Boxee (the xbmc fork) have announced their own .nfo format - can xmg (and apps using it) support it?

[gavinc]

http://support.boxee.tv/entries/510214-creating-an-nfo-file-for-boxee-1-1

Didn't know if here or on the CP lighthouse was a better place to log this...

[dmwyatt]

The answer is "sure!". The problem is getting someone to write the code. :D

[gavinc]

Hah ha!

Checking out the source, am I right in thinking that "all" additional code would need to do is optionally dump out the data in the boxee specific format?

If I was to break out the python books, do you have a preferred style of specifying the choice of output format?

Another param, 'format' that takes 'xbmc' or 'boxee' for the MetaGen function?

G

[dmwyatt]

As it stands, the only thing xmg puts in a nfo file is the imdb url for the movie, since that is all that is needed to make sure that XBMC scrapes the right movie...there's not any need to generate a full .nfo.

If this isn't good enough for Boxee you'd probably want to add a keyword argument to write_nfo so that it's signature looks like this:

This way, the default is xbmc-style nfo files and you don't break existing code, and then add another method like this:

    #code to generate nfo

And then add code to write_nfo that checks the value of the 'format' parameter and if it specfies 'boxee' call _generate_boxee_nfo.

[tester22]

Shouldn't be to hard to do, could give it a try when I find time. The spec for Boxee/plex seems to be the same as used in xbmc: http://wiki.xbmc.org/index.php?title=Import_-_Export_Library#Movies

Most of those values are available in the json response from themovie db: http://jsonformatter.curiousconcept.com/#https://dl.dropbox.com/s/vdt7ljtvaqmbbcy/json.dump?dl=1

For me the best approach sound like using a mix of url and xml, like suggested here: http://wiki.xbmc.org/index.php?title=Import_-_Export_Library#Video_nfo_files_containing_a_mix_of_XML_and_URL

I have spent 30 min and this is the current output of the script now: https://dl.dropbox.com/s/s6hfeihiwz2c9kp/movie.nfo?dl=1

Should work but havn't tested it in plex yeat.

[tester22]

The current implementation does write out all fields in the already implemented api response from themoviedb. I could change this so it makes another call to collect more data, but is that interesting for anyone?

[gavinc]

It's interesting to me, as I would hope therms/ruud would flow it back uphill to the couchpotato app, and an option to output movie.nfo in boxee format would be made available.

BTW boxee changed their URLs, that spec of theirs is now at : http://boxee.zendesk.com/entries/510215-creating-an-nfo-file

[tester22]

Ok. will have a look on additional fields.

In the meantime you could download the raw file from here: https://github.com/tester22/CouchPotato/commit/a0344a0a202b4f6c6d89c5d42c0c742e0d107c7c

and replace the existing file in CP. This is needed since there are some CP specific changes to this compared to xmg master.

[dmwyatt]

I will talk to ruud about making sure xmg in cp is unaltered from my repo. That's just silly.

On Friday, July 1, 2011, tester22 reply@reply.github.com wrote:

Ok. will have a look on additional fields.

In the meantime you could download the raw file from here: https://github.com/tester22/CouchPotato/commit/a0344a0a202b4f6c6d89c5d42c0c742e0d107c7c

and replace the existing file in CP. This is needed since there are some CP specific changes to this compared to xmg master.

Reply to this email directly or view it on GitHub: https://github.com/therms/xmg/issues/1#issuecomment-1483691

Regards,

Dustin

[tester22]

I am working on improving the code a little bit and have a question for you guys. All responses from themoviedb contains all fields but they might be empty. If a field is empty how should I handle that? A field without value ? or as in the xbmc wiki Or to not write out the field in the xml file?

What do you think?