Live TV metadata scrapping

[Jdiesel 1114]

What fields are used for the live TV metadata scraping? Just wondering what I need to include in my TVXML to get metadata scraping to work. I have program title and description currently in my TVXML. 

[Luke 37120]

This will be your best reference:

 

http://wiki.xmltv.org/index.php/XMLTVFormat

[Jdiesel 1114]

I believe I have everything setup correctly as per the XMLTV and everything works with the exception of grabbing the show metadata. Below is an example from my XML:

<channel id="42.*****.tv">
                <display-name lang="en">Discovery</display-name>
                <display-name>42</display-name>
                <display-name>42 Discovery</display-name>
                <icon src="http://*****.*****.tv/assets/images/channels/42.png" />
                <url>http://*****.*****.tv:9999/viewms/c42.stream/playlist.m3u8</url>
        </channel>



<programme start="20170202140000" stop="20170202150000" channel="42.*****.tv">
                <title lang="en">Moonshiners</title>
                <desc lang="en">Tim Smith's whiskey time machine.</desc>
                <category lang="en">GeneralTV</category>
</programme>

This is what I see under the channel

 

I guess my question is, is it using the <title> </title> field to determine the show name to grab the metadata for? If not which field is needed?

[Luke 37120]

No, we don't do any internet metadata lookups so the xml data is expected to contain everything.

[Jdiesel 1114]

This setting is a bit misleading. What does it do exactly?

 

[Luke 37120]

At this time it is only used for movies. The problem with attempting to match programs is first the performance of adding http requests for every guide entry, and most importantly, the tvdb/moviedb search engines are designed to pretty much always return something even when not correct. This will lead to lots of incorrect metadata in the guide and therefore reports of "emby not working".

[Jdiesel 1114]

Makes sense, thanks for the explanation. 

[Luke 37120]

Thanks. It is easier to get better matches with movies because we also have the year. We don't have the series year with the guide data. We sometimes have the episode year, but it's the series year that could help us better filter the results.

[bardmaster 17]

As an add-on question:  Is there any way to stop the nfo files from being created?  I unchecked this box hoping for that result since, as you indicate above @@Luke, metadata is only downloaded for recorded movies (it would be nice to have this as a caption below the check-box), and all that is created is very basic nfo files that don't help with anything.

 

My main reason for asking is:  I employ MCEBuddy running as a service in the background on my server, and it is replacing the original video file with a comskip version.  MCEBuddy actually deletes the nfo with the same name as the video file, but it does not know to delete the "tvshow.nfo" file thus leaving clutter in the folder after every recording.

 

So unchecking this box should lead to NO metadata files of any sort being created, in my opinion.

 

ps. of note:  MCEBuddy is actually able to identify season and episode# when it scrapes TVDB.  Take a look at the below screenshot, resulting after running MCEBuddy on last night's new episode of Jane the Virgin (with auto-delete of source files disabled).  The MCEBuddy version is the second MP4.  So this scraping appears very feasible...

[Luke 37120]

No there isn't, because let's say in the future you uninstall/reinstall Emby Server from scratch. You probably are going to want a consistent result with what you are seeing now and the nfo files help ensure that.

[bardmaster 17]

Is there any available option to control the filenaming convention for recorded live TV?  At the very least the spec, but I would love to have modification options...like the option to not create a folder by showname.

Edited April 26, 2017 by bardmaster

[Luke 37120]

Not at the moment, no.