def get(self):
expected_inputs = ('activitystreams', 'html', 'json-mf2')
input = util.get_required_param(self, 'input')
if input not in expected_inputs:
raise exc.HTTPBadRequest('Invalid input: %s, expected one of %r' %
(input, expected_inputs))
url = util.get_required_param(self, 'url')
# check if request is cached
cache = self.request.get('cache', '').lower() != 'false'
cache_key = 'U %s' % url
cached = memcache.get(cache_key) if cache else None
if cached:
logging.info('Serving cached response %r', cache_key)
url = cached['url']
body = cached['body']
else:
# fetch url
try:
resp = util.urlopen(url)
except (ValueError, httplib.InvalidURL) as e:
self.abort(400, str(e))
# other exceptions are handled by webutil.handlers.handle_exception(),
# which uses interpret_http_exception(), etc.
if url != resp.geturl():
url = resp.geturl()
logging.info('Redirected to %s', url)
body = resp.read()
if cache:
logging.info('Caching response in %r', cache_key)
memcache.set(cache_key, {'url': url, 'body': body}, URL_CACHE_TIME)
# decode data
mf2 = None
if input == 'html':
mf2 = mf2py.parse(doc=body, url=url)
elif input == 'json-mf2':
mf2 = json.loads(body)
mf2.setdefault('rels', {}) # mf2util expects rels
actor = None
title = None
if mf2:
actor = microformats2.find_author(
mf2, fetch_mf2_func=lambda url: mf2py.parse(url=url))
title = mf2util.interpret_feed(mf2, url).get('name')
if input == 'activitystreams':
activities = json.loads(body)
elif input == 'html':
activities = microformats2.html_to_activities(body, url, actor)
elif input == 'json-mf2':
activities = [microformats2.json_to_object(item, actor=actor)
for item in mf2.get('items', [])]
self.write_response(source.Source.make_activities_base_response(activities),
url=url, actor=actor, title=title)
def html_to_atom(html, url=None, fetch_author=False, reader=True):
"""Converts microformats2 HTML to an Atom feed.
Args:
html: string
url: string URL html came from, optional
fetch_author: boolean, whether to make HTTP request to fetch rel-author link
reader: boolean, whether the output will be rendered in a feed reader.
Currently just includes location if True, not otherwise.
Returns:
unicode string with Atom XML
"""
if fetch_author:
assert url, 'fetch_author=True requires url!'
parsed = mf2py.parse(doc=html, url=url)
actor = microformats2.find_author(
parsed, fetch_mf2_func=lambda url: mf2py.parse(url=url))
return activities_to_atom(
microformats2.html_to_activities(html, url, actor),
actor,
title=mf2util.interpret_feed(parsed, url).get('name'),
xml_base=util.base_url(url),
host_url=url,
reader=reader)
def add_subscription(origin, feed_url, type, tags=None):
feed = Feed.query.filter_by(feed=feed_url, type=type).first()
if not feed:
name = None
if type == "html":
flask.current_app.logger.debug("mf2py parsing %s", feed_url)
resp = util.requests_get(feed_url)
feed_text = resp.text if "charset" in resp.headers.get("content-type", "") else resp.content
parsed = mf2util.interpret_feed(mf2py.parse(doc=feed_text, url=feed_url), feed_url)
name = parsed.get("name")
elif type == "xml":
flask.current_app.logger.debug("feedparser parsing %s", feed_url)
parsed = feedparser.parse(feed_url, agent=util.USER_AGENT)
if parsed.feed:
name = parsed.feed.get("title")
else:
flask.current_app.logger.error("unknown feed type %s", type)
flask.abort(400)
if not name:
p = urllib.parse.urlparse(origin)
name = p.netloc + p.path
feed = Feed(name=name[:140], origin=origin, feed=feed_url, type=type)
if feed:
db.session.add(feed)
flask_login.current_user.subscriptions.append(Subscription(feed=feed, name=feed.name, tags=tags))
db.session.commit()
# go ahead and update the fed
tasks.q.enqueue(tasks.update_feed, feed.id)
return feed
def convert_mf2util():
def dates_to_string(json):
if isinstance(json, dict):
return {k: dates_to_string(v) for (k, v) in json.items()}
if isinstance(json, list):
return [dates_to_string(v) for v in json]
if isinstance(json, datetime.date) or isinstance(json, datetime.datetime):
return json.isoformat()
return json
url = request.args.get('url')
as_feed = request.args.get('as-feed')
op = request.args.get('op')
if url:
try:
d = mf2py.parse(url=url)
if op == 'post-type-discovery':
entry = mf2util.find_first_entry(d, ['h-entry', 'h-event'])
return jsonify({'type': mf2util.post_type_discovery(entry)})
if as_feed == 'true' or mf2util.find_first_entry(d, ['h-feed']):
json = mf2util.interpret_feed(d, url)
else:
json = mf2util.interpret(d, url)
return jsonify(dates_to_string(json))
except:
current_app.logger.exception('running mf2util service')
return jsonify({'error': str(sys.exc_info()[0])})
return """
def get_title(mf2):
"""Returns the author of a page as a ActivityStreams actor dict.
