Python util.domain_from_link函数代码示例

  def _urls_and_domains(self, auth_entity, user_url):
    """Returns this user's valid (not webmention-blacklisted) URLs and domains.

    Converts the auth entity's user_json to an ActivityStreams actor and uses
    its 'urls' and 'url' fields. May be overridden by subclasses.

    Args:
      auth_entity: oauth_dropins.models.BaseAuth
      user_url: string, optional URL passed in when authorizing

    Returns: ([string url, ...], [string domain, ...])
    """
    actor = self.gr_source.user_to_actor(json.loads(auth_entity.user_json))
    logging.debug('Converted to actor: %s', json.dumps(actor, indent=2))

    urls = []
    for url in util.trim_nulls(util.uniquify(
        [user_url] + [actor.get('url')] +
        [u.get('value') for u in actor.get('urls', [])])):
      domain = util.domain_from_link(url)
      if domain and not util.in_webmention_blacklist(domain.lower()):
        urls.append(url)

    urls = util.dedupe_urls(urls)
    domains = [util.domain_from_link(url).lower() for url in urls]
    return urls, domains

  def test_domain_from_link(self):
    self.assertEqual('localhost', util.domain_from_link('http://localhost/foo'))
    self.assertEqual('a.b.c.d', util.domain_from_link('http://a.b.c.d/foo'))
    for good_link in ('asdf.com', 'www.asdf.com', 'https://asdf.com/',
                      'asdf.com/foo?bar#baz', 'm.asdf.com', 'asdf.com:1234',
                      'mobile.asdf.com/foo?bar#baz', '//asdf.com/foo/bar',
                      'https://m.asdf.com/foo?bar#baz'):
      actual = util.domain_from_link(good_link)
      self.assertEqual('asdf.com', actual, '%s returned %s' % (good_link, actual))

    self.assertEqual('asdf.com.', util.domain_from_link('http://asdf.com./x'))

    for bad_link in '', '  ', 'a&b.com', 'http://', 'file:///':
      self.assertEquals(None, util.domain_from_link(bad_link))

  def _urls_and_domains(self, auth_entity, user_url):
    """Returns this user's valid (not webmention-blacklisted) URLs and domains.

    Converts the auth entity's user_json to an ActivityStreams actor and uses
    its 'urls' and 'url' fields. May be overridden by subclasses.

    Args:
      auth_entity: oauth_dropins.models.BaseAuth
      user_url: string, optional URL passed in when authorizing

    Returns: ([string url, ...], [string domain, ...])
    """
    actor = self.gr_source.user_to_actor(json.loads(auth_entity.user_json))
    logging.debug('Converted to actor: %s', json.dumps(actor, indent=2))

    candidates = util.trim_nulls(util.uniquify(
        [user_url] + microformats2.object_urls(actor)))

    if len(candidates) > MAX_AUTHOR_URLS:
      logging.warning('Too many profile links! Only resolving the first %s: %s',
                      MAX_AUTHOR_URLS, candidates)

    urls = []
    for i, url in enumerate(candidates):
      url, domain, send = util.get_webmention_target(url, resolve=i < MAX_AUTHOR_URLS)
      if send:
        urls.append(url)

    urls = util.dedupe_urls(urls)  # normalizes domains to lower case
    domains = [util.domain_from_link(url) for url in urls]
    return urls, domains

def _process_syndication_urls(source, permalink, syndication_urls):
  """Process a list of syndication URLs looking for one that matches the
  current source.  If one is found, stores a new SyndicatedPost in the
  db.

  Args:
    source: a models.Source subclass
    permalink: a string. the current h-entry permalink
    syndication_urls: a collection of strings. the unfitered list
      of syndication_urls
  """

  results = {}
  # save the results (or lack thereof) to the db, and put them in a
  # map for immediate use
  for syndication_url in syndication_urls:
    # follow redirects to give us the canonical syndication url --
    # gives the best chance of finding a match.
    syndication_url = util.follow_redirects(syndication_url).url
    # source-specific logic to standardize the URL. (e.g., replace facebook
    # username with numeric id)
    syndication_url = source.canonicalize_syndication_url(syndication_url)
    # check that the syndicated url belongs to this source TODO save future
    # lookups by saving results for other sources too (note: query the
    # appropriate source subclass by author.domains, rather than
    # author.domain_urls)
    if util.domain_from_link(syndication_url) == source.AS_CLASS.DOMAIN:
      logging.debug('saving discovered relationship %s -> %s',
                    syndication_url, permalink)
      relationship = SyndicatedPost.insert(
        source, syndication=syndication_url, original=permalink)
      results.setdefault(syndication_url, []).append(relationship)
  return results

  def post(self):
    source = self.load_source()

