[BUG] Incorrect scraping and garbage leftover in .nfo files.

[McMovies 7]

Seems the movie identification is bugged recently.
Seen is few time in recent last 4-6 week but thought its edge-cases and ignored it. But recently realized its not edge case. and thought to alert the devs. 

Majorly 2 issues.
1. if a move is misidentified, and a trailer is saved in .nfo file. with this format <trailer>plugin://plugin.video.youtube/?action=play_video&amp;videoid=XXXXXXX</trailer> 
Re-identifying the movie to correct it, will replace everything but not the trailer section, it will show the trailer of old mis-identified movie.
The only way to fix this issue is to delete the .nfo file, rename the movie folder to something else, force/manually identify, and later re name the movie back to original name. 

2. partial-name match and mis-identification of movie.
for example "/Deva (2025)/Deva (2025).mp4" , is matching with this movie:https://www.themoviedb.org/movie/811941-part-1  and not the actual movie: https://www.themoviedb.org/movie/1195430
i wonder why is scrapping a wrong movie with totally different year and different name as Devara : Part 1 (2024)
its seems like the scrapper is acting like a lazy, never bother to check year, rather declaring result whenever it partially match something.


Steps to produce this bug:
0. Open terminal
1. Navigate to your movie library directory & create a folder called "Deva (2025)"
2. Make a dummy file inside - navigate to you directory in terminal & use this command--> touch "Deva (2025).mp4" )
3. Go to emby dashboard and do a library scan.
4. It will pick the wrong movie. -- hence follow step-5
5. Select the movie and click the three dots, click "identify"
6. in MovieDb Id: section paste this value "1195430"
7. Do a library scan/ refresh metadata.. resrart emby server...  do all what you can think of.
8. Now go to the movie and play the trailer... it will play wrong trailer.

repeat step 5 to 8 in as many folder you can think of, given the first movie must have trailer. even if the second one don't have trailer it will retain the trailer code <trailer>plugin://plugin.video.youtube/?action=play_video&amp;videoid=XXXXXXX</trailer>   form the first scrapped item.

[visproduction 315]

My Version 4.9.0.42 beta creates video youtube trailer lines differently.  Why is there a difference?

Quote

<trailer>plugin://plugin.video.youtube/play/?video_id=lmB3FUVQ_28</trailer>

Also try testing to see if the old code trailer playsback:

1. Try the same page with a different browser.
2. Try clearing the browser cache.
3. Try clicking Admin / Library / Scan Library Files button

 

[Luke 42077]

Quote

1. if a move is misidentified, and a trailer is saved in .nfo file. with this format <trailer>plugin://plugin.video.youtube/?action=play_video&amp;videoid=XXXXXXX</trailer> 
Re-identifying the movie to correct it, will replace everything but not the trailer section, it will show the trailer of old mis-identified movie.
The only way to fix this issue is to delete the .nfo file, rename the movie folder to something else, force/manually identify, and later re name the movie back to original name. 

Hi, try refreshing the metadata on it.

[Luke 42077]

Quote

2. partial-name match and mis-identification of movie.
for example "/Deva (2025)/Deva (2025).mp4" , is matching with this movie:https://www.themoviedb.org/movie/811941-part-1  and not the actual movie: https://www.themoviedb.org/movie/1195430
i wonder why is scrapping a wrong movie with totally different year and different name as Devara : Part 1 (2024)
its seems like the scrapper is acting like a lazy, never bother to check year, rather declaring result whenever it partially match something.

Hi there, please attach the Emby server log from when the problem occurred:

How to Report a Problem

Thanks!

 

[Luke 42077]

@McMovies?

[Happy2Play 9780]

Emby has somewhat always been that way as the reason "Remove Identification" option was added as Reidentifying something will not completely clear all existing fields/image it only replaces existing info.

But would need to see the server log for that specific example to see the server queries.

Testing Deva (2025)\Deva (2025).strm I got this.  As you can see Emby queried ‌&query=Deva&language=en-US&year=2025 and got correct result just like it being first on the website.

deva — The Movie Database (TMDB)

2025-04-27 08:45:46.989 Info TheMovieDb: MovieDbProvider: Finding id for item: Deva
2025-04-27 08:45:46.994 Info HttpClient: GET https://api.themoviedb.org/3/search/movie?api_key=‌f6bd687ffa63cd282b6ff2c6877f2669‌&query=Deva&language=en-US&year=2025
2025-04-27 08:45:47.168 Info HttpClient: GET https://api.themoviedb.org/3/movie/1195430?api_key=‌f6bd687ffa63cd282b6ff2c6877f2669‌&append_to_response=alternative_titles,reviews,casts,releases,images,keywords,trailers&language=en-US&include_image_language=en-US,null

But to add TMDB has some quirks with year as there are different release dates in each country so a movie can be returned with plus or minus a year in some/most cases. 

Example for your wrong movie

 

Edited April 27, 2025 by Happy2Play