From 4a294764152c3878927095f5c7efe955489ec7f9 Mon Sep 17 00:00:00 2001
From: echel0n <justin.tabish@gmail.com>
Date: Mon, 21 Jul 2014 15:16:04 -0700
Subject: [PATCH] Added code to perform disposal of bs4 class when finished as
 well as perform garbage collection afterwards, this fixed a memory leak issue
 SR was experiancing during backlog/manual/failed searches.

---
 sickbeard/providers/bitsoup.py      |  7 ++++++-
 sickbeard/providers/freshontv.py    |  6 ++++++
 sickbeard/providers/hdtorrents.py   |  5 +++++
 sickbeard/providers/iptorrents.py   |  5 +++++
 sickbeard/providers/kat.py          | 20 +++++++++++++------
 sickbeard/providers/publichd.py     |  9 +++++++--
 sickbeard/providers/scc.py          |  9 +++++++++
 sickbeard/providers/torrentbytes.py |  5 +++++
 sickbeard/providers/torrentleech.py |  5 +++++
 sickbeard/scene_numbering.py        |  1 -
 sickbeard/tv.py                     | 31 +++++++++++++----------------
 tests/torrent_tests.py              |  9 +++++++--
 12 files changed, 83 insertions(+), 29 deletions(-)

diff --git a/sickbeard/providers/bitsoup.py b/sickbeard/providers/bitsoup.py
index d66facd0..6e19257e 100644
--- a/sickbeard/providers/bitsoup.py
+++ b/sickbeard/providers/bitsoup.py
@@ -21,6 +21,7 @@ import traceback
 import datetime
 import urlparse
 import time
+import gc
 import sickbeard
 import generic
 from sickbeard.common import Quality, cpu_presets
@@ -174,7 +175,11 @@ class BitSoupProvider(generic.TorrentProvider):
                     
                     torrent_table = html.find('table', attrs={'class': 'koptekst'})
                     torrent_rows = torrent_table.find_all('tr') if torrent_table else []
-                    
+
+                    # cleanup memory
+                    html.decompose()
+                    gc.collect()
+
                     #Continue only if one Release is found
                     if len(torrent_rows) < 2:
                          logger.log(u"The Data returned from " + self.name + " do not contains any torrent",
diff --git a/sickbeard/providers/freshontv.py b/sickbeard/providers/freshontv.py
index b52b8f92..e034c95b 100755
--- a/sickbeard/providers/freshontv.py
+++ b/sickbeard/providers/freshontv.py
@@ -21,6 +21,7 @@ import traceback
 import datetime
 import urlparse
 import time
+import gc
 import sickbeard
 import generic
 from sickbeard.common import Quality, cpu_presets
@@ -199,6 +200,11 @@ class FreshOnTVProvider(generic.TorrentProvider):
 
                     torrent_table = html.find('table', attrs={'class': 'frame'})
                     torrent_rows = torrent_table.findChildren('tr') if torrent_table else []
+
+                    # cleanup memory
+                    html.decompose()
+                    gc.collect()
+
                     #Continue only if one Release is found                    
                     if len(torrent_rows) < 2:
                         logger.log(u"The Data returned from " + self.name + " do not contains any torrent",
diff --git a/sickbeard/providers/hdtorrents.py b/sickbeard/providers/hdtorrents.py
index f176e542..79bb8374 100644
--- a/sickbeard/providers/hdtorrents.py
+++ b/sickbeard/providers/hdtorrents.py
@@ -22,6 +22,7 @@ import re
 import traceback
 import datetime
 import urlparse
+import gc
 import sickbeard
 import generic
 from sickbeard.common import Quality, cpu_presets
@@ -203,6 +204,10 @@ class HDTorrentsProvider(generic.TorrentProvider):
                     #Get first entry in table
                     entries = html.find_all('td', attrs={'align': 'center'})
 