    # validate URL, find silo post
    url = util.get_required_param(self, 'url')
    domain = util.domain_from_link(url)
    path = urlparse.urlparse(url).path
    msg = 'Discovering now. Refresh in a minute to see the results!'

    if domain == source.GR_CLASS.DOMAIN:
      post_id = source.GR_CLASS.post_id(url)
      if post_id:
        type = 'event' if path.startswith('/events/') else None
        util.add_discover_task(source, post_id, type=type)
      else:
        msg = "Sorry, that doesn't look like a %s post URL." % source.GR_CLASS.NAME

    elif util.domain_or_parent_in(domain, source.domains):
      synd_links = original_post_discovery.process_entry(source, url, {}, False, [])
      if synd_links:
        for link in synd_links:
          util.add_discover_task(source, source.GR_CLASS.post_id(link))
        source.updates = {'last_syndication_url': util.now_fn()}
        models.Source.put_updates(source)
      else:
        msg = 'Failed to fetch %s or find a %s syndication link.' % (
          util.pretty_link(url), source.GR_CLASS.NAME)

    else:
      msg = 'Please enter a URL on either your web site or %s.' % source.GR_CLASS.NAME

    self.messages.add(msg)
    self.redirect(source.bridgy_url(self))

  def new(handler, auth_entity=None, **kwargs):
    """Creates and returns a WordPress for the logged in user.

    Args:
      handler: the current RequestHandler
      auth_entity: oauth_dropins.wordpress.WordPressAuth
    """
    auth_domain = auth_entity.key.id()
    site_info = WordPress.get_site_info(handler, auth_entity)
    if site_info is None:
      return

    urls = util.dedupe_urls(util.trim_nulls(
      [site_info.get('URL'), auth_entity.blog_url]))
    domains = [util.domain_from_link(u) for u in urls]

    avatar = (json.loads(auth_entity.user_json).get('avatar_URL')
              if auth_entity.user_json else None)
    return WordPress(id=domains[0],
                     auth_entity=auth_entity.key,
                     name=auth_entity.user_display_name(),
                     picture=avatar,
                     superfeedr_secret=util.generate_secret(),
                     url=urls[0],
                     domain_urls=urls,
                     domains=domains,
                     site_info=site_info,
                     **kwargs)

  def new(handler, auth_entity=None, **kwargs):
    """Creates and returns a WordPress for the logged in user.

    Args:
      handler: the current RequestHandler
      auth_entity: oauth_dropins.wordpress.WordPressAuth
    """
    # Fetch blog's site info
    auth_domain = auth_entity.key.id()
    site_info = json.loads(auth_entity.urlopen(
        API_SITE_URL % auth_entity.blog_id).read())
    site_url = site_info.get('URL')
    if site_url:
      domains = [util.domain_from_link(site_url), auth_domain]
      urls = [site_url, auth_entity.blog_url]
    else:
      domains = [auth_domain]
      urls = [auth_entity.blog_url]

    avatar = (json.loads(auth_entity.user_json).get('avatar_URL')
              if auth_entity.user_json else None)
    return WordPress(id=domains[0],
                     auth_entity=auth_entity.key,
                     name=auth_entity.user_display_name(),
                     picture=avatar,
                     superfeedr_secret=util.generate_secret(),
                     url=urls[0],
                     domain_urls=urls,
                     domains=domains,
                     site_info=site_info,
                     **kwargs)

def handle_feed(feed, source):
  """Handles a Superfeedr JSON feed.

