Store only last 20 articles for ecrans backend

modules/ecrans/backend.py

@@ -18,11 +18,13 @@
 # along with weboob. If not, see <http://www.gnu.org/licenses/>.
 "backend for http://www.ecrans.fr"
 
-from weboob.capabilities.messages import ICapMessages
+import time 
+
+from weboob.capabilities.messages import ICapMessages, Thread
 from weboob.tools.capabilities.messages.GenericBackend import GenericNewspaperBackend
 from .browser import NewspaperEcransBrowser
-from .tools import rssid
+from .tools import rssid, url2id
-
+from weboob.tools.newsfeed import Newsfeed
 
 class NewspaperEcransBackend(GenericNewspaperBackend, ICapMessages):
     MAINTAINER = 'Julien Hebert'
@@ -36,4 +38,30 @@ class NewspaperEcransBackend(GenericNewspaperBackend, ICapMessages):
     RSS_FEED = 'http://www.ecrans.fr/spip.php?page=backend'
     RSSID = rssid
 
+    def set_message_read(self, message):
+        self.storage.set(
+            'seen',
+            message.thread.id,
+            'comments',
+            self.storage.get(
+                'seen',
+                message.thread.id,
+                'comments',
+                default=[]) + [message.id])
+
+        lastpurge = self.storage.get('lastpurge', default=0)
+        l = []
+        if time.time() - lastpurge > 7200:
+            self.storage.set('lastpurge', time.time())
+            # Get lasts 20 articles
+            for id in self.storage.get('seen', default={}): 
+                 l.append((int(url2id(id)), id))
+            l.sort()
+            l.reverse()
+            tosave = [v[1] for v in l[0:19]]
+            toremove = set([v for v in self.storage.get('seen', default={})]).difference(tosave)
+            for id in toremove: 
+                self.storage.delete('seen', id)
+                
+        self.storage.save()
 

modules/ecrans/tools.py

@@ -33,7 +33,12 @@ def id2url(_id):
 
 def url2id(url):
     "return an id from an url"
-    return url
+    regexp =  re.compile("(^.*),([0-9]+)\.html$")
+    match = regexp.match(url)
+    if match: 
+        return match.group(2)
+    else:
+        raise ValueError("Can't find an id for the url")
 
 def rssid(entry):
     return url2id(entry.id)