Args:
mf2: dict, parsed mf2 object (ie return value from mf2py.parse())
Returns: string title, possibly ellipsized
"""
lines = mf2util.interpret_feed(mf2, '').get('name', '').splitlines()
if lines:
return util.ellipsize(lines[0])
return ''
def get_title(mf2):
"""Returns an mf2 object's title, ie its name.
Args:
mf2: dict, parsed mf2 object (ie return value from mf2py.parse())
Returns: string title, possibly ellipsized
"""
lines = mf2util.interpret_feed(mf2, '').get('name', '').splitlines()
if lines:
return util.ellipsize(lines[0])
return ''
def get_title(mf2):
"""Returns an mf2 object's title, ie its name.
Args:
mf2: dict, parsed mf2 object (ie return value from mf2py.parse())
Returns: string title, possibly ellipsized
"""
lines = mf2util.interpret_feed(mf2, '').get('name', '').splitlines()
if lines:
return util.ellipsize(lines[0])
return ''
def process_html_feed_for_new_entries(feed, content, backfill, now):
# strip noscript tags before parsing, since we definitely aren't
# going to preserve js
content = re.sub('</?noscript[^>]*>', '', content)
parsed = mf2util.interpret_feed(
mf2py.parse(url=feed.feed, doc=content), feed.feed)
hfeed = parsed.get('entries', [])
for hentry in hfeed:
entry = hentry_to_entry(hentry, feed, backfill, now)
if entry:
current_app.logger.debug('built entry: %s', entry.permalink)
yield entry
def html_to_atom(html, url=None, **kwargs):
"""Converts microformats2 HTML to an Atom feed.
Args:
html: string
url: string URL html came from, optional
Returns: unicode string with Atom XML
"""
parsed = mf2py.parse(doc=html, url=url)
return activities_to_atom(microformats2.html_to_activities(html, url),
microformats2.find_author(parsed),
title=mf2util.interpret_feed(parsed,
url).get('name'),
xml_base=util.base_url(url),
host_url=url)
def get(self):
expected_inputs = ('activitystreams', 'html', 'json-mf2', 'jsonfeed')
input = util.get_required_param(self, 'input')
if input not in expected_inputs:
raise exc.HTTPBadRequest('Invalid input: %s, expected one of %r' %
(input, expected_inputs))
url, body = self._urlopen(util.get_required_param(self, 'url'))
# decode data
mf2 = None
if input == 'html':
mf2 = mf2py.parse(doc=body, url=url)
elif input == 'json-mf2':
mf2 = json.loads(body)
mf2.setdefault('rels', {}) # mf2util expects rels
actor = None
title = None
if mf2:
def fetch_mf2_func(url):
_, doc = self._urlopen(url)
return mf2py.parse(doc=doc, url=url)
actor = microformats2.find_author(mf2,
fetch_mf2_func=fetch_mf2_func)
title = mf2util.interpret_feed(mf2, url).get('name')
if input == 'activitystreams':
activities = json.loads(body)
elif input == 'html':
activities = microformats2.html_to_activities(body, url, actor)
elif input == 'json-mf2':
activities = [
microformats2.json_to_object(item, actor=actor)
for item in mf2.get('items', [])
]
elif input == 'jsonfeed':
activities, actor = jsonfeed.jsonfeed_to_activities(
json.loads(body))
self.write_response(
source.Source.make_activities_base_response(activities),
url=url,
actor=actor,
title=title)
def find_possible_feeds(origin):
# scrape an origin source to find possible alternative feeds
try:
resp = util.requests_get(origin)
except requests.exceptions.RequestException as e:
flask.flash("Error fetching source {}".format(repr(e)))
flask.current_app.logger.warn("Subscribe failed for %s with error %s", origin, repr(e))
return None
feeds = []
xml_feed_types = [
"application/rss+xml",
"application/atom+xml",
"application/rdf+xml",
"application/xml",
"text/xml",
]
xml_mime_types = xml_feed_types + ["text/xml", "text/rss+xml", "text/atom+xml"]
content_type = resp.headers["content-type"]
content_type = content_type.split(";", 1)[0].strip()
if content_type in xml_mime_types:
feeds.append({"origin": origin, "feed": origin, "type": "xml", "title": "untitled xml feed"})
elif content_type == "text/html":
parsed = mf2py.parse(doc=resp.text, url=origin)
# if text/html, then parse and look for h-entries
hfeed = mf2util.interpret_feed(parsed, origin)
if hfeed.get("entries"):
ftitle = hfeed.get("name") or "untitled h-feed"
feeds.append({"origin": origin, "feed": resp.url, "type": "html", "title": ftitle[:140]})
# look for link="feed"
for furl in parsed.get("rels", {}).get("feed", []):
fprops = parsed.get("rel-urls", {}).get(furl, {})
if not fprops.get("type") or fprops.get("type") == "text/html":
feeds.append({"origin": origin, "feed": furl, "type": "html", "title": fprops.get("title")})
# then look for link rel="alternate"
for link in parsed.get("alternates", []):
if link.get("type") in xml_feed_types:
feeds.append({"origin": origin, "feed": link.get("url"), "type": "xml", "title": link.get("title")})
return feeds
def test_h_feed_excludes_rel_syndication():
"""Represents a feed that (incorrectly) includes page-scoped
rel=syndication values in the feed itself. If we're not careful,
these values will be slurped into every entry in the feed.
