary type="html"><p>Today I had a Vienna Beef hot dog. It was pink, plump and perfect.</p></summary>
830 .. _django/utils/feedgenerator.py: http://code.djangoproject.com/browser/django/trunk/django/utils/feedgenerator.py
831 .. _Python datetime object: http://www.python.org/doc/current/lib/module-datetime.html