Group matching RSS results

[arucard]

I would like to request a feature for the RSS feeds. I would like to be able to group together all results that match the same filter on an RSS feed. This would only have to be grouped for matches in the current, latest rss feed. These grouped results would then be sent to the queue as a single download.

This would be very useful for automatically downloading things that are spread across multiple posts, with rar files separated from par files for example. This could then fail post-processing (if one of the rar files needed repair) even though the par files are already in the queue. Additionally, this feature could prevent your queue and history from cluttering up, because you would only have a single entry for every thing you download.

So the feature would work roughly like this. You'd have an additional switch option available, labeled as something like "Group matching RSS results". Now let's say your RSS feed has a filter named "test". When the latest RSS feed is retrieved, let's say it contains the following posts:
- test.rar (generally containing all rar files and a single par2 file)
- test.par2 (generally containing all par2 files)
- test.nfo (generally containing an nfo file and an srr file)
Now, when this is matched against the filters, you would normally get 3 separate entries in your queue, one for each RSS entry. But since all these RSS entries would have been matched against the "test" filter, they would be grouped together as a single entry, named after the first matched RSS entry (or something else). You then have a single entry in your queue, and later in your history, representing this download. You can also perform all post-processing correctly and run post-processing scripts on all available data at once.

Note that the next time the RSS feed is retrieved and checked, all information about previously matching entries can be forgotten.

Of course, there is still the matter of multiple filters matching (like with "Requires" filters). Here you could choose to either match the entire set of matched filters or, probably the more simple solution, just to check the last matched filter and group on that.

I hope this feature makes it into this already excellent program. It would really clean up my queue and history and, more importantly, make my post-processing scripts work correctly when the data is spread over multiple posts.

[shypike]

So you're asking for SABnzbd to compensate for a low quality indexer?
One that doesn't group together what goes together.
Do you setup filters for each and every show separately?
I'm not convinced that this would a good investment, as it would take a lot
of implementation time for a small audience and make things rather complicated to operate.

[arucard]

I don't know how much this would take to implement, it seemed like something that could just work together with the checking of the filters against the RSS feed. But I haven't seen the source code, so if you say it's too complicated I'll have to take your word for it.

As for the indexer, I don't know if I'm using a low quality indexer. I'm using binsearch which is shown on the SABnzbd wiki and it works fine for me for the most part. Since I prefer not to register or use a paid indexing site, the only other site that's suggested on the wiki seems to be nzbindex which does seem to group posts together a bit better, but the wiki said that the RSS always seems like new nzb's so I didn't use that one.

I'm not sure how this would make things more complicated to operate when it just adds a single checkbox to change the behaviour from its current state, but perhaps I'm understanding you incorrectly. It also seems to me that whatever filter you've used on your RSS feed, it might be useful to have the option to force everything that matches that filter to be downloaded as a single nzb. I just assumed that it wouldn't be too hard to create a new nzb from multiple existing nzb's (which would then be sent to the queue), but i could be wrong about that.

Either way, if the cost really outweighs the benefit, then that's just too bad. But I do hope it can find its way into a future release. Until then, I'll just check out some different indexing sites, perhaps that'll help too.