"""
parsed = {
"items":[{
"type": ["h-entry"], "properties": {
"name": ["First Post"],
"url": ["http://example.com/first-post"],
"content": [{
"html": "This is the body of the first post",
"value": "This is the body of the first post",
}],
"syndication": [
"https://twitter.com/example_com/123456",
"https://www.facebook.com/example.com/123456",
],
},
}, {
"type": ["h-event"], "properties": {
"name": ["Second Post"],
"url": ["http://example.com/second-post"],
"content": [{
"html": "This is the body of the second post",
"value": "This is the body of the second post",
}],
"syndication": [
"https://twitter.com/example_com/7891011",
"https://www.facebook.com/example.com/7891011",
],
},
}], "rels": {
"syndication": [
"https://twitter.com/example_com/123456",
"https://twitter.com/example_com/7891011",
"https://www.facebook.com/example.com/123456",
"https://www.facebook.com/example.com/7891011"
],
}
}
result = mf2util.interpret_feed(parsed, 'http://example.com')
assert result['entries'][0]['syndication'] == ["https://twitter.com/example_com/123456", "https://www.facebook.com/example.com/123456"]
assert result['entries'][1]['syndication'] == ["https://twitter.com/example_com/7891011", "https://www.facebook.com/example.com/7891011"]
def convert_mf2util():
def dates_to_string(json):
if isinstance(json, dict):
return {k: dates_to_string(v) for (k, v) in json.items()}
if isinstance(json, list):
return [dates_to_string(v) for v in json]
if isinstance(json, datetime.date) or isinstance(json, datetime.datetime):
return json.isoformat()
return json
url = request.args.get('url')
if url:
d = mf2py.Parser(url=url).to_dict()
if mf2util.find_first_entry(d, ['h-feed']):
json = mf2util.interpret_feed(d, url)
else:
json = mf2util.interpret(d, url)
return jsonify(dates_to_string(json))
return """
def process_html_feed_for_new_entries(feed, content, backfill, now, fetch_mf2_func):
# strip noscript tags before parsing, since we definitely aren't
# going to preserve js
content = re.sub('</?noscript[^>]*>', '', content, flags=re.IGNORECASE)
# look for a <base> element
doc = bs4.BeautifulSoup(content, 'html5lib')
base_el = doc.find('base')
base_href = base_el.get('href') if base_el else None
parsed = mf2util.interpret_feed(
mf2py.parse(doc, feed.feed),
source_url=feed.feed, base_href=base_href,
fetch_mf2_func=fetch_mf2_func)
hfeed = parsed.get('entries', [])
for hentry in hfeed:
current_app.logger.debug('building entry: %s', hentry.get('url'))
entry = hentry_to_entry(hentry, feed, backfill, now)
if entry:
current_app.logger.debug('built entry: %s', entry.permalink)
yield entry
def add_subscription(origin, feed_url, type, tags=None):
feed = Feed.query.filter_by(feed=feed_url, type=type).first()
if not feed:
name = None
if type == 'html':
flask.current_app.logger.debug('mf2py parsing %s', feed_url)
resp = util.requests_get(feed_url)
feed_text = resp.text if 'charset' in resp.headers.get(
'content-type', '') else resp.content
parsed = mf2util.interpret_feed(
mf2py.parse(doc=feed_text, url=feed_url), feed_url)
name = parsed.get('name')
elif type == 'xml':
flask.current_app.logger.debug('feedparser parsing %s', feed_url)
parsed = feedparser.parse(feed_url, agent=util.USER_AGENT)
if parsed.feed:
name = parsed.feed.get('title')
else:
flask.current_app.logger.error('unknown feed type %s', type)
flask.abort(400)
if not name:
p = urllib.parse.urlparse(origin)
name = p.netloc + p.path
feed = Feed(name=name[:140], origin=origin, feed=feed_url, type=type)
if feed:
db.session.add(feed)
flask_login.current_user.subscriptions.append(
Subscription(feed=feed, name=feed.name, tags=tags))
db.session.commit()
# go ahead and update the fed
tasks.q.enqueue(tasks.update_feed, feed.id)
return feed
def html_to_atom(html, url=None, fetch_author=False):
"""Converts microformats2 HTML to an Atom feed.
