[compcentral]

Question:

I don't understand why the tv library isn't as flexible as the other are.

Answer:

Music isn't a problem because music has a several defacto standards for attaching metadata directly to the media so we don't have to infer anything based off the path or filename.

Just to elaborate... Music files (like .mp3, .m4a, .wma) support having built in ID3 (or similar) tags that include information about the artist, song, album, etc. These files could be stored in random folders 100 levels deep and all would be just fine.

Video files do not have this capability. The end.

I would do as suggested and store your files using the normal, accepted folder
schema. I have close to 6TB of videos and I have no problems doing this at all. XBMC will handle the genre groupings for you.

[kraqh3d]

True, but you have multiple TV show sources (England vs America) which adds to it as you need to account for each of them, doubling the annoyance.

I wasn't 100% accurate earlier. The scraping limitation is really this:

<FOLDER WERE CONTENT IS SET>/<TV SHOW NAME>/<IGNORE ANY FOLDERS HERE>/media.s01e01.avi

It's just that setting content on the source applies to most users.

Now, if you select "Folder contains a single TV Show" then that folder name (where you set content with that option) is assumed to be the TV Show Name.

And all of this is actually kept in the video database, in the path table specifically. So, you could hack this all to work quite easily by modifying the path table accordingly. Set content on one of your Genre folders and then look at the path table in the video database and you'll see what I mean. You can duplicate that entry for all your other genre folders manually and rescan and it'll pick up all the other show folder. Just blank out the hash. (Actually duplicating it won't hurt as its very very unlikely you'll get a hash collision). This is just a quicker way to Set Content for all your genres than doing it thru the GUI.

[TheDoughboy]

compcentral -

I do see the point, but I'm looking at the best of both worlds. For WAF, I need to keep the files separate. Like I said when I was using other software it was because all the interface presented me with was a file list. I considered switching over to combing all the genre folders into one but again the WAF came into affect. My wife doesn't want the kids to see certain content. So I can either have everything in the proper folder like you suggest and then go and cherry pick which folders go into each source which is time consuming. Or I leave things as they are quick pick out the genre folder and stack them inside a source name TV, which took me five minutes. For example my horror movies are kept away from the library and I just browse them from the video file list. That way when we have movie night the kids never see the movies my wife doesn't want them to see. Yes I do know there are parental controls, but my wife doesn't even want the kids to see the movie posters and ask questions.

And like I said the solution for now is to go into my tv source and just run a scan on each genre folder and the shows scan correctly for the most part. It just that automated scans don't work. I have to manually do it.

The other solution would be to add the genre folders as sources and the automated scans work.

Just to elaborate... Music files (like .mp3, .m4a, .wma) support having built in ID3 (or similar) tags that include information about the artist, song, album, etc. These files could be stored in random folders 100 levels deep and all would be just fine.

I understand that, what I don't understand is what the purpose of nfo files is then? If I have a nfo file why doesn't it truly act like a ID3 tag? I do understand that the ID3 is directly inside the file, but then why not just tell the scraper to look for the nfo before looking at the video file?

I know that it says in the documentation that it will look at the nfo first, but I have had experiences where certain files it won't, ones that aren't in any database (tvdb, imdb.) For example I had a Max Headroom video that wasn't in the online databases, I created an empty nfo file and used Media Companion to make sure the nfo was correct and had a screenshot. XBMC initially scanned the file as a PBS Special, created the PBS Special show and added the video there. I removed the entry from the library went to Max Headroom and scanned for new content, this time it added the video correctly.
==========================
kraqh3d -

And all of this is actually kept in the video database, in the path table specifically. So, you could hack this all to work quite easily by modifying the path table accordingly. Set content on one of your Genre folders and then look at the path table in the video database and you'll see what I mean. You can duplicate that entry for all your other genre folders manually and rescan and it'll pick up all the other show folder. Just blank out the hash. (Actually duplicating it won't hurt as its very very unlikely you'll get a hash collision). This is just a quicker way to Set Content for all your genres than doing it thru the GUI.

I hope I'm understanding you correctly, but I think I did this already. I went to the TV source and made sure that the option for single show was not on, then I went to the genre and did the same thing. Then I went to each individual show and made the single show option was on. Still didn't work.

I'm still actively looking for solutions like this - http://forum.xbmc.org/showthread.php?tid=51614 and http://forum.xbmc.org/showthread.php?tid=68621 just to see what info I can gleam. I'll delve into code soon.

[kraqh3d]

No regexp today or custom scraper is going to help because the code makes assumptions about where to get the show name, and that's your problem. You MUST set content either one folder above the show names, or on the show name with "Single Show" enabled, but nowhere else, and not both.

In your case, you want to set content on the Genre folder. Instead of doing it through the GUI for every genre which can be time consuming, you can do this:

Close Xbmc and rename your video library database to keep a backup. Start Xbmc back up and enter video files mode. (Xbmc will create a new library database file.)

Enter one of your TV show sources (ie "T:\America, etc"). You should now see your list of genres. Navigate to the first one, bring up the context menu, and Set Content to TV Shows. Select a scraper because you have to. DO NOT select Single TV Show or anything else. Now close Xbmc down again.

Open the video database file in a SQLite browser. (There are tons of them. Use google.) Open the path table. You'll see a row for "T:\America, etc\GENRE" where strContent is "tv shows" and some other stuff. Copy that row as many times as you need, changing GENRE each time to match your genres. Start Xbmc back up, and scan. It should now scan them all.

Short of this, you have to change the source code in VideoInfoScanner.cpp. Trace thru it but I think you'll need to muck with RetrieveInfoForTVShow().

[TheDoughboy]

What you said worked, not automated scraping works fine. I kind of feel stupid for not thinking of that myself. Thanks for the help.

I'm looking at the VideoInfoScanner.cpp now to see what I can gleam. I know that the code is telling me it always puts a priority on the nfo file when scanning. The question I have to answer is why it can't look for that file when scanning. My thinking, and I could be wrong, is to tell it to only scrape when it finds that file. Tell the scraper that when it finds a tvshow.nfo file the folder contains a single show. That way you wouldn't have be worried about deepness in folders. Under my thinking the scraper would ignore folders that didn't have the tvshow.nfo.

I don't know how far I'll get but I'm going to give it a shot.