Emby Server 4.7.2.0 not reading folders like "john doe-tmdb-#####"

[pju 10]

After a long derailment - I've followed the path with <LockData>False</LockData> and I also have added a TMDBid where possible.
And it for sure has helped.
But I'm still fighting an issue with some actors not being picked up although they exists in the metadata folder and other actors in the same series are OK..
In the below example:

• Marta Dusseldorp has a legacy folder "Marta Dusseldorp-tmdb-152523" which has been migrated. Apparently this folder is not used, because a new one "Marta Dusseldorp" is created when I add "BlackJack" to a library.
• Tony Barry also have a legacy folder, but the info here is not used.

As far as I can see, 'series.xml' is read correctly, all metadata is there, just some of actors are not getting their metadata.

Any ideas...?

series.xml person.nfo person.nfo

[Happy2Play 9780]

My guess is Emby only read the xml and has no idea it needs to link only NAME to already known name-providerid as the provided metadata.

So these are not equal without additional lookups

 

    <name>Jason Isaacs</name>
    <role>Gabriel Lorca</role>
    <type>Actor</type>

 

    <name>Jason Isaacs</name>
    <role>Gabriel Lorca</role>
    <type>Actor</type>
    <tmdbid>11355</tmdbid>
    <tvdbid>346159</tvdbid>

At the same time the xml plugin would have to be updated to support reading person providerids if they existed

In theory this existing data somewhat becomes useless as it is unlinkable without querying the Series again to get providerid linked data.
 

[pju 10]

Mmmm, could be.  But for me it does not explain why most actors correctly gets metdata found in the legacy xmn files.

Also it looks liek pünktchen does some magic in his migration plugin.
Because if I include for instance BlackJack in a library before running the migratioin plugin, actors get their data.
If I load the series after running the plugin, some actors do not get their data.

[pju 10]

Having done quite a bit of further testing, I must admit that I still can't quite figure out what happens behind the scene.
But, I can see that with some work, a full migration from my very legacy data is possible with a very high rate of success.

So, thanks to everyone for support - mostly of course @pünktchenfor guidance and for a great plugin.