Page 39 of 55

Re: Media Rover

Posted: May 21st, 2010, 1:08 pm
by kierse
wally007 wrote: What do you think about making it official , kierse ? Sources need updating.
I created a task in the Github issue tracker to create a new source plugin for nzbsRus
wiggy9906 wrote: www.nzbindex.nl also has a working rss feed. Build the feed from there their website and then add &nzblink=1 to the URL. Again set provider as mytvnzb.
While it would definitely be possible to add support (it has an RSS feed), the report titles are no where near as clean as what you get from any of the other indexing sites.  MR has to be able to parse the report title and identify the series, season, episode number, etc in order to successfully compare it to what's on your filesystem.  I'm not very optimistic that we would get very reliable results from nzbindex.nl but I could be wrong.  How has your experience been using it?  Does it work properly or does MR end up scheduling a bunch of random stuff?

Kierse

Re: Media Rover

Posted: May 21st, 2010, 2:33 pm
by rollingeyeball
kierse wrote: Hey rollingeyeball, good to have you back!  Are you running the latest code from the dev branch?  Also, did you replace the MR application directory when you upgraded or did you just unpack it over top of the existing one?

Kierse
I created a brand new directory for the downloaded files (/dev branch) and wrote new config files to the default location, which is the same location as last time.

Can I suggest you try NZBclub instead of nzbindex? I might not even use it myself, but NZBClub gives the best results, hands down.
I think as long as people customise their searches adequately, it'd be possible to get results.

Re: Media Rover

Posted: May 21st, 2010, 3:16 pm
by wiggy9906
kierse wrote:
wiggy9906 wrote: www.nzbindex.nl also has a working rss feed. Build the feed from there their website and then add &nzblink=1 to the URL. Again set provider as mytvnzb.
While it would definitely be possible to add support (it has an RSS feed), the report titles are no where near as clean as what you get from any of the other indexing sites.  MR has to be able to parse the report title and identify the series, season, episode number, etc in order to successfully compare it to what's on your filesystem.  I'm not very optimistic that we would get very reliable results from nzbindex.nl but I could be wrong.  How has your experience been using it?  Does it work properly or does MR end up scheduling a bunch of random stuff?

Kierse
I've been using rss feed from nzbindex.nl since Monday (5 days). No false positives. I can't say how many matches were made as I have 3 other feeds configured and not checked what sources are from where. When I set it up I configured it as the only feed and removed a few of my recent downloaded episodes. It found then and downloaded them fine.

I guess it helps having the correct search set up.  I search for Xvid in groups a.b.boneless a.b.multimedia a.b.tv and a.b.tvseries.

Wiggy

sqlite3.OperationalError: database is locked

Posted: May 22nd, 2010, 9:17 am
by paulkoan
Hello, I tried this with dev release 176 and 194, and am getting the following traceback.

I tried deleting the metadata.db  and got the same error, and there isn't another instance of mediarover running.  I am not using sorting.

Any ideas what can be holding the db open?

Paul

Code: Select all

Traceback (most recent call last):
  File "/home/share/apps/tv/mediarover/mediarover/mediarover.py", line 19, in <module>
    mediarover.run()
  File "/home/share/apps/tv/mediarover/kierse-mediarover-c434494/mediarover/__init__.py", line 101, in run
    scheduler(broker, options)
  File "/home/share/apps/tv/mediarover/kierse-mediarover-c434494/mediarover/__init__.py", line 133, in scheduler
    __scheduler(broker, options)
  File "/home/share/apps/tv/mediarover/kierse-mediarover-c434494/mediarover/__init__.py", line 348, in __scheduler
    in_progress = set([row['title'] for row in broker['metadata_data_store'].list_in_progress()])
  File "/home/share/apps/tv/mediarover/kierse-mediarover-c434494/mediarover/ds/metadata.py", line 57, in list_in_progress
    return self.__dbh.execute("SELECT title, type, quality FROM in_progress").fetchall()
sqlite3.OperationalError: database is locked

Re: Media Rover

Posted: May 22nd, 2010, 8:33 pm
by pilGrim
wiggy9906 wrote: www.nzbindex.nl also has a working rss feed. Build the feed from there their website and then add &nzblink=1 to the URL. Again set provider as mytvnzb.