+                    # cleanup memory
+                    html.decompose()
+                    gc.collect()
+
                     if not entries:
                         logger.log(u"The Data returned from " + self.name + " do not contains any torrent",
                                    logger.DEBUG)
diff --git a/sickbeard/providers/iptorrents.py b/sickbeard/providers/iptorrents.py
index 3e2caf15..9dccd217 100644
--- a/sickbeard/providers/iptorrents.py
+++ b/sickbeard/providers/iptorrents.py
@@ -21,6 +21,7 @@ import re
 import traceback
 import datetime
 import urlparse
+import gc
 import sickbeard
 import generic
 from sickbeard.common import Quality, cpu_presets
@@ -182,6 +183,10 @@ class IPTorrentsProvider(generic.TorrentProvider):
                     torrent_table = html.find('table', attrs={'class': 'torrents'})
                     torrents = torrent_table.find_all('tr') if torrent_table else []
 
+                    # cleanup memory
+                    html.decompose()
+                    gc.collect()
+
                     #Continue only if one Release is found                    
                     if len(torrents) < 2:
                         logger.log(u"The Data returned from " + self.name + " do not contains any torrent",
diff --git a/sickbeard/providers/kat.py b/sickbeard/providers/kat.py
index 2ecf7c34..50c7b6b1 100644
--- a/sickbeard/providers/kat.py
+++ b/sickbeard/providers/kat.py
@@ -19,7 +19,7 @@
 
 from __future__ import with_statement
 
-import time
+import gc
 import sys
 import os
 import traceback
@@ -120,8 +120,12 @@ class KATProvider(generic.TorrentProvider):
             return None
 
         try:
-            soup = BeautifulSoup(data, features=["html5lib", "permissive"])
-            file_table = soup.find('table', attrs={'class': 'torrentFileList'})
+            html = BeautifulSoup(data, features=["html5lib", "permissive"])
+            file_table = html.find('table', attrs={'class': 'torrentFileList'})
+
+            # cleanup memory
+            html.decompose()
+            gc.collect()
 
             if not file_table:
                 return None
@@ -248,11 +252,15 @@ class KATProvider(generic.TorrentProvider):
                     continue
 
                 try:
-                    soup = BeautifulSoup(html, features=["html5lib", "permissive"])
+                    html = BeautifulSoup(html, features=["html5lib", "permissive"])
 
-                    torrent_table = soup.find('table', attrs={'class': 'data'})
+                    torrent_table = html.find('table', attrs={'class': 'data'})
                     torrent_rows = torrent_table.find_all('tr') if torrent_table else []
 
+                    # cleanup memory
+                    html.decompose()
+                    gc.collect()
+
                     #Continue only if one Release is found
                     if len(torrent_rows) < 2:
                         logger.log(u"The data returned from " + self.name + " does not contain any torrents",
@@ -284,7 +292,7 @@ class KATProvider(generic.TorrentProvider):
                                 logger.DEBUG)
                             continue
 
-                        #Check number video files = episode in season and find the real Quality for full season torrent analyzing files in torrent 
+                        #Check number video files = episode in season and find the real Quality for full season torrent analyzing files in torrent
                         if mode == 'Season' and search_mode == 'sponly':
                             ep_number = int(epcount / len(set(allPossibleShowNames(self.show))))
                             title = self._find_season_quality(title, link, ep_number)
diff --git a/sickbeard/providers/publichd.py b/sickbeard/providers/publichd.py
index ff365d6c..fddf0fe4 100644
--- a/sickbeard/providers/publichd.py
+++ b/sickbeard/providers/publichd.py
@@ -25,6 +25,7 @@ import traceback
 import urllib, urlparse
 import re
 import datetime
+import gc
 
 import sickbeard
 import generic
@@ -152,11 +153,15 @@ class PublicHDProvider(generic.TorrentProvider):
                 html = os.linesep.join([s for s in html.splitlines() if not optreg.search(s)])
 
                 try:
-                    soup = BeautifulSoup(html, features=["html5lib", "permissive"])
+                    html = BeautifulSoup(html, features=["html5lib", "permissive"])
 