Args:
html: string
url: string URL html came from, optional
fetch_author: boolean, whether to make HTTP request to fetch rel-author link
Returns:
unicode string with Atom XML
"""
if fetch_author:
assert url, 'fetch_author=True requires url!'
parsed = mf2py.parse(doc=html, url=url)
actor = microformats2.find_author(
parsed, fetch_mf2_func=lambda url: mf2py.parse(url=url))
return activities_to_atom(
microformats2.html_to_activities(html, url, actor),
actor,
title=mf2util.interpret_feed(parsed, url).get('name'),
xml_base=util.base_url(url),
host_url=url)
def get(self):
expected_inputs = ('activitystreams', 'html', 'json-mf2')
input = util.get_required_param(self, 'input')
if input not in expected_inputs:
raise exc.HTTPBadRequest('Invalid input: %s, expected one of %r' %
(input, expected_inputs))
# fetch url
url = util.get_required_param(self, 'url')
resp = util.urlopen(url)
if url != resp.geturl():
url = resp.geturl()
logging.info('Redirected to %s', url)
body = resp.read()
# decode data
mf2 = None
if input == 'activitystreams':
activities = json.loads(body)
elif input == 'html':
activities = microformats2.html_to_activities(body, url)
mf2 = mf2py.parse(doc=body, url=url)
elif input == 'json-mf2':
mf2 = json.loads(body)
mf2['rels'] = {} # mf2util expects rels
activities = [microformats2.json_to_object(item)
for item in mf2.get('items', [])]
author = None
title = None
if mf2:
author = microformats2.find_author(mf2)
title = mf2util.interpret_feed(mf2, url).get('name')
self.write_response(source.Source.make_activities_base_response(activities),
url=url, actor=author, title=title)
def convert_mf2util():
def dates_to_string(json):
if isinstance(json, dict):
return {k: dates_to_string(v) for (k, v) in json.items()}
if isinstance(json, list):
return [dates_to_string(v) for v in json]
if isinstance(json, datetime.date) or isinstance(json, datetime.datetime):
return json.isoformat()
return json
url = request.args.get('url')
as_feed = request.args.get('as-feed')
if url:
try:
d = mf2py.parse(url=url)
if as_feed == 'true' or mf2util.find_first_entry(d, ['h-feed']):
json = mf2util.interpret_feed(d, url)
else:
json = mf2util.interpret(d, url)
return jsonify(dates_to_string(json))
except:
return jsonify({'error': str(sys.exc_info()[0])})
return """
def get(self):
expected_inputs = ('activitystreams', 'html', 'json-mf2')
input = util.get_required_param(self, 'input')
if input not in expected_inputs:
raise exc.HTTPBadRequest('Invalid input: %s, expected one of %r' %
(input, expected_inputs))
url = util.get_required_param(self, 'url')
# check if request is cached
cache = self.request.get('cache', '').lower() != 'false'
cache_key = 'U %s' % url
cached = memcache.get(cache_key) if cache else None
if cached:
logging.info('Serving cached response %r', cache_key)
url = cached['url']
body = cached['body']
else:
# fetch url
try:
resp = util.urlopen(url)
except (ValueError, httplib.InvalidURL) as e:
self.abort(400, str(e))
except Exception as e:
if util.is_connection_failure(e):
# HTTP 504 Gateway Timeout
self.abort(504, str(e))
raise
if url != resp.geturl():
url = resp.geturl()
logging.info('Redirected to %s', url)
body = resp.read()
if cache:
logging.info('Caching response in %r', cache_key)
memcache.set(cache_key, {'url': url, 'body': body}, URL_CACHE_TIME)
# decode data
mf2 = None
if input == 'html':
mf2 = mf2py.parse(doc=body, url=url)
elif input == 'json-mf2':
mf2 = json.loads(body)
mf2.setdefault('rels', {}) # mf2util expects rels
actor = None
title = None
if mf2:
actor = microformats2.find_author(
mf2, fetch_mf2_func=lambda url: mf2py.parse(url=url))
title = mf2util.interpret_feed(mf2, url).get('name')
if input == 'activitystreams':
activities = json.loads(body)
elif input == 'html':
