[xodi]

Hi, I manually download the various artworks (poster/fanart, etc., but no actor thumbnails as I don't want to keep duplicate actor thumbnails) for movie "2001: A Space Odyssey" (https://www.imdb.com/title/tt0062622/) and created a NFO file for it. The NFO file e has the following content:

xml:


<?xml version="1.0" encoding="UTF-8" standalone="yes"?>

<movie>

    <title>2001: A Space Odyssey</title>

    <originaltitle>2001: A Space Odyssey</originaltitle>

    <rating>8</rating>

    <votes>5367</votes>

    <top250>0</top250>

    <year>1968</year>

    <tagline>The ultimate trip</tagline>

    <runtime>149</runtime>

    <mpaa>U</mpaa>

    <playcount>0</playcount>

    <id>tt0062622</id>

    <tmdbid>62</tmdbid>

    <set>

        <name>The Space Odyssey Series</name>

        <overview/>

    </set>

    <sorttitle></sorttitle>

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

    <watched>false</watched>

    <credits>Stanley Kubrick</credits>

    <credits>Arthur C. Clarke</credits>

    <director>Stanley Kubrick</director>

    <studio>Stanley Kubrick Productions</studio>

    <studio>Metro-Goldwyn-Mayer</studio>

    <genre>Science Fiction</genre>

    <genre>Mystery</genre>

    <genre>Adventure</genre>

    <country>United Kingdom</country>

    <country>United States of America</country>

</movie>

All artworks worked fine except the actor thumbnails. I tried both TMDB scrapper and UMS scrapper, but none of them would scrap actors information (and the actor thumbnails) for the movie. If I added the actors element to the NFO file, like the following, the actor name appears, but no thumbnails for it.

xml:


    <actor>

        <name>Keir Dullea</name>

        <role>Dr. David Bowman</role>

    </actor>

I thought Kodi will use information in the NFO file, and for any information not exist in the NFO file, it will download from internet, but seems this is not the case. How can I ask Kodi to scrap the actor information?

BTW, I had "Download Actor Thumbnails" option enabled in Kodi settings. From the progress bar at the top right corner during the scrapper, it seems Kodi never queried internet to get information for the movie. The whole scrapping process was too fast.

[Karellen]

I thought Kodi will use information in the NFO file

Yes it does.

for any information not exist in the NFO file, it will download from internet

Yes and no.
Yes if there is a link to the missing information
No if there is no link.

If you had the following actors data, then Kodi would know what to do...

xml:
    <actor>

        <name>Ian McShane</name>

        <role>Mr. Wednesday</role>

        <order>1</order>

        <thumb>http://image.tmdb.org/t/p/original/pY9ud4BJwHekNiO4MMItPbgkdAy.jpg</thumb>

    </actor>

Simple answer is that your NFO Files are incomplete. As well as some of your tags are outdated (assuming you are using v18). See the NFO Files link in My Signature below.

How can you scrape the Actors information? You can't do that individually. You need to scrape the movie again without the NFO File. Why do you need the NFO File? As I said, it is incomplete.

[Karellen]

@xodi

Here, have my NFO File for the movie... https://paste.kodi.tv/suhamuvoku

Also take note of the difference.

[xodi]

Thanks for the information. I use NFO file because I want to change something, like adding an additional tag to a movie. Kodi database corrupted before several times due to exception, I needed to add these tags again and again when rebuilding database. So I wanted to add tag to the NFO file, so each time to rescrape the movie, the tag will be added automatically. And I want to keep the NFO file as simple as it can be, so I didn't keep any URLs in the NFO file.

From the log, seems Kodi thought the NFO file is a "full" NFO, I didn't find in the WiKi (https://kodi.wiki/view/NFO_files) which thing made Kodi though it's full NFO. Seems I need make a metadata NFO. How can I do that?

[Karellen]

These are the types of NFO Files that Kodi core uses...

xml:
NO_NFO       = 0, //!< No info found

FULL_NFO     = 1, //!< Full info specified

URL_NFO      = 2, //!< A URL to grab info from was found

OVERRIDE_NFO = 3, //!< Override info was found

COMBINED_NFO = 4, //!< A URL to grab info from + override info was found

ERROR_NFO    = 5, //!< Error processing info

TITLE_NFO = 6 //!< At least Title was read (and optionally the Year)

https://github.com/xbmc/xbmc/blob/eb161c...nner.h#L39

Basically the Full NFO means a metadata nfo. ie, it has required information in it that can be scanned into the library directly. It does not mean it is "full and complete". This makes it distinct from URLNFO which is a Parsing NFO which contains the link to the movie to be scraped.

I use NFO file because I want to change something, like adding an additional tag to a movie

Sure, that is reasonable. But an easier way to accomplish this is to scrape the movie with Kodi, Export to Separate Files, make the changes in the NFO File and Refresh back into the library.

Or, to cut out a few steps you can make the changes using Chorus2 Web Browser, then Export to Separate files.

I will move this thread out of Developers section and up to Metadata Add-ons.

[xodi]

Thank you very much @Karellen, these information are very helpful. I just found another way for this. It seems AEON MQ8 skin can download actor thumbnails automatically with the help of Skin Helper Service from TMDB website. I only need put the actor name in the NFO file. It works perfectly and I don't need save any URL in my NFO. Thanks again for your detailed information about the scraping.

[Karellen]

Thread marked solved.