Merged in jaesivsm/pyaggr3g470r (pull request #15)

the icon feature

2 files changed, 115 insertions, 13 deletions

diff --git a/pyaggr3g470r/lib/crawler.py b/pyaggr3g470r/lib/crawler.py
index 324f0d8e..2ba5403a 100644
--- a/pyaggr3g470r/lib/crawler.py
+++ b/pyaggr3g470r/lib/crawler.py
@@ -24,7 +24,7 @@ from datetime import datetime
 from time import strftime, gmtime
 from concurrent.futures import ThreadPoolExecutor
 from requests_futures.sessions import FuturesSession
-from pyaggr3g470r.lib.utils import default_handler
+from pyaggr3g470r.lib.utils import default_handler, construct_feed_from
 
 logger = logging.getLogger(__name__)
 logging.captureWarnings(True)
@@ -136,7 +136,7 @@ class PyAggUpdater(AbstractCrawler):
         self.feed = feed
         self.entries = entries
         self.headers = headers
-        self.parsed_feed = parsed_feed.get('feed', {})
+        self.parsed_feed = parsed_feed
         super(PyAggUpdater, self).__init__(auth)
 
     def to_article(self, entry):
@@ -188,19 +188,26 @@ class PyAggUpdater(AbstractCrawler):
                      self.headers.get('etag', ''),
                      self.headers.get('last-modified', ''))
 
-        dico = {'error_count': 0, 'last_error': None,
-                'etag': self.headers.get('etag', ''),
-                'last_modified': self.headers.get('last-modified',
-                    strftime('%a, %d %b %Y %X %Z', gmtime())),
-                'site_link': self.parsed_feed.get('link')}
+        up_feed = {'error_count': 0, 'last_error': None,
+                   'etag': self.headers.get('etag', ''),
+                   'last_modified': self.headers.get('last-modified',
+                                    strftime('%a, %d %b %Y %X %Z', gmtime()))}
+        fresh_feed = construct_feed_from(url=self.feed['link'],
+                                         fp_parsed=self.parsed_feed,
+                                         feed=self.feed)
+        for key in ('description', 'site_link', 'icon'):
+            if fresh_feed.get(key) and fresh_feed[key] != self.feed.get(key):
+                up_feed[key] = fresh_feed[key]
         if not self.feed.get('title'):
-            dico['title'] = self.parsed_feed.get('title', '')
+            up_feed['title'] = fresh_feed.get('title', '')
+
         logger.info('%r %r - pushing feed attrs %r',
                     self.feed['id'], self.feed['title'],
-                    {key: "%s -> %s" % (dico[key], self.feed.get(key))
-                     for key in dico if dico[key] != self.feed.get(key)})
-        if any([dico[key] != self.feed.get(key) for key in dico]):
-            future = self.query_pyagg('put', 'feed/%d' % self.feed['id'], dico)
+                    {key: "%s -> %s" % (up_feed[key], self.feed.get(key))
+                     for key in up_feed if up_feed[key] != self.feed.get(key)})
+        if any([up_feed[key] != self.feed.get(key) for key in up_feed]):
+            future = self.query_pyagg('put',
+                    'feed/%d' % self.feed['id'], up_feed)
             future.add_done_callback(self.get_counter_callback())
 
 
@@ -265,7 +272,7 @@ class FeedCrawler(AbstractCrawler):
                     self.feed['id'], self.feed['title'])
 
         ids, entries = [], {}
-        parsed_response = feedparser.parse(response.text)
+        parsed_response = feedparser.parse(response.content)
         for entry in parsed_response['entries']:
             entry_ids = extract_id(entry)
             entry_ids['feed_id'] = self.feed['id']
diff --git a/pyaggr3g470r/lib/utils.py b/pyaggr3g470r/lib/utils.py
index a4f4b3ec..a0154b7f 100644
--- a/pyaggr3g470r/lib/utils.py
+++ b/pyaggr3g470r/lib/utils.py
@@ -1,4 +1,9 @@
 import types
+import urllib
+import requests
+import feedparser
+from bs4 import BeautifulSoup, SoupStrainer
+
 
