[multicore]

Hello,
In one of the recent version (can't point which), Kodi has fail to scrap video file info or maybe became more strict with certain file name conventions/formats?
I have files in my "older" video collection that were scanned with success before.
Some file name examples:

• Ai_1080p
  
• Austin_Powers_in_Goldmember_2002_m720p_BDRip_AAC_x264-m4c14k_Uploaded.By.KizMark
  
• Annie Hall
  
• alice in wonderland 1080
  
• Airplane.1980.m-720p.ManO.-.Aftms
  
• Beetlejuice.1988.720p.BluRay.DTS.x264-ESiR
  

I checked with wiki here and here, but couldn't make much sense from this. I remember that Kodi used to be "smart" enough to parse the correct data.
Here is relevant part from the log: https://paste.kodi.tv/simaquhupa.kodi
I can see the HTTP request to the API and the response with cURL. I tried it myself and found that the file with different name format (like after removing video info and other release group stuff) will provide a valid JSON document with the movie information.

Any idea what can I do to solve this or why Kodi's video scrap system has changed?

EDIT:
I noticed that "Annie Hall" was actually looking for a different title, prrobably from the Matroska metadata:
22:41:16.145 T:12244   DEBUG: CScraperUrl::Get: Using "UTF-8" charset for "https://api.tmdb.org/3/search/movie?api_key=f090bb54758cabf231fb605d3e3e0468&query=Annie.Hall.1977.720p.HDTV.x264%20SZ&year=&language=en"

[Karellen]

Update your scraper.

[multicore]

Sorry for the lazy question, but what scraper you suggest to use?
Currently, I got The Movie Database and Universal Movie Scraper set.

BTW, I'm using nightly build version of 18.0 from two days ago.

[Karellen]

Use the same one. Just update it. You are using an old version.

Also, name your files correctly. Your naming is terrible and you will always have trouble scraping them.

[multicore]

Thanks for the swift reply!

The naming is not mine. This how most of the pirated movies (I hope that I'm not violating any law here) named for years.
I noticed that the srcaper is running queries with movie metadata instead of file name now. This is why even "normal" names fail to be scanned.

Example:
Bad:
https://api.tmdb.org/3/search/movie?api_...anguage=en

Good (after manual fixing the URL):
https://api.tmdb.org/3/search/movie?api_...anguage=en

Notice it add ampersand notation and details from the title tag that get the scrap service confused.

[Karellen]

Yes, that is obvious that you are using pirated movies, which we dont assist with, hence my very brief replies.

Now I am not sure what you are trying to convey with your example, but I have advised you how to fix. It is up to you if you want to heed my advice.

The excuse about file naming is pretty poor.

Good luck.

[multicore]

Hi Karellen,
I really appreciate you will to help and I'm not using the naming format here as an excuse here.
The thing is, a while ago, I could enter manual search text string if getting info was failing. It seems like some patterns in metadata of the video files now prevent that.

BTW, I just upgraded all the video info plugin and result are the same. I trying swapping scrapers now.

[multicore]

In case anyone find interest in this forum thread, I fixed the problem by editing Matroska media container metadata by myself.
That's the only way an API call to the scrapper service will give a good response.

What was probably changed is the way scrapper is looking for information.
Right now, its parsed from "Title" metadata, which can have junk not related to the movie title at all so file naming has no effect here.

The tool that helped me with this task was MKVToolNix.
The "Header Editor" can modify and commit the changes in metadata.
Other video containers will need their tools for the job.