  Creates :class:`models.BlogPost` entities and adds propagate-blogpost tasks
  for new items.

  http://documentation.superfeedr.com/schema.html#json
  http://documentation.superfeedr.com/subscribers.html#pubsubhubbubnotifications

  Args:
    feed: unicode string, Superfeedr JSON feed
    source: Blogger, Tumblr, or WordPress
  """
  logging.info('Source: %s %s', source.label(), source.key.string_id())
  logging.info('Raw feed: %s', feed)

  if source.status != 'enabled':
    logging.info('Dropping because source is %s', source.status)
    return
  elif 'webmention' not in source.features:
    logging.info("Dropping because source doesn't have webmention feature")
    return

  for item in json.loads(feed).get('items', []):
    url = item.get('permalinkUrl') or item.get('id')
    if not url:
      logging.error('Dropping feed item without permalinkUrl or id!')
      continue

    # extract links from content, discarding self links.
    #
    # i don't use get_webmention_target[s]() here because they follows redirects
    # and fetch link contents, and this handler should be small and fast and try
    # to return a response to superfeedr successfully.
    #
    # TODO: extract_links currently has a bug that makes it drop trailing
    # slashes. ugh. fix that.
    content = item.get('content') or item.get('summary', '')
    links = [util.clean_url(util.unwrap_t_umblr_com(l))
             for l in util.extract_links(content)
             if util.domain_from_link(l) not in source.domains]

    unique = []
    for link in util.dedupe_urls(links):
      if len(link) <= _MAX_STRING_LENGTH:
        unique.append(link)
      else:
        logging.info('Giving up on link over %s chars! %s', _MAX_STRING_LENGTH, link)

    logging.info('Found links: %s', unique)
    if len(url) > _MAX_KEYPART_BYTES:
      logging.warning('Blog post URL is too long (over 500 chars)! Giving up.')
      bp = models.BlogPost(id=url[:_MAX_KEYPART_BYTES], source=source.key,
                           feed_item=item, failed=unique)
    else:
      bp = models.BlogPost(id=url, source=source.key, feed_item=item, unsent=unique)

    bp.get_or_save()

 def post(self):
   logging.debug('Params: %s', self.request.params)
   if self.lease(ndb.Key(urlsafe=self.request.params['key'])):
     # skip "self" links to this blog's domain
     source_domains = self.entity.source.get().domains
     to_send = set()
     for url in self.entity.unsent:
       link_domain = util.domain_from_link(url)
       if link_domain and link_domain not in source_domains:
         to_send.add(url)
     self.entity.unsent = list(to_send)
     self.send_webmentions()

  def search_for_links(self):
    """Searches for activities with links to any of this source's web sites.

    G+ search supports OR:
    https://developers.google.com/+/api/latest/activities/search

    Returns: sequence of ActivityStreams activity dicts
    """
    query = ' OR '.join(
      '"%s"' % util.fragmentless(url) for url in self.domain_urls
      if not util.in_webmention_blacklist(util.domain_from_link(url)))
    return self.get_activities(
      search_query=query, group_id=gr_source.SEARCH, etag=self.last_activities_etag,
      fetch_replies=False, fetch_likes=False, fetch_shares=False, count=50)

  def _urls_and_domains(auth_entity, blog_name=None):
    """Returns this blog's URL and domain.

    Args:
      auth_entity: oauth_dropins.tumblr.TumblrAuth
      blog_name: which blog. optional. matches the 'name' field for one of the
        blogs in auth_entity.user_json['user']['blogs'].

    Returns: ([string url], [string domain])
    """
    for blog in json.loads(auth_entity.user_json).get('user', {}).get('blogs', []):
      if ((blog_name and blog_name == blog.get('name')) or
          (not blog_name and blog.get('primary'))):
        return [blog['url']], [util.domain_from_link(blog['url']).lower()]

    return [], []

  def _url_and_domain(auth_entity, blog_name=None):
    """Returns the blog URL and domain.

    Args:
      auth_entity: oauth_dropins.tumblr.TumblrAuth
      blog_name: which blog. optional. matches the 'name' field for one of the
        blogs in auth_entity.user_json['user']['blogs'].

    Returns: (string url, string domain, boolean ok)
    """
    for blog in json.loads(auth_entity.user_json).get('user', {}).get('blogs', []):
      if ((blog_name and blog_name == blog.get('name')) or
          (not blog_name and blog.get('primary'))):
        return blog['url'], util.domain_from_link(blog['url']), True

    return None, None, False

def handle_feed(feed, source):
  """Handles a Superfeedr JSON feed.

