[Cop E Ryte]

I'm hoping there's someone out there that can help me with an issue I recently discovered I'm having with XBMC. I have a .mkv file of the movie S.W.A.T. which is being read by XBMC as a movie titled "A Thousand Clouds of Peace." I checked the file and the way it was titled and all seems to be correct. I even renamed the file SWAT as well as S.W.A.T. [2003] and none of it seems to work. I'm using IMDB as the database to draw the movie info but this is bugging me out. I'd rather not own the movie S.W.A.T. (especially since it's not so great) if the software misreads it. Does XBMC have a problem reading acronyms or is there something else I'm missing. Please help if you can.

Thanks.

[althekiller]

What is the full file and/or directory name of the movie? If there happens to be an nfo file for that movie, make sure it has the correct imdb URL. A debug log of a scan attempt may be useful, pastebin it.

[TomJensen]

This happens to me on about 20% of the movies when I tried the built-in IDMB scraper. IMDB lists the most popular titles first, which may have the search string as part of the name, or as an alternate name. I'm guessing the built-in scraper simply takes the top entry, without going down to the "Exact Title" category and take the top entry from there.

There was no way to correct this, given that the virtual keyboard doesn't work (at least in the official Atlantis build), and it would've been a clunky option anyway, trying to enter characters with a remote. In the end, I ditched the built-in scraper and used the Media Companion (3rd-party scraper) instead. It's not perfect either, but the hit rate is pretty decent, and I can edit any mistakes with a keyboard.

Edit: Your problem is the [2003] suffix. I think some of the problems I had were the same type--I used (R5) and (screener) in some of the filenames.

[jmarshall]

Test cases where the scraper fails to find the appropriate movie yet it exists on an imdb search are always most welcome.

[theophile]

If you highlight the movie and select "refresh" from the context menu, it will ask you the title first. Just enter the correct title and off you go.

[digitalhigh]

If you highlight the movie and select "refresh" from the context menu, it will ask you the title first. Just enter the correct title and off you go.

I prefer to use one of the media downloader programs...either Movie Companion or Media Plus. Takes all the guesswork out of whether or not the scraper will get the right movie. And it's really not the scraper's fault either...any time you automate a large job like scanning a big movie library, there's bound to be some error somewhere along the way.

Just take the time to organize your library once with the proper posters/fanart/.nfo files, and you'll be a lot happier with the end results. I never had a perfect 1:1 DB scan until I manually organized everything...

[TomJensen]

@jmarshall

The IMDB problem I experienced (and also for the OP's case) is easily reproducible. When a movie title is short, say one or two words, then any extraneous info like (screener) or (R5) or (2003) suffixes will throw off the search [I'm assuming the scraper uses IMDB's search engine].

For example, "S.W.A.T." will return the correct entry at the top (at least if you want the 2003 title and not the earlier one), while "S.W.A.T. [2003]" won't even show the movies. Likewise, "Defiance" will pop up the 2008 movie at the top, while "Defiance (R5)" drops the 2008 entry down to 11th place. Similar results for Push, etc. In the OP's case, the erroneous movie selected was also made in 2003, which was part of the search term.

It's understandable. Given the title is short, any extraneous info will get a heavier weighing and confuse the search.

Suggest that the scraper filters out any trailing element enclosed in brackets or parens, or any trailing element that indicates a year, e.g. 19## or 20##.

[CrashX]

In some case the year becomes usefull ... ie I am legend was released twice.

[jmarshall]

All that stuff is cleaned off the filename before we send it to imdb.