I've been using rss feed from nzbindex.nl since Monday (5 days). No false positives. I can't say how many matches were made as I have 3 other feeds configured and not checked what sources are from where. When I set it up I configured it as the only feed and removed a few of my recent downloaded episodes. It found then and downloaded them fine.

I guess it helps having the correct search set up.  I search for Xvid in groups a.b.boneless a.b.multimedia a.b.tv and a.b.tvseries.

Wiggy
I set up a NZBIndex.nl feed as suggested above, I remed out the other feeds so this was my only feed.  When running in my debugger I can see in the output that not one show on my hard drive was matched; all valid shows that should match in the rss show skipping, not watching series.  It seems that media rover is not parsing the name out of the title at all for me.  I did add the &nzblink=1 to the end of the rss feed and do have good results showing the the rss due to my search parameters.

Are my results unique, meaning some settings are wrong, or is this feed not workable?

pilgrim

Re: sqlite3.OperationalError: database is locked

Posted: May 22nd, 2010, 10:35 pm
by paulkoan
paulkoan wrote:
I tried deleting the metadata.db  and got the same error, and there isn't another instance of mediarover running.  I am not using sorting.

Any ideas what can be holding the db open?
It was on an nfs share - only one process openining it as far as I can tell, so I guess this is just a nfs locking issue as usual.  I moved the db out of the share and now it works.

nfs - both excellent and crap simultaneously.

Re: Media Rover

Posted: May 23rd, 2010, 2:08 am
by kierse
Hi wiggy9906,
wiggy9906 wrote: There is no comma in file, I've tried season and Season, the docs indicate the season is ignored in any case and only the numerical value read. File is in series root.
There was indeed a problem with the code handling the .ignore files.  I just pushed a patch to Github.  Update your code and try again

Kierse

Re: Media Rover

Posted: May 23rd, 2010, 2:22 am
by kierse
rollingeyeball wrote: I created a brand new directory for the downloaded files (/dev branch) and wrote new config files to the default location, which is the same location as last time.
I had a look at the code and it is most definitely looking for the 'single_episode' template variable.  My production machine (which is running the latest code from dev) works as expected (using 'single_episode') and I haven't been able to reproduce this problem.  What version of code from the dev branch are you running? 
rollingeyeball wrote: Can I suggest you try NZBclub instead of nzbindex? I might not even use it myself, but NZBClub gives the best results, hands down.
I think as long as people customise their searches adequately, it'd be possible to get results.
I updated the request ticket for nzbsRus to include NZBclub and nzbindex.  I'll create them both and let the users decide which one works best

Kierse

Re: sqlite3.OperationalError: database is locked

Posted: May 23rd, 2010, 2:27 am
by kierse
Hey paulkoan,
paulkoan wrote: I am not using sorting.
Thanks for bringing this one to my attention.  When quality management is turned off MR doesn't need to connect to the database and therefore shouldn't be doing so.  I'll update the code to reflect this.

Kierse

p.s. In case you have quality management turned on: MR updates the quality tables in the database when sorting a downloaded episode.  If you aren't using the sorting script there isn't any reason to have quality management turned on.

Re: Media Rover

Posted: May 23rd, 2010, 2:49 am
by rollingeyeball
Ahh. I found my issue. Since I made a new folder, and didn't overwrite the last MR folder my sorting batch needed to be edited. The batch was calling the old sorting.

/facepalm

But thanks for letting me know it was me :)

Re: Media Rover

Posted: May 23rd, 2010, 4:24 am
by wiggy9906
kierse wrote: Hi wiggy9906,
wiggy9906 wrote: There is no comma in file, I've tried season and Season, the docs indicate the season is ignored in any case and only the numerical value read. File is in series root.
There was indeed a problem with the code handling the .ignore files.  I just pushed a patch to Github.  Update your code and try again

Kierse
Thanks Kierse, I updated and the .ignore files are now working as expected.

Thank you for your excellent program that remains working while all others fail.

