fix: properly handle missing title in cache

bin/background_indexer.py

@@ -54,8 +54,6 @@ class BackgroundIndexer(AbstractManager):
                 self.logger.debug(f'{uuid_path.parent} is locked, pickle generated by another process.')
                 continue
 
-            got_new_captures = True
-            max_captures -= 1
             with uuid_path.open() as f:
                 uuid = f.read()
 
@@ -80,6 +78,8 @@ class BackgroundIndexer(AbstractManager):
                 self.lookyloo.get_crawled_tree(uuid)
                 self.lookyloo.trigger_modules(uuid, auto_trigger=True)
                 self.logger.info(f'Pickle for {uuid} build.')
+                got_new_captures = True
+                max_captures -= 1
             except MissingUUID:
                 self.logger.warning(f'Unable to find {uuid}. That should not happen.')
             except NoValidHarFile as e:

lookyloo/capturecache.py

@@ -74,20 +74,13 @@ class CaptureCache():
 
         if cache_entry.get('title') is not None:
             self.title: str = cache_entry['title']
-        else:
-            # This shouldn't happen, but if it does, we need the key to exist.
-            self.logger.warning(f'Title missing in cache for {self.uuid}.')
-            self.title = ''
+
         if cache_entry.get('timestamp'):
             try:
                 self.timestamp: datetime = datetime.strptime(cache_entry['timestamp'], '%Y-%m-%dT%H:%M:%S.%f%z')
             except ValueError:
                 # If the microsecond is missing (0), it fails
                 self.timestamp = datetime.strptime(cache_entry['timestamp'], '%Y-%m-%dT%H:%M:%S%z')
-        else:
-            # This shouldn't happen, but if it does, we need the key to exist.
-            self.logger.warning(f'Timestamp missing in cache for {self.uuid}.')
-            self.timestamp = datetime.fromtimestamp(0)
 
         self.redirects: List[str] = json.loads(cache_entry['redirects']) if cache_entry.get('redirects') else []
 

website/web/__init__.py

@@ -774,7 +774,7 @@ def tree(tree_uuid: str, node_uuid: Optional[str]=None):
         return render_template('tree.html', tree_json=ct.to_json(),
                                info=cache,
                                tree_uuid=tree_uuid, public_domain=lookyloo.public_domain,
-                               screenshot_thumbnail=b64_thumbnail, page_title=cache.title,
+                               screenshot_thumbnail=b64_thumbnail, page_title=cache.title if hasattr(cache, 'title') else '',
                                screenshot_size=screenshot_size,
                                meta=meta, enable_mail_notification=enable_mail_notification,
                                enable_monitoring=lookyloo.monitoring_enabled,