Trailer naming conventions in XBMC - a discovery that I thought I would share

Hello again everyone,

This may or may not exist somewhere in some post, but I never saw it so I thought it might be useful to others. Ever since I first installed XBMC and then Aeon, I have been struggling with why one particular movie would just never get scraped properly (or even pickup and hand-made nfo file). The name of the movie is called "The Long, Long, Trailer" a funny movie with Lucille Ball and Desi Arnaz. After going through weeks of machinations, removing parts of the file name like the comma or the extra "long" from the name, it turns out that the problem was having the word "Trailer" in the name. It occurred to me just this morning that it may have something to with "Trailer" causing XBMC to assume it was a movie trailer and therefore would not pick up the nfo or do a proper scrape. So I removed the word "Trailer" from the name of the file and lo and behold, it scraped like a champ! So if you have any movies with that word in the title (there are a few such as Trailer Park Boys, Mickey's Trailer, Trailer Park Of Terror, Trailer Horn, etc) then you may need to remove that word from the name before you try to scrape it.

I just thought I would share my discovery and save anyone else going through the hair-pulling experience that I did.

Interesting observation, but this has to do with XBMC and not Aeon. You should post this over in the general area so the developers can see if something can be done to fix the issue.

What is the exact filename?

Cheers,
Jonathan

We should probably tighten up the trailer detection. Forcing moviename-trailer.ext isn't all that restrictive. Right now we match on "[-._ ](sample|trailer)[-._ ]".

I was bored so I tested this out...As a rule of thumb I lookup the movie in IMDB then name the folder to what IMDB goes by. So for this movie I used "The Long, Long Trailer (1953)"....Ember picked it up correctly. Unless I overlooked it (I deleted the created .nfo before using XBMC) XBMC missed it. I didn't even see a wierd new movie at all, but I was looking through 300 movies so I could have passed it.

**EDIT**

Even using the Ember .nfo XBMC still didn't scrape into the library.

Indeed, ideally we'd only match <separator>trailer if we also find the same movie file. The same can probably be applied to samples as well?

In either case, having the space in that regexp is plain silly.

Hi guys....thanks for taking a look at this! Looks like a bona-fide issue that could use some sorting out.

Happy 4th of July!!!

I've made the changes in r21388. Let the bitching about trailers being added to the library commence!

I'm using an xbox with r21829 and I just noticed that trailers being downloaded from the Trailer Addict plugin are being added to the library. They are being named "movie - trailer.flv" with one space before the dash and two spaces after the dash.

This is correctly picked up by the scraper as a trailer, but when scanning for new content, it's also picked up as a separate movie. If I manually rename it to movie-trailer.flv then there is no problem scanning as a trailer and it does not get picked up as a separate movie.

Since the trailer scraper picks up the form of movie - trailer.extension, shouldn't the movie scraper disregard this format? I rarely use the trailer feature so it's not really an issue for me, but I can see it being a problem with others.

To be fair, you did ask for it althekiller... :0)