-                    torrent_table = soup.find('table', attrs={'id': 'torrbg'})
+                    torrent_table = html.find('table', attrs={'id': 'torrbg'})
                     torrent_rows = torrent_table.find_all('tr') if torrent_table else []
 
+                    # cleanup memory
+                    html.decompose()
+                    gc.collect()
+
                     #Continue only if one Release is found
                     if len(torrent_rows) < 2:
                         logger.log(u"The Data returned from " + self.name + " do not contains any torrent",
diff --git a/sickbeard/providers/scc.py b/sickbeard/providers/scc.py
index 2f908245..1d5d3753 100644
--- a/sickbeard/providers/scc.py
+++ b/sickbeard/providers/scc.py
@@ -22,6 +22,7 @@ import re
 import traceback
 import datetime
 import urlparse
+import gc
 import sickbeard
 import generic
 from sickbeard.common import Quality, cpu_presets
@@ -203,6 +204,10 @@ class SCCProvider(generic.TorrentProvider):
                         torrent_table = html.find('table', attrs={'id': 'torrents-table'})
                         torrent_rows = torrent_table.find_all('tr') if torrent_table else []
 
+                        # cleanup memory
+                        html.decompose()
+                        gc.collect()
+
                         #Continue only if at least one Release is found
                         if len(torrent_rows) < 2:
                             if html.title:
@@ -222,10 +227,14 @@ class SCCProvider(generic.TorrentProvider):
                                     url = all_urls[1]
                                 else:
                                     url = all_urls[0]
+
                                 title = link.string
                                 if re.search('\.\.\.', title):
                                     details_html = BeautifulSoup(self.getURL(self.url + "/" + link['href']))
                                     title = re.search('(?<=").+(?<!")', details_html.title.string).group(0)
+                                    details_html.decompose()
+                                    gc.collect()
+
                                 download_url = self.urls['download'] % url['href']
                                 id = int(link['href'].replace('details?id=', ''))
                                 seeders = int(result.find('td', attrs={'class': 'ttr_seeders'}).string)
diff --git a/sickbeard/providers/torrentbytes.py b/sickbeard/providers/torrentbytes.py
index 979d6e96..c75d0e87 100644
--- a/sickbeard/providers/torrentbytes.py
+++ b/sickbeard/providers/torrentbytes.py
@@ -21,6 +21,7 @@ import traceback
 import datetime
 import urlparse
 import time
+import gc
 import sickbeard
 import generic
 from sickbeard.common import Quality, cpu_presets
@@ -175,6 +176,10 @@ class TorrentBytesProvider(generic.TorrentProvider):
                     torrent_table = html.find('table', attrs={'border': '1'})
                     torrent_rows = torrent_table.find_all('tr') if torrent_table else []
 
+                    # cleanup memory
+                    html.decompose()
+                    gc.collect()
+
                     #Continue only if one Release is found                    
                     if len(torrent_rows) < 2:
                         logger.log(u"The Data returned from " + self.name + " do not contains any torrent",
diff --git a/sickbeard/providers/torrentleech.py b/sickbeard/providers/torrentleech.py
index f323b81a..5aa98cca 100644
--- a/sickbeard/providers/torrentleech.py
+++ b/sickbeard/providers/torrentleech.py
@@ -21,6 +21,7 @@ import traceback
 import datetime
 import urlparse
 import time
+import gc
 import sickbeard
 import generic
 from sickbeard.common import Quality, cpu_presets
@@ -178,6 +179,10 @@ class TorrentLeechProvider(generic.TorrentProvider):
                     torrent_table = html.find('table', attrs={'id': 'torrenttable'})
                     torrent_rows = torrent_table.find_all('tr') if torrent_table else []
 
