problem with finding metadata for swedish tv series

[n3mmr]

The series is named "Jakten på lyckan". Note the a-ring character in the second word.

It is available on thetvdb.com, it has show id 224331.

However, I cannot make serviio serve this as a series.

The files are named Jakten.på.lyckan.S01E01.SVT.Play.mp4 and Jakten.på.lyckan.S01E02.SVT.Play.mp4. They're in a directory named "Jakten pa lyckan

That ought to be right, oughtn't it?

What is wrong? Is there a simple fix?
Or is there some problem with searching thetvdb in non-ASCII scripts?

[Cerberus]

you need to remove the dots from the name or it will not work


Jakten på lyckan S01E01 SVT Play.mp4

NOT

Jakten.på.lyckan.S01E01.SVT.Play.mp4

[zip]

The dots should not matter, they are removed before making the request. Try renaming it with ascii characters if it makes any difference.

[n3mmr]

The dots should not matter, they are removed before making the request. Try renaming it with ascii characters if it makes any difference.

What ascii characters should that be?? thetvdb.com has it listed with non-ascii... There is no entry for an ascii-only name...

'

[n3mmr]

Pushing the check for video metadata only produces the line:

2011-02-01 17:38:32,166 INFO [LibraryManager] Forcing metadata update for 'VIDEO' media files

with no useful info

[zip]

Jakten.pa.lyckan......

Pushing the button marks video items as 'ready for update' and they'll be updated next time the thread runs or you press Force refresh. In your case, though, when you rename the file, it'll be removed and added again so no need to 'update'.

[n3mmr]

Jakten.pa.lyckan......

Pushing the button marks video items as 'ready for update' and they'll be updated next time the thread runs or you press Force refresh. In your case, though, when you rename the file, it'll be removed and added again so no need to 'update'.

There are more problems... serviio suddenly decided the well known movie "Aftershock" (New Zealand) was really Tangshan dadizhen
Note, Aftershock has been correctly named for a long time under 0.4.2, which I no longer have available.

Go figure. "Adieu deGaulle" has been renamed "Movies, Aspirin and Vultures" 0.4.2 couldn't find a name at all for that.

Hmm.

Bug???

Actually, these two problems are actually good arguments for multi-tier metadata mgmt..
The movies I have on my server are NOT the movies known about in the database.

[zip]

Searching for "Aftershock" on themoviedb.org actually returns the name you'referring to (aka). Is that not the movie?
The online databases are living things so a movie found one day may not be found the other, nothing to do with serviio version (unless there is actual code change in the file name parsing mechanism)

what's the file name of the other one (de gaulle)? It indeed doesn't exist on themoviedb.org

[n3mmr]

Searching for "Aftershock" on themoviedb.org actually returns the name you'referring to (aka). Is that not the movie?
The online databases are living things so a movie found one day may not be found the other, nothing to do with serviio version (unless there is actual code change in the file name parsing mechanism)

what's the file name of the other one (de gaulle)? It indeed doesn't exist on themoviedb.org

No, it isn't that movie at all. Mine is a movie made in New Zealand a few years ago.

This is a major problem with movies. Serviio DOES return the wrong name sometimes, but most often the right name: I think my idea for using the file name as an alternative in a controlled, per-movie way is very reasonable.

And DON'T say "update the databases"... there are at least 3 movies by the name Aftershock.

[zip]

What if we came up with a sidecar file, which would include (let's say) themoviedb.org or imdb id that would be used instead of filename parsing?

e.g.: serviio.mdt (includes imdbid=XXXXXXXX)

[n3mmr]

What if we came up with a sidecar file, which would include (let's say) themoviedb.org or imdb id that would be used instead of filename parsing?

e.g.: serviio.mdt (includes imdbid=XXXXXXXX)

The interesting movies do not always have entries in either imdb or themoviedb. They might be local, or even private, and they should still be presented inside the structured menu system.

I think we should not assume that these online databases contain all there is to know or that any one of them could be some kind of acceptable Arbiter Elegantiorum.

Also: with the sidecar the user must first personally find the info, then make serviio find it again.
What would be the gain??

[zip]

The gain is to help Serviio with parsing the files.

[n3mmr]

The gain is to help Serviio with parsing the files.

It wouldn't help with my examples, since the databases do NOT contain any info on these movies.

And the databases cannot possibly have info on privately produced movies, like "Jenny takes off", "Mom in the making" and such.

I guess what I want is an easy way to put "one-offs" into the context menu system.

It is a bit silly to have to go thru the folders for most of the interesting stuff, when the "Series" -> "Series_name" -> Episodes-in-the-right-order is so great a way to do it.

Well. What is needed may be a separate daemon: one that answers all "online metadata requests" from serviio.
Then that daemon could have a more complex metadata management.
Going out to various databases, looking in a local file or do something from the file name?

What would it take to write a daemon like that? I suppose you format an xml stream, send it off to a server, and expect another xml data blob back from the server??

Is there some sort of documentation for what you do and what you expect?

[zip]

Well. What is needed may be a separate daemon: one that answers all "online metadata requests" from serviio.
Then that daemon could have a more complex metadata management.
Going out to various databases, looking in a local file or do something from the file name?

What would it take to write a daemon like that? I suppose you format an xml stream, send it off to a server, and expect another xml data blob back from the server??

Is there some sort of documentation for what you do and what you expect?

There would have to be a new metadata extractor supporting your wrapper daemon I think. The current extractors got to the predefined API web addresses.

Create a ticket on bitbucket for your use case (not the daemon, but your requirements of multiple extractors at the same time), link this topic to it and I'll see what I can do

[n3mmr]

Well. What is needed may be a separate daemon: one that answers all "online metadata requests" from serviio.
Then that daemon could have a more complex metadata management.
Going out to various databases, looking in a local file or do something from the file name?

What would it take to write a daemon like that? I suppose you format an xml stream, send it off to a server, and expect another xml data blob back from the server??

Is there some sort of documentation for what you do and what you expect?

There would have to be a new metadata extractor supporting your wrapper daemon I think. The current extractors got to the predefined API web addresses.

Create a ticket on bitbucket for your use case (not the daemon, but your requirements of multiple extractors at the same time), link this topic to it and I'll see what I can do

Good man.

Coupladays!

Ahem: Bitbucket?? Can't seem to find my way..... Please show me...

[zip]

viewtopic.php?f=5&t=2

[Cerberus]

did this bitbucket get added or would you like me to do it for you n3mmr ??

[n3mmr]

did this bitbucket get added or would you like me to do it for you n3mmr ??

I did it just now.

Feel free to add to it, a comment, perhaps.

[n3mmr]

did this bitbucket get added or would you like me to do it for you n3mmr ??

I did it just now.

Feel free to add to it, a comment, perhaps.

issue #261

[Cerberus]

did this bitbucket get added or would you like me to do it for you n3mmr ??

I did it just now.

Feel free to add to it, a comment, perhaps.

issue #261

cool nope not something im intrested in at all tbh i have metadata turned off on my setup..