From d39dacb5929d94a593bd4e01271e0b6ff0e5b79c Mon Sep 17 00:00:00 2001
From: Samuel Clay <samuel@ofbrooklyn.com>
Date: Thu, 20 May 2010 15:13:25 -0400
Subject: [PATCH] Cleanup in prep for move from urllib to eventlets.

---
 apps/rss_feeds/models.py | 36 ++++++++++++++++++------------------
 utils/feed_fetcher.py    |  9 ++-------
 2 files changed, 20 insertions(+), 25 deletions(-)

diff --git a/apps/rss_feeds/models.py b/apps/rss_feeds/models.py
index 3249bbc70..3bb348845 100644
--- a/apps/rss_feeds/models.py
+++ b/apps/rss_feeds/models.py
@@ -1,25 +1,17 @@
+import time
+import settings
+import difflib
+import datetime
+import hashlib
 from django.db import models
 from django.db import IntegrityError
-from django.contrib.auth.models import User
-from django.contrib.contenttypes.models import ContentType
-from django.core import serializers
 from django.core.cache import cache
-from utils import feedparser, object_manager, json
-from utils.dateutil.parser import parse as dateutil_parse
-from utils.feed_functions import encode, prints, mtime, levenshtein_distance
-import time, datetime, random
-from django.utils.http import urlquote
-from django.utils.safestring import mark_safe
+from utils import json
+from utils.feed_functions import levenshtein_distance
 from utils.story_functions import format_story_link_date__short
 from utils.story_functions import format_story_link_date__long
 from utils.story_functions import pre_process_story
 from utils.compressed_textfield import StoryField
-from django.db.models import Q
-import settings
-import logging
-import difflib
-import datetime
-import hashlib
 from utils.diff import HTMLDiff
 
 USER_AGENT = 'Protopub v1.0 - protopub.com'
@@ -102,7 +94,7 @@ class Feed(models.Model):
                 existing_story, story_has_changed = self._exists_story(story, story_content, existing_stories)
                 story_author, _ = self._save_story_author(story.get('author'))
                 if existing_story is None:
-                    pub_date = datetime.datetime.timetuple(story.get('published'))
+                    # pub_date = datetime.datetime.timetuple(story.get('published'))
                     # logging.debug('- New story: %s %s' % (pub_date, story.get('title')))
                     
                     s = Story(story_feed = self,
@@ -332,7 +324,15 @@ class StoryAuthor(models.Model):
         
     def __unicode__(self):
         return '%s - %s' % (self.feed, self.author_name)
-        
+
+class FeedPage(models.Model):
+    feed = models.OneToOneField(Feed, related_name="feed_page")
+    page_data = StoryField(null=True, blank=True)
+
+class FeedXML(models.Model):
+    feed = models.OneToOneField(Feed, related_name="feed_xml")
+    rss_xml = StoryField(null=True, blank=True)
+    
 class Story(models.Model):
     '''A feed item'''
     story_feed = models.ForeignKey(Feed, related_name="stories")
@@ -372,7 +372,7 @@ class FeedUpdateHistory(models.Model):
     
     def __unicode__(self):
         return "[%s] %s feeds: %s seconds" % (
-            fetch_date.strftime('%F %d'),
+            self.fetch_date.strftime('%F %d'),
             self.number_of_feeds,
             self.seconds_taken,
         )
diff --git a/utils/feed_fetcher.py b/utils/feed_fetcher.py
index 3c1449c2a..a44789f08 100644
--- a/utils/feed_fetcher.py
+++ b/utils/feed_fetcher.py
@@ -1,21 +1,17 @@
 from apps.rss_feeds.models import Feed, Story, FeedUpdateHistory
 from django.core.cache import cache
-from apps.reader.models import UserSubscription, UserStory
+from apps.reader.models import UserSubscription
 from apps.rss_feeds.importer import PageImporter
 from utils import feedparser
 from django.db import transaction
 from django.db.models import Q
-from utils.dateutil.parser import parse as dateutil_parse
 from utils.story_functions import pre_process_story
 import sys
 import time
 import logging
 import datetime
-# import threading
 import traceback
 import multiprocessing
-import Queue
-import datetime
 import random
 import socket
 
@@ -69,7 +65,7 @@ class FetchFeed:
             print(log_msg)
             return FEED_SAME, None
             
-        next_scheduled_update = self.set_next_scheduled_update()
+        self.set_next_scheduled_update()
         
         # we check the etag and the modified time to save bandwith and avoid bans
         try:
@@ -205,7 +201,6 @@ class ProcessFeed:
 
 
         # Compare new stories to existing stories, adding and updating
-        num_entries = len(self.fpf.entries)
         start_date = datetime.datetime.now()
         end_date = datetime.datetime.now()
         story_guids = []