将未转义的 html 插入 django 制作的 rss feed 中

Mar*_*ark 5 xml django rss escaping feed

我正在尝试使用 django 使用feedgenerator.Rss201rev2Feed创建播客 rss 提要 作为提要生成器,它的工作方式与 BeautifulSoup 有点相反:将信息放入适当的 xml 标签中

它运行良好,但我不想转义所有html

特别是,我希望<itunes:summary>rss feed 的值显示如下: <itunes:summary><![CDATA[Link to <a href="http://www.website.com">the website</a>]]></itunes:summary> 按照Apple 规范

如果我在普通视图中渲染 html,我可以在 html 模板中使用|safe过滤器。我现在需要类似的东西,以有选择地防止<在 rss feed 中转义。

也就是说,我需要 rss 出现<![CDATA[...]]而不是转义&lt;![CDATA[...]]&gt;

然而,Django似乎“无论如何,Django都会自动转义RSS提要(或任何与此相关的XML)中的特殊字符,无论您是否通过安全过滤器传递它”(请参阅​​this 2009 Question

到目前为止还没有运气:

因此,迄今为止使用mark_safe 的尝试已被证明是无用的。

我也不确定如何解释一种想法,将“autoescape=False 传递给 django.contrib.synmination.feeds 中的 render() 调用”。

添加到 addQuickElement 注释的建议, escape=False返回了错误

 handler.addQuickElement(u'itunes:summary',item['summary'], escape=False)
 TypeError: addQuickElement() got an unexpected keyword argument 'escape'
Run Code Online (Sandbox Code Playgroud)

这是一个老问题,但到目前为止我找不到任何解决方案。

有人知道一种巧妙的方法来<![CDATA[...出现在最终提要中,而不是转义到 吗&lt;![CDATA[...


编辑:这是我发布这个问题时当前形式的代码(我还没有尝试合并@Lego的答案)

import mimetypes

from django.conf import settings
from django.contrib.syndication.views import Feed

# For customising the feed
from django.utils.feedgenerator import Rss201rev2Feed
from django.utils import feedgenerator
# see also https://github.com/blancltd/blanc-basic-podcast/blob/master/blanc_basic_podcast/podcast/itunesfeed.py
# and https://github.com/aneumeier/feeds/blob/master/feeds/rss.py
# and https://github.com/blancltd/blanc-basic-podcast/blob/master/blanc_basic_podcast/podcast/feeds.py
# and https://docs.djangoproject.com/en/1.7/ref/contrib/syndication/#custom-feed-generators

from django.contrib.auth.models import User
from django.shortcuts import get_object_or_404
from django.utils.translation import ugettext_lazy as _
from django.contrib.sites.models import Site

from audiotracks.models import get_track_model, Playlist
Track = get_track_model()

ITEMS_PER_FEED = getattr(settings, 'AUDIOTRACKS_PODCAST_LIMIT', 99)
# MarkAdded @ToDo revisit that default maximum num. tracks per feed

from django.core.urlresolvers import reverse, reverse_lazy


from django_slack import slack_message




######################################################################
##### try adapting code from https://github.com/CaptainHayashi/django-lass-uryplayer/blob/master/uryplayer/feeds.py

from django.utils.feedgenerator import Rss201rev2Feed
from django.contrib.syndication.views import Feed
from django.contrib.sites.models import Site
from django.db.models import permalink
# from uryplayer.models import Podcast
import datetime
# MarkAdded in attempt to have un-escaped <![CDATA[...]]
from django.utils.safestring import mark_safe


# from /sf/ask/19262211/
try:
    from html.parser import HTMLParser  # py3
except ImportError:
    from HTMLParser import HTMLParser  # py2

unescape = HTMLParser().unescape
# print(unescape("&gt;"))
# That proved useless so far



class iTunesPodcastsFeedGenerator(Rss201rev2Feed):

    def rss_attributes(self):
        return {u"version": self._version, u"xmlns:atom": u"http://www.w3.org/2005/Atom", u'xmlns:itunes': u'http://www.itunes.com/dtds/podcast-1.0.dtd'}

    def add_root_elements(self, handler):
        super(iTunesPodcastsFeedGenerator, self).add_root_elements(handler)
        handler.addQuickElement(u'itunes:subtitle', self.feed['subtitle'])
        handler.addQuickElement(u'itunes:author', self.feed['author_name'])
        # handler.addQuickElement(u'itunes:summary', mark_safe(self.feed['description']))
        handler.addQuickElement(u'itunes:summary', unescape(mark_safe(self.feed['description'])))

