Difference between revisions of "WDLXTV MediaLibrary"

Revision as of 10:29, 8 October 2011

Preamble

This is still a work in progress and heavily beta!

This page will be updated continuously while there is ongoing development.
If you intend to play around with this you should bring basic linux skills and be able to work on the shell.

Changelog

v1.3.11
 - added ExtSheet support for UMSP plugin (WDTVExt plugin included)
 - added MediaMark support for UMSP plugin (WDTVExt plugin included)

v1.3.10
 - fixed TMDB double imports
 - fixed web plugin sql query

v1.3.9
 - fixed tgmd import
 - fixed sqlite web tool buttons
 - fixed a couple of sql related bugs
 - added support for setting cache dir (eg. config_tool -c MOVIESCAN_CACHE_DIR='/tmp/mnt/whatever/.tmdb')

v1.3.8
 - initial beta release

Introduction

History

This project is something I was carrying around with me for quite some time now and never got to finish it.
It all started when I started playing around with sqlite and the media library that comes with the WD firmware. It always bothered me that it was only working with locally attached media and not with xmounted network shares. (yeah, there is the .addMounts hack, but that's not what I had in mind.) So I tried inserting my own entries into the database just to discover that DMAOSD periodically deletes them again. Also there are quite some problems when you try to use sqlite with a database file on a network share (locking issue) which drove me to implement WDLXTV-DB (which is in the firmware for quite some time now and can already be used by MediaMark for example).

wdlxtv-ml.app.bin

Here's the second step now and I've been working on it for quite some time.
This app.bin provides a script ('moviescan') which will import mediafiles into the database and offers the possibility to search TMDB (http://www.themoviedb.org/) for movie info (rating, cast, etc, ...) and put this info into the database as well. Optionally you can import this information from .tgmd files if available.

What is this good for?

You can use this database for many cool things, I included two small examples:

• a very simple media browser
  This will only show the files in your library and allows you to toggle the MediaMark for the file and edit some file properties like title, year, imdb id, tmdb id.
• a UMSP plugin
  This will enable you to browse your files sorted by different criteria: year, rating, latest additions, genre.

These are meant as examples and I hope you come up with some even more sophisticated additions.You will find some develpoment informations further down in this article.

Download

• wdlxtv-ml.app-1.3.9.tgz
• wdlxtv-ml.app-1.3.10.tgz
• wdlxtv-ml.app-1.3.11.tgz

Installation

• get WDLXTV_DB up and running
• download wdlxtv-ml.app.tgz
• extract wdlxtv-ml.app.bin and .wdtvext-plugins to USB drive and connect it to WDTV
  
• to enable Moviesheets in UMSP plugin set EXTSHEET_USE_SQL=ON (available in WEC)

config_tool -c EXTSHEET_USE_SQL=ON

• reboot

Usage

This app.bin provides a shell script moviescan which imports data into the sqlite database.

Also included is a simple web plugin and UMSP plugin as examples.

Help

Run the script without any arguments or -h to get a little help:

# moviescan 
  moviescan version 1.3.8 by recliq

  usage: moviescan [options] update <PATH> [<PATH> [<PATH>]] [<DEPTH>]
         moviescan [options] init
         moviescan [options] clean

  options:
	-a|--auto [<num>|s]	enable auto chooser, default: s (skip)
	-c|--use-cache		enable tmdb file cache
	-d|--debug		enable debug output
	-e|--use-dirname	use dirname instead of filename for search
	-g|--use-tgmd		use .tgmd files as info source
	-i|--no-vidinfo		disable video info
	-l|--lang		set TMDB language (two letter country code), default: en
	-n|--use-nfo		search .nfo files for imdb id
	-s|--simul		do a dry-run without changing any data
	-t|--use-tmdb		use TMDB as info source

I will explain the important options later on.

Init database

The first thing you have to do is to create the needed database structures and initialize the database:

moviescan init

This will drop any existing databases and data without asking you again!

Import files

Then you are ready to import some files, in this case i'll import from my xmounted TestShare

moviescan update /tmp/media/usb/TestShare

Options

• -h or --help
  Display commandline help.
  

• -v or --verbose
  Enable verbose mode.
  

• -d or --debug
  Enable debug mode. (lot's of output!)
  

• -i or --no-vidinfo
  Disable getting video info with ffmpeg.
  

• -a or --auto [<num> | s]
  Enable auto mode. If multiple results are found on TMDB search the script will not ask to select a title but will always auto answer with the given number or skip.
  Default: s (skip)

• -g or --use-tgmd
  Searches for a .tgmd file and uses the info provided to feed the database.
  

• -t or --use-tmdb
  Searches for movie information using TMDB (themoviedb.org).

• -c or --use-cache *EXPERIMENTAL*
  Enable TMDB file cache.
  You have to set this path before you can use this feature:

  config_tool -c MOVIESCAN_CACHE_DIR='/tmp/mnt/whatever/.tmdb'

• -e or --use-dirname
  Use the parent dirname instead of the filename to build the search name.
  

• -n or --use-nfo
  Search for imdb id in existing .nfo files.
  

• -l or --lang
  Set TMDB language for results (two letter country code), falls back to en if language is not available.
  Default: en

Development

SQLite table schemas

• Table video

CREATE TABLE video(
    hash VARCHAR(16) primary key,
    title VARCHAR,
    file VARCHAR,
    path VARCHAR,
    size UNSIGNED INTEGER,
    mtime INTEGER,
    mtime_date VARCHAR(15),
    duration UNSIGNED INTEGER,
    bitrate VARCHAR(16),
    vcodec VARCHAR(32),
    height SMALLINT,
    width SMALLINT,
    fps FLOAT,
    type VARCHAR(20),
    year VARCHAR(4),
    imdb VARCHAR(32),
    acodec VARCHAR(32),
    channels VARCHAR(16),
    abitrate VARCHAR(16),
    tvdb VARCHAR(32),
    epid VARCHAR(32),
    tmdb UNSIGNED INTEGER
);

• Table movies

CREATE TABLE movies (
    id INTEGER PRIMARY KEY,
    imdb VARCHAR(32),
    tmdb UNSIGNED INTEGER,
    title VARCHAR,
    originaltitle VARCHAR,
    tagline VARCHAR,
    plot VARCHAR,
    year UNSIGNED SMALLINT,
    runtime UNSIGNED SMALLINT,
    rating float,
    genre VARCHAR,
    actor VARCHAR,
    studio VARCHAR,
    coutry VARCHAR,
    releasedate VARCHAR(12),
    director VARCHAR,
    writers VARCHAR,
    certification VARCHAR(4),
    mpaa VARCHAR
);

ToDo

• add support for TV series (TVDB)
• get MediaMark working
• get Moviesheets working
• get Linksheets working

Bugs and Discussion

WDLXTV Forum - WDLXTV Media Library (beta)

--Recliq 2011