activities = microformats2.html_to_activities(body, url, actor)
elif input == 'json-mf2':
activities = [microformats2.json_to_object(item, actor=actor)
for item in mf2.get('items', [])]
self.write_response(source.Source.make_activities_base_response(activities),
url=url, actor=actor, title=title)
def find_possible_feeds(origin):
# scrape an origin source to find possible alternative feeds
try:
resp = util.requests_get(origin)
except requests.exceptions.RequestException as e:
flask.flash('Error fetching source {}'.format(repr(e)))
flask.current_app.logger.warn(
'Subscribe failed for %s with error %s', origin, repr(e))
return None
feeds = []
xml_feed_types = [
'application/rss+xml',
'application/atom+xml',
'application/rdf+xml',
'application/xml',
'text/xml',
]
xml_mime_types = xml_feed_types + [
'text/xml',
'text/rss+xml',
'text/atom+xml',
]
html_feed_types = [
'text/html',
'application/xhtml+xml',
]
content_type = resp.headers['content-type']
content_type = content_type.split(';', 1)[0].strip()
if content_type in xml_mime_types:
feeds.append({
'origin': origin,
'feed': origin,
'type': 'xml',
'title': 'untitled xml feed',
})
elif content_type in html_feed_types:
parsed = mf2py.parse(doc=resp.text, url=origin)
# if text/html, then parse and look for h-entries
hfeed = mf2util.interpret_feed(parsed, origin)
if hfeed.get('entries'):
ftitle = hfeed.get('name') or 'untitled h-feed'
feeds.append({
'origin': origin,
'feed': resp.url,
'type': 'html',
'title': ftitle[:140]
})
# look for link="feed"
for furl in parsed.get('rels', {}).get('feed', []):
fprops = parsed.get('rel-urls', {}).get(furl, {})
if not fprops.get('type') or fprops.get('type') in html_feed_types:
feeds.append({
'origin': origin,
'feed': furl,
'type': 'html',
'title': fprops.get('title'),
})
# then look for link rel="alternate"
for link in parsed.get('alternates', []):
if link.get('type') in xml_feed_types:
feeds.append({
'origin': origin,
'feed': link.get('url'),
'type': 'xml',
'title': link.get('title'),
})
return feeds
def find_possible_feeds(origin):
# scrape an origin source to find possible alternative feeds
try:
resp = util.requests_get(origin)
except requests.exceptions.RequestException as e:
flask.flash('Error fetching source {}'.format(repr(e)))
flask.current_app.logger.warn('Subscribe failed for %s with error %s',
origin, repr(e))
return None
feeds = []
xml_feed_types = [
'application/rss+xml',
'application/atom+xml',
'application/rdf+xml',
'application/xml',
'text/xml',
]
xml_mime_types = xml_feed_types + [
'text/xml',
'text/rss+xml',
'text/atom+xml',
]
html_feed_types = [
'text/html',
'application/xhtml+xml',
]
content_type = resp.headers['content-type']
content_type = content_type.split(';', 1)[0].strip()
if content_type in xml_mime_types:
feeds.append({
'origin': origin,
'feed': origin,
'type': 'xml',
'title': 'untitled xml feed',
})
elif content_type in html_feed_types:
parsed = mf2py.parse(doc=resp.text, url=origin)
# if text/html, then parse and look for h-entries
hfeed = mf2util.interpret_feed(parsed, origin)
if hfeed.get('entries'):
ftitle = hfeed.get('name') or 'untitled h-feed'
feeds.append({
'origin': origin,
'feed': resp.url,
'type': 'html',
'title': ftitle[:140]
})
# look for link="feed"
for furl in parsed.get('rels', {}).get('feed', []):
fprops = parsed.get('rel-urls', {}).get(furl, {})
if not fprops.get('type') or fprops.get('type') in html_feed_types:
feeds.append({
'origin': origin,
'feed': furl,
'type': 'html',
'title': fprops.get('title'),
})
# then look for link rel="alternate"
for link in parsed.get('alternates', []):
if link.get('type') in xml_feed_types:
feeds.append({
'origin': origin,
'feed': link.get('url'),
'type': 'xml',
'title': link.get('title'),
})
return feeds