  Creates BlogPost entities and adds propagate-blogpost tasks for new items.

  http://documentation.superfeedr.com/schema.html#json
  http://documentation.superfeedr.com/subscribers.html#pubsubhubbubnotifications

  Args:
    feed: string, Superfeedr JSON feed
    source: Blogger, Tumblr, or WordPress
  """
  logging.info('Source: %s %s', source.label(), source.key.string_id())
  logging.info('Raw feed: %s', feed)

  if source.status != 'enabled':
    logging.warning('Dropping because source is %s', source.status)
    return
  elif 'webmention' not in source.features:
    logging.warning("Dropping because source doesn't have webmention feature")
    return

  for item in json.loads(feed).get('items', []):
    url = item.get('permalinkUrl') or item.get('id')
    if not url:
      logging.error('Dropping feed item without permalinkUrl or id!')
      continue

    source.preprocess_superfeedr_item(item)
    # extract links from content, discarding self links.
    #
    # i don't use get_webmention_target[s]() here because they follows redirects
    # and fetch link contents, and this handler should be small and fast and try
    # to return a response to superfeedr successfully.
    #
    # TODO: extract_links currently has a bug that makes it drop trailing
    # slashes. ugh. fix that.
    content = item.get('content') or item.get('summary', '')
    links = [l for l in util.extract_links(content)
             if util.domain_from_link(l) not in source.domains]

    logging.info('Found links: %s', links)
    models.BlogPost(id=url,
                    source=source.key,
                    feed_item=item,
                    unsent=links,
                    ).get_or_save()

  def _urls_and_domains(self, auth_entity, user_url):
    """Returns this user's valid (not webmention-blacklisted) URLs and domains.

    Converts the auth entity's user_json to an ActivityStreams actor and uses
    its 'urls' and 'url' fields. May be overridden by subclasses.

    Args:
      auth_entity: :class:`oauth_dropins.models.BaseAuth`
      user_url: string, optional URL passed in when authorizing

    Returns:
      ([string url, ...], [string domain, ...])
    """
    actor = self.gr_source.user_to_actor(json.loads(auth_entity.user_json))
    logging.debug('Converted to actor: %s', json.dumps(actor, indent=2))

    candidates = util.trim_nulls(util.uniquify(
        [user_url] + microformats2.object_urls(actor)))

    if len(candidates) > MAX_AUTHOR_URLS:
      logging.info('Too many profile links! Only resolving the first %s: %s',
                   MAX_AUTHOR_URLS, candidates)

    urls = []
    for i, url in enumerate(candidates):
      final, domain, ok = util.get_webmention_target(url, resolve=i < MAX_AUTHOR_URLS)
      if ok:
        final = final.lower()
        if util.schemeless(final).startswith(util.schemeless(url.lower())):
          # redirected to a deeper path. use the original higher level URL. #652
          final = url
        # If final has a path segment check if root has a matching rel=me.
        match = re.match(r'^(https?://[^/]+)/.+', final)
        if match and i < MAX_AUTHOR_URLS:
          root = match.group(1)
          resp = util.requests_get(root)
          resp.raise_for_status()
          data = util.mf2py_parse(resp.text, root)
          me_urls = data.get('rels', {}).get('me', [])
          if final in me_urls:
            final = root
        urls.append(final)

    urls = util.dedupe_urls(urls)  # normalizes domains to lower case
    domains = [util.domain_from_link(url) for url in urls]
    return urls, domains

  def authorize(self):
    """Check for a backlink to brid.gy/publish/SILO."""
    bases = set()
    if util.domain_from_link(self.request.host_url) == 'brid.gy':
      bases.add('brid.gy')
      bases.add('www.brid.gy')  # also accept www
    else:
      bases.add(self.request.host_url)

    expected = ['%s/publish/%s' % (base, self.source.SHORT_NAME) for base in bases]

    if self.entity.html:
      for url in expected:
        if url in self.entity.html or urllib.quote(url, safe='') in self.entity.html:
          return True

    self.error("Couldn't find link to %s" % expected[0])
    return False