        # handler.addQuickElement(u'itunes:image', self.feed['iTunes_image_url'])
        handler.addQuickElement('itunes:image', '' , { 'href' : self.feed['iTunes_image_url']})
        # that's from https://gitorious.org/podjango/podjango/commit/621857be0a3d7c44f1925c7daf471c38ea62c180?diffmode=sidebyside

        handler.addQuickElement(u'itunes:explicit', self.feed['iTunes_explicit'])
        handler.startElement(u"itunes:owner", {})
        handler.addQuickElement(u'itunes:name', self.feed['iTunes_name'])
        handler.addQuickElement(u'itunes:email', self.feed['iTunes_email'])
        handler.endElement(u"itunes:owner")

        # @ToDo: add categories

    def add_item_elements(self,  handler, item):
        super(iTunesPodcastsFeedGenerator, self).add_item_elements(handler, item)
        handler.addQuickElement(u'itunes:summary', unescape(item['summary']))
        handler.addQuickElement(u'itunes:explicit',item['explicit'])
        # handler.addQuickElement(u'itunes:image', item['iTunes_image_url'])
        handler.addQuickElement(u'itunes:image', '' , { 'href' : self.feed['iTunes_image_url']})


    # def __unicode__(self):
    #     return unicode(self.order_num)

class iTunesPodcastPost():
    def __init__(self, podcast): # note: podcast here = Track for me
        self.id = podcast.id
        # self.date_submitted = podcast.date_submitted
        self.pub_date = podcast.pub_date
        self.title = podcast.title if podcast.title else "Track"
        self.summary = unescape(podcast.description) if podcast.description else "Cool thing"
        # self.description = mark_safe("<![CDATA[%s]]>" % (podcast.description)) if podcast.description else ""
        self.description = podcast.description if podcast.description else "Cool thing"

        self.enclosure_url = podcast.awe_url # defined in models.py

        self.enclosure_length = podcast.size if podcast.size else 1 # u'unkown duration'
        self.enclosure_mime_type = u'audio/mpeg' # @ToDo generalise once we have other types
        self.explicit = u'yes' if podcast.explicit else u'no'
        self.url = podcast.get_absolute_url

        self.iTunes_image_url = podcast.main_image_url # if podcast.main_image_url else 'http://fun.com'

        self.length = podcast.time_duration if podcast.time_duration else 11 # "TBD"
        self.user_id = podcast.user_id
        self.user = User.objects.get(id = podcast.user_id)
        self.slug = podcast.slug
        self.duration = podcast.time_duration if podcast.time_duration else "5:00" # "Duration TBC"

        # if categories:
        #     categories = [to_unicode(c) for c in categories]
        # see https://docs.djangoproject.com/en/1.7/_modules/django/utils/feedgenerator/#SyndicationFeed.add_root_elements

    def __unicode__(self):
        return self.title

    def get_absolute_url(self):
        # return "%s" % self.url()
        # user = User.objects.get(id=self.user_id)
        return reverse('track_detail', args=[self.user.username, self.slug]) 


class iTunesPodcastsFeed(Feed):
    """
    A feed of podcasts for iTunes and other compatible podcatchers.
    Based on https://github.com/CaptainHayashi/django-lass-uryplayer/blob/master/uryplayer/feeds.py
    """

    def get_object(self, request, username, playlist_slug):
        self.request = request
        # return get_object_or_404(User, username=username)
        user = get_object_or_404(User, username=username)
        return get_object_or_404(Playlist, user_id=user.id, slug=playlist_slug)

    def link(self, playlist):
        # return self.request.build_absolute_uri("/@%s/" % user.username)
        user = User.objects.get(id=playlist.user_id)
        return reverse('playlist_index', args=[user.username, playlist.slug])

    def title(self, playlist):

        return playlist.title


    # description_template = mark_safe("defaults/playlist_description_missing.html")
    # not using that

    def description(self, playlist):
        if playlist.description:
            return playlist.description
            # return mark_safe("<![CDATA[%s]]>" % (playlist.description))
            # No, I won't wrap in CDATA until I can avoid escaping the "<" signs here

        else:
            return "[Auto text] The creator has not written a description."
            # return render_to_string("defaults/playlist_description_missing.txt")
            # pass


    def iTunes_image_url(self, obj): # TypeError: coercing to Unicode: need string or buffer, instancemethod found
        if obj.main_image_url:
            return unicode(obj.main_image_url) # obj.main_image_url
        else:
            return u'https://dl.dropboxusercontent.com/u/16441973/publicstatic/img/playlist-icon.png'


    # author_name = 'University Radio York'
    # modified from https://github.com/aneumeier/feeds/blob/master/feeds/rss.py
    def author_name(self, obj): # obj = playlist
        """
        Return the author for this feed.
        The feed is in `obj`, provided by `get_object`
        """
        if obj.author:
            return u"%s" % obj.author
        else:
            return 'Playlist created by %s' % (obj.user.username)

    def subtitle(self, obj): # obj = playlist
        """
        Return the author for this feed.
        The feed is in `obj`, provided by `get_object`
        """
        if obj.subtitle:
            return u"%s" % obj.author
        else:
            return '%s created in 2015' % (obj.title)

