ncarthy Wrote:I had also noticed that GetAlbums does the limit first, then the sorting. That is to say, if you specify start 0 and end 10 with sort ascending then it returns the first 10 albums it finds in the database and then sorts them. Those 10 albums are sorted by albumid I think. This behaviour can be demonstrated by doing a second query with a start of 10, end of 20 and comparing the results. It won't be the next 10 albums in alphabetical order.

GetArtists does it correctly, but at the very high cost of querying the database for ALL artists, then sorting and applying the limit.

Neither way is right... in an ideal world the sorting/limiting should be done by the database, not the libjson code. But that would require changes to the database code, outside of libjsonrpc.

This also leads to an inconsistency in the "total" parameter returned by libjsonrpc: For GetAlbums with start 0 and end 10 it will return 10, for GetArtists it will return the number of Artists in the database. I opened a Trac ticket for the totals problem in September (http://trac.xbmc.org/ticket/10146)

ncarthy Wrote:My workaround was to load all the albums into memory and do my own sorting/limiting.

vinc Wrote:Is there any way to get library information from a tv-show episode or movie? Stuff like plot, release year, rating etc.? Can't find the right method to call and "VideoLibrary.GetMovieInfo" seems to be deprecated?

Montellese Wrote:You need to use the "VideoLibrary.GetMovies" method and provide an array of all the fields you want. (AFAIK) Possible field values for movies are:

Code:

"title", "genre", "year", "rating", "director", "file", "trailer", "tagline", "plot", "plotoutline", "originaltitle", "lastplayed", "duration", "playcount", "writer", "studio", "mpaa", "movieid"

Montellese Wrote:You need to use the "VideoLibrary.GetMovies" method and provide an array of all the fields you want. (AFAIK) Possible field values for movies are:

Code:

"title", "genre", "year", "rating", "director", "file", "trailer", "tagline", "plot", "plotoutline", "originaltitle", "lastplayed", "duration", "playcount", "writer", "studio", "mpaa", "movieid"

vinc Wrote:I think I didn't explain good enough what I wanted to do; I want to query the information for a SINGLE movie of tvshow, should I be using the above method? I can't seem to get it working.

wuench Wrote:So I get that I can get the fanart, thumbnail, etc locations with VideoPlaylist.GetItems.

Is there a method to retrieve JPG's from the HTTP interface using the special:// URLs it returns?

Montellese Wrote:Ah sorry misunderstood you there. Dharma does not contain methods to get the details of a single movie/episode/song but such methods are already implemented in the current development source code.


Do you want to know how you can download files from the paths provided by the json rpc responses? If so give "Files.Download" a try. It takes an xbmc-specific path and returns an URL from where you can download the specified file.

Montellese Wrote:I'm not sure I understand what you want. The "id" tag should be used to match a response from XBMC to a request you sent yourself. Furthermore if you get a json object without an "id" tag it indicates that you received an announcement and not a response.

If you write your own parser you should basically count the number of { and } to determine the end of a message. And you should use a json library which can read a json object and detect the available fields and their values and not use regular expressions because there are alot of different response objects and I don't think you would want to write regex's for all of them.

Xedar Wrote:Hi Montellese

I want the "id", alot of them! in every { }
like this

Code:

{
   "id" : 50,
   "jsonrpc" : "2.0",
   "result" : {
      "end" : 2,
      "movies" : [
         {
           "id" : 50,
           "fanart" : "special://masterprofile/Thumbnails/Video/Fanart/a8c81876.tbn",
            "file" : "/home/xedar/movietest/Arn.vid.vägens.ände/arnvägen.mkv",
            "label" : "Arn - The Kingdom at Road's End",
            "movieid" : 1,
            "originaltitle" : "Arn - Riket vid vägens slut",
            "rating" : 9.0,
            "thumbnail" : "special://masterprofile/Thumbnails/Video/a/a8c81876.tbn"
         },
         {
            "id" : 50,
            "fanart" : "special://masterprofile/Thumbnails/Video/Fanart/7655f397.tbn",
            "file" : "/home/xedar/movietest/Sommar.med.göran/sommargöran.mkv",
            "label" : "Sommaren med Göran – En midsommarnattskomedi",
            "movieid" : 2,
            "originaltitle" : "Sommaren med Göran – En midsommarnattskomedi",
            "rating" : 4.50,
            "thumbnail" : "special://masterprofile/Thumbnails/Video/7/7655f397.tbn"
         }
      ],
      "start" : 0,
      "total" : 2
   }
}

Problem is iam limited to none JSON, i send request to raw TCP and match with regex.
We use Iviewer for Ipad http://www.commandfusion.com/

right now I do it ugly, matching the "movieid" field instead of id # for movies and yes regex can be a pain when parsing alot...
I can understand its not a good feature to change standard JSON outputs.

Best regards.