Wiggy

Re: Media Rover

Posted: May 23rd, 2010, 4:39 am
by paulkoan
Hello Kierse - is there any chance that "trunk" could be accepted as a version for sabnzbd or that we could turn off version checking in the conf?

I update quite regularly and have to edit out the version check each time. 

Cheers,

Paul

Re: Media Rover

Posted: May 23rd, 2010, 3:11 pm
by kierse
Hey paulkoan,
paulkoan wrote: Hello Kierse - is there any chance that "trunk" could be accepted as a version for sabnzbd or that we could turn off version checking in the conf?

I update quite regularly and have to edit out the version check each time. 
Done, MR will now look for a flag in the sabnzbd queue section called '__check_version__' before performing the version check.  Adding this variable and setting it to False will cause MR to skip the version check.

Kierse

Re: Media Rover

Posted: May 23rd, 2010, 6:56 pm
by wally007
MR is crashing due to error in RSS feed ( nzb matrix feeds are offline at the moment ).

EDIT -> Actually nzbmatrix MOVED their feeds and in place of old feeds is redirect. Still problem remains , it'd better if MR just ignored ONE source error and continued with other sources.

Here's the log

Code: Select all

2010-05-24 01:45 INFO - processing 'nzbmatrix SD' items - __init__.py:374
2010-05-24 01:45 ERROR - list index out of range - __init__.py:133
Traceback (most recent call last):
  File "/Users/vladik1/Downloads/mediarover/mediarover/__init__.py", line 131, in scheduler
    __scheduler(broker, options)
  File "/Users/vladik1/Downloads/mediarover/mediarover/__init__.py", line 375, in __scheduler
    for item in source.items():
  File "/Users/vladik1/Downloads/mediarover/mediarover/source/nzbmatrix/__init__.py", line 40, in items
    item = NzbmatrixItem(rawItem, self.type(), self.priority(), self.quality())
  File "/Users/vladik1/Downloads/mediarover/mediarover/source/nzbmatrix/item.py", line 90, in __init__
    self.__download = self.__parseItem()
  File "/Users/vladik1/Downloads/mediarover/mediarover/source/nzbmatrix/item.py", line 63, in __parseItem
    if re.match("TV", self._report_category()):
  File "/Users/vladik1/Downloads/mediarover/mediarover/source/nzbmatrix/item.py", line 77, in _report_category
    return re.match("(\w+):", self.__item.getElementsByTagName("category")[0].childNodes[0].data).group(1)
IndexError: list index out of range
Traceback (most recent call last):
  File "/Users/vladik1/Downloads/mediarover/mediarover.py", line 19, in <module>
    mediarover.run()
  File "/Users/vladik1/Downloads/mediarover/mediarover/__init__.py", line 99, in run
    scheduler(broker, options)
  File "/Users/vladik1/Downloads/mediarover/mediarover/__init__.py", line 131, in scheduler
    __scheduler(broker, options)
  File "/Users/vladik1/Downloads/mediarover/mediarover/__init__.py", line 375, in __scheduler
    for item in source.items():
  File "/Users/vladik1/Downloads/mediarover/mediarover/source/nzbmatrix/__init__.py", line 40, in items
    item = NzbmatrixItem(rawItem, self.type(), self.priority(), self.quality())
  File "/Users/vladik1/Downloads/mediarover/mediarover/source/nzbmatrix/item.py", line 90, in __init__
    self.__download = self.__parseItem()
  File "/Users/vladik1/Downloads/mediarover/mediarover/source/nzbmatrix/item.py", line 63, in __parseItem
    if re.match("TV", self._report_category()):
  File "/Users/vladik1/Downloads/mediarover/mediarover/source/nzbmatrix/item.py", line 77, in _report_category
    return re.match("(\w+):", self.__item.getElementsByTagName("category")[0].childNodes[0].data).group(1)
IndexError: list index out of range

Re: Media Rover

Posted: May 23rd, 2010, 7:13 pm
by kierse
wally007 wrote: ...it'd better if MR just ignored ONE source error and continued with other sources.
It's supposed to be doing that.  I'll have a look and get back to you