+                    # cleanup memory
+                    html.decompose()
+                    gc.collect()
+
                     #Continue only if one Release is found                    
                     if len(torrent_rows) < 2:
                         logger.log(u"The Data returned from " + self.name + " do not contains any torrent",
diff --git a/sickbeard/scene_numbering.py b/sickbeard/scene_numbering.py
index e82ad1b3..7e2be838 100644
--- a/sickbeard/scene_numbering.py
+++ b/sickbeard/scene_numbering.py
@@ -441,7 +441,6 @@ def get_xem_absolute_numbering_for_show(indexer_id, indexer):
 
     xem_refresh(indexer_id, indexer)
 
-
     result = {}
     myDB = db.DBConnection()
     rows = myDB.select(
diff --git a/sickbeard/tv.py b/sickbeard/tv.py
index a5013928..1176201a 100644
--- a/sickbeard/tv.py
+++ b/sickbeard/tv.py
@@ -227,11 +227,6 @@ class TVShow(object):
 
     def getEpisode(self, season=None, episode=None, file=None, noCreate=False, absolute_number=None, forceUpdate=False):
 
-        # Load XEM data to DB for show
-        sickbeard.scene_numbering.xem_refresh(self.indexerid, self.indexer, force=forceUpdate)
-
-        ep = None
-
         # if we get an anime get the real season and episode
         if self.is_anime and absolute_number and not season and not episode:
             myDB = db.DBConnection()
@@ -269,21 +264,23 @@ class TVShow(object):
             else:
                 ep = TVEpisode(self, season, episode)
 
-            # get scene absolute numbering
-            ep.scene_absolute_number = sickbeard.scene_numbering.get_scene_absolute_numbering(self.indexerid,
-                                                                                              self.indexer,
-                                                                                              ep.absolute_number)
-
-            # get scene season and episode numbering
-            ep.scene_season, ep.scene_episode = sickbeard.scene_numbering.get_scene_numbering(self.indexerid,
-                                                                                              self.indexer,
-                                                                                              season, episode)
-
             if ep != None:
+                # Load XEM data to DB for show
+                sickbeard.scene_numbering.xem_refresh(self.indexerid, self.indexer, force=forceUpdate)
+
+                # get scene absolute numbering
+                ep.scene_absolute_number = sickbeard.scene_numbering.get_scene_absolute_numbering(self.indexerid,
+                                                                                                  self.indexer,
+                                                                                                  ep.absolute_number)
+
+                # get scene season and episode numbering
+                ep.scene_season, ep.scene_episode = sickbeard.scene_numbering.get_scene_numbering(self.indexerid,
+                                                                                                  self.indexer,
+                                                                                                  season, episode)
+
                 self.episodes[season][episode] = ep
 
-        epObj = self.episodes[season][episode]
-        return epObj
+        return self.episodes[season][episode]
 
     def should_update(self, update_date=datetime.date.today()):
 
diff --git a/tests/torrent_tests.py b/tests/torrent_tests.py
index 677af49d..11bf4e13 100644
--- a/tests/torrent_tests.py
+++ b/tests/torrent_tests.py
@@ -22,6 +22,7 @@ from __future__ import with_statement
 import unittest
 import sys, os.path
 import urlparse
+import gc
 
 sys.path.append(os.path.abspath('..'))
 sys.path.append(os.path.abspath('../lib'))
@@ -40,11 +41,15 @@ class TorrentBasicTests(test.SickbeardTestDBCase):
         if not html:
             return
 
-        soup = BeautifulSoup(html, features=["html5lib", "permissive"])
+        html = BeautifulSoup(html, features=["html5lib", "permissive"])
 
-        torrent_table = soup.find('table', attrs={'class': 'data'})
+        torrent_table = html.find('table', attrs={'class': 'data'})
         torrent_rows = torrent_table.find_all('tr') if torrent_table else []
 
+        # cleanup memory
+        html.decompose()
+        gc.collect()
+
         #Continue only if one Release is found
         if len(torrent_rows) < 2:
             print(u"The data returned does not contain any torrents")