 def default_handler(obj):
     """JSON handler for default query formatting"""
@@ -12,3 +17,93 @@ def default_handler(obj):
         return str(obj)
     raise TypeError("Object of type %s with value of %r "
                     "is not JSON serializable" % (type(obj), obj))
+
+
+def try_keys(dico, *keys):
+    for key in keys:
+        if key in dico:
+            return dico[key]
+    return
+
+
+def rebuild_url(url, base_split):
+    split = urllib.parse.urlsplit(url)
+    if split.scheme and split.netloc:
+        return url  # url is fine
+    new_split = urllib.parse.SplitResult(
+            scheme=split.scheme or base_split.scheme,
+            netloc=split.netloc or base_split.netloc,
+            path=split.path, query='', fragment='')
+    return urllib.parse.urlunsplit(new_split)
+
+
+def construct_feed_from(url=None, fp_parsed=None, feed=None, query_site=True):
+    if url is None and fp_parsed is not None:
+        url = fp_parsed.get('url')
+    if url is not None and fp_parsed is None:
+        response = requests.get(url, verify=False)
+        fp_parsed = feedparser.parse(response.content)
+    assert url is not None and fp_parsed is not None
+    feed = feed or {}
+    split = urllib.parse.urlsplit(url)
+    if not fp_parsed['bozo']:
+        feed['link'] = url
+        feed['site_link'] = try_keys(fp_parsed['feed'], 'href', 'link')
+        feed['title'] = fp_parsed['feed'].get('title')
+        feed['description'] = try_keys(fp_parsed['feed'], 'subtitle', 'title')
+        feed['icon'] = try_keys(fp_parsed['feed'], 'icon')
+    else:
+        feed['site_link'] = url
+
+    if feed.get('site_link'):
+        feed['site_link'] = rebuild_url(feed['site_link'], split)
+        split = urllib.parse.urlsplit(feed['site_link'])
+
+    if feed.get('icon'):
+        feed['icon'] = rebuild_url(feed['icon'], split)
+
+    if not feed.get('site_link') or not query_site \
+            or all(bool(feed.get(key)) for key in ('link', 'title', 'icon')):
+        return feed
+
+    response = requests.get(feed['site_link'], verify=False)
+    bs_parsed = BeautifulSoup(response.content, 'html.parser',
+                           parse_only=SoupStrainer('head'))
+
+    if not feed.get('title'):
+        try:
+            feed['title'] = bs_parsed.find_all('title')[0].text
+        except Exception:
+            pass
+
+    def check_keys(**kwargs):
+        def wrapper(elem):
+            for key, vals in kwargs.items():
+                if not elem.has_attr(key):
+                    return False
+                if not all(val in elem.attrs[key] for val in vals):
+                    return False
+            return True
+        return wrapper
+
+    if not feed.get('icon'):
+        icons = bs_parsed.find_all(check_keys(rel=['icon', 'shortcut']))
+        if not len(icons):
+            icons = bs_parsed.find_all(check_keys(rel=['icon']))
+        if len(icons) >= 1:
+            feed['icon'] = rebuild_url(icons[0].attrs['href'], split)
+        else:  # trying the default one
+            icon = rebuild_url('/favicon.ico', split)
+            if requests.get(icon, verify=False).ok:
+                feed['icon'] = icon
+
+    if not feed.get('link'):
+        alternate = bs_parsed.find_all(check_keys(rel=['alternate'],
+                type=['application/rss+xml']))
+        if len(alternate) == 1:
+            feed['link'] = rebuild_url(alternate[0].attrs['href'], split)
+        elif len(alternate) > 1:
+            feed['link'] = rebuild_url(alternate[0].attrs['href'], split)
+            feed['other_link'] = [rebuild_url(al.attrs['href'], split)
+                                  for al in alternate[1:]]
+    return feed