    # def summary(self, obj):
    #     return obj.description

    # @ToDo: finish adapting rest of this from the hard-coded URY values to actual values for my implementation

    iTunes_name = u'Hard-coded iTunes name for now'
    iTunes_email = u'm@rkmoriarty.com' 
    # @ToDo: make dynamic, not hard-coded


    iTunes_explicit = u'no'
    feed_type = iTunesPodcastsFeedGenerator
    feed_copyright = "Copyright 1967-%s University Radio York" % datetime.date.today().year



    def feed_extra_kwargs(self, playlist):
        extra = {}
        extra['iTunes_name'] = self.iTunes_name
        extra['iTunes_email'] = self.iTunes_email
        # extra['iTunes_image_url'] = self.iTunes_image_url
        def get_image(self, playlist):
            if playlist.main_image_url:
                return playlist.main_image_url
            else:
                return "https://dl.dropboxusercontent.com/u/16441973/publicstatic/img/rss_playlist_icon_placeholder.png"
                # @ToDo: replace with Awesound logo
            # return render_to_string("defaults/playlist_description_missing.txt")
            # pass

        extra['iTunes_image_url'] = get_image(self, playlist)
        extra['iTunes_explicit'] = self.iTunes_explicit

        return extra


    def items(self, playlist):
        """
        Returns a list of items to publish in this feed.
        """
        posts = playlist.tracks.all().order_by('-pub_date').order_by("-created_at")[:99]
        posts = [iTunesPodcastPost(item) for item in posts]
        return posts

    def item_extra_kwargs(self, item):
        return {'summary':unescape(mark_safe(item.description)), 
            'explicit':item.explicit,   
            'iTunes_image_url':item.iTunes_image_url}
            # was summary: item.summary

    # MarkAdded
    def item_link(self, item):
        # return item.enclosure_length
        if item.user_id:
            # we have a normal track created by a user
            # user = User.objects.get(id = item.user_id)
            return reverse('track_detail', args=[item.user.username, item.slug])
        else:
            # we have a funny track without a user, e.g., created via command line
            return 'Exception:TrackWithoutUser'

    def item_pubdate(self, item):
        return item.pub_date

    def item_enclosure_url(self, item):
        return item.enclosure_url


    def item_enclosure_length(self, item):
        # return item.enclosure_length
        return item.length

    def item_enclosure_mime_type(self, item):
        # return item.enclosure_mime_type
        return 'audio/mpeg' # @ToDo: make dynamic

    def item_description(self, item):
        # return item.summary
        if item.description:
            return unescape(mark_safe(item.description))
        else:
            return "User has not written a description. This is an automatic message"


# current_site = Site.objects.get_current()
current_site = 'https://greatsite.com'
iTunes_feed = iTunesPodcastsFeed()



### the above will be called if both username and playlist_slug are deteced in the url
### there are two older methods to handle other situations

class AllTracks(Feed):
    #
    # working old method, not relevant to html escaping question
    #


class UserTracks(AllTracks):
    #
    # working old method, not relevant to my question
    #

all_tracks = AllTracks()
user_tracks = UserTracks()

### note, both of those are also subject to full html escaping also



def choose_feed(request, *args, **kwargs):
    """
    Pick up the user feed or the global feed depending on whether or not the
    URL contains a username parameter
    """
    # feed = user_tracks if 'username' in kwargs else all_tracks
    if 'username' in kwargs:
        if 'playlist_slug' in kwargs:
            # feed = podcast_feed  
            slug = kwargs['playlist_slug']
            feed = iTunes_feed
            if request.user:
                user = request.user
                slack_message('slackmessages/playlist_feed.slack', { #django_slack/slackmessages/
                    'playlist': Playlist.objects.get(slug=slug),
                    'user':user,
                    })
        else:

            feed = user_tracks
    else:
        feed = all_tracks     


    return feed.__call__(request, *args, **kwargs)
Run Code Online (Sandbox Code Playgroud)

小智 1

因此,基于文档handler是一个XMLGenerator,并且调用addQuickElement假设所有内容都是字符数据。这就是为什么它被逃脱了。

您可能需要做的是使用覆盖SyndicationFeed.add_item_elements(self, handler, item)并插入元素,并使用添加标签。aaddQuickElementitunes:summarystartElementendElement

class iTunesFeed(Rss201rev2Feed):
    def add_item_elements(self, handler item):
        super(iTunesFeed, self).add_root_elements(handler)
        handler.startElement('itunes:summary')
        handler.characters('Link to ')            
        handler.addQuickElement('a', 'the website', {'href':'http://www.website.com'})
        handler.endElement('itunes:summary')
Run Code Online (Sandbox Code Playgroud)

这可能不是 100% 有效,但应该已经非常接近了。