Create Plex Collection based on IMDB List


here i have a short Python script (Tested with python 2.7 on a qnap x64 plex installation)

For this script to work you need to get a plex token and fill in the first few lines.

This Script retrieves the IMDB CSV Export of a List and matches the entries to a Plex Library and Tags the Entries in the library

# -*- coding: utf-8 -*-

import requests
from math import ceil
from plexapi.server import PlexServer
from lxml.html import parse
import re
import csv
import urllib2

### Plex server details ###

### Existing movie library details ###
LIST_ID = 'ls074285945'
IMDB_LIST_URL = '{}/export'.format(LIST_ID)

def add_collection(library_key, rating_key):
    headers = {"X-Plex-Token": PLEX_TOKEN}
    params = {"type": 1,
              "id": rating_key,
              "collection[0].tag.tag": IMDB_COLLECTION_NAME,
              "collection.locked": 1

    url = "{base_url}/library/sections/{library}/all".format(base_url=PLEX_URL, library=library_key)
    r = requests.put(url, headers=headers, params=params)

def find_imdb_id(imdb_url):
    """Extract id from imdb_url"""
    re1 = '.*?'
    re2 = '(\\d+)'
    regex = re.compile(re1 + re2, re.IGNORECASE|re.DOTALL)
    match =
    if not match:
        return False
    imdb_id = "tt" +
    return imdb_id
def run_imdb_list():
        plex = PlexServer(PLEX_URL, PLEX_TOKEN)
        print("No Plex server found at: {base_url}".format(base_url=PLEX_URL))
        print("Exiting script.")
        return [], 0

    # Get the IMDB  list
    print("Retrieving the IMDB list...")
    i = 1
    list_picture_ids = []
    list_picture_titles = []
    list_picture_years = []
    title_lookup = {}
    response = urllib2.urlopen(IMDB_LIST_URL)
    reader = csv.reader(response)
    titles =
    reader = csv.DictReader(response, titles)
    for row in reader:
        imdburl = row['URL']
        imdb_id = find_imdb_id(imdburl)
        if imdb_id:
            title_lookup[imdb_id] = row['Title'] + " (" + row['Year'] + ")"
            print "ERROR: Unable to find IMDB ID for %s" % (row['Title'])

    # Get list of movies from the Plex server
    print("Retrieving a list of movies from the '{library}' library in Plex...".format(library=MOVIE_LIBRARY_NAME))
        movie_library = plex.library.section(MOVIE_LIBRARY_NAME)
        movie_library_key = movie_library.key
        all_movies = movie_library.all()
        print("The '{library}' library does not exist in Plex.".format(library=MOVIE_LIBRARY_NAME))
        print("Exiting script.")

    # Create a dictionary of {imdb_id: movie}
    imdb_map = {}
    for m in all_movies:
        if 'imdb://' in m.guid:
            imdb_id = m.guid.split('imdb://')[1].split('?')[0]
            if imdb_id in list_picture_ids:
                imdb_map[imdb_id] = m

    print("Setting the collection for the '{}' library...".format(MOVIE_LIBRARY_NAME))
    in_library_idx = []
    for i, imdb_id in enumerate(list_picture_ids):
        movie = imdb_map.pop(imdb_id, None)
        if movie:
            add_collection(movie_library_key, movie.ratingKey)

    return len(list_picture_ids)

if __name__ == "__main__":
    import time
    print time.strftime("%d/%m/%Y")

    list_count = run_imdb_list()
   # raw_input("Press Enter to finish...")

Hope this helps you.

The Script is based on the following but modified to Download the CSV instead of parsing the HTML. Here the source

Plex Sammlung nach imdb Liste erstellen


heute mal etwas spezielleres. Solltet ihr einen Plex Server haben und Film-Listen bei IMDB erstellt haben so kann euch das folgende Script helfen. Das Script nimmt den CSV Export von imDB und matcht dies zu deinen Plex Filmen und erstellt eine entsprechende Sammlung. Danke an robot2xl aus dem Plex Forum für die Vorlage. Ich habe das Script noch erweitert und meines ermittlt nun von den Plex Filmen welche per themoviedb getaggt wurden die entsprechende imdb id. – Das ganze bei heruntergefahrenem Plex und unter Linux mit Python.

Take imdb exported list and
create tag and tag all owned movies from CSV


import os
import sqlite3
import re
import csv
import getopt
import sys
import urllib
import json

DRYRUN = True #Don't UPDATE databse
REPORT = False #Print out unowned films

def db_exec(cursor, query):
    """Wrap DB query so we can print in DRYRUN mode"""
    if not DRYRUN:
        #print query

def connect_db():
    """Connect to Plex DB (linux only at the moment)"""
    #homedir = os.path.expanduser("~")
    #conn = sqlite3.connect(
    #        '%s/Library/Application Support/Plex Media Server/Plug-in Support/'
    #        'Databases/com.plexapp.plugins.library.db' % (homedir))
    conn = sqlite3.connect(
            '/volume1/Plex/Library/Application Support/'
            'Plex Media Server/Plug-in Support/Databases/'
    cursor = conn.cursor()
    return (conn, cursor)

def close_db(cursor):
    """Close DB cursor"""

def generate_field_string(user_fields):
    """Take existing user_fields from DB and add locks for tag"""
    fields = ""

    if user_fields:
        fields = re.split('=', user_fields)[1]

    if fields:
        items = fields.split('|')
        items = [int(x) for x in items]
        myset = set(items) #get rid of dupes
        items = list(myset)
        items = [str(x) for x in items]
        fields = '|'.join(items)
        fields = "15|16"

    new_fields = "lockedFields=%s" % fields
    return new_fields

def remove_tag(tags_collection, tag_to_remove):
    """ Build up collection string minus tag_to_remove """
    new_tags = ""
    items = tags_collection.split('|')
    for i in items:
        if i != tag_to_remove:
            if not new_tags:
                new_tags = i
                new_tags = "%s|%s" % (new_tags, i)

    return new_tags

def update_metadata_items(cursor, metadata_id, tag_title, untag=False):
    """Update tags and user_fields in metadata_items table"""
    cursor.execute('SELECT tags_collection, user_fields '
    'FROM metadata_items WHERE id="%s"' % (metadata_id))

    item = cursor.fetchone()
    tags_collection = item[0]
    user_fields = item[1]

    new_field_string = generate_field_string(user_fields)

    if tags_collection:
        items = tags_collection.split('|')
        for i in items:
            if i == tag_title and not untag:
        tags_collection = "%s|%s" % (tags_collection, tag_title)
        tags_collection = tag_title

    if untag:
        #Passed argument to unwind the tag set for film
        tags_collection = remove_tag(tags_collection, tag_title)

    db_exec(cursor, "UPDATE metadata_items SET tags_collection='%s',"
            "user_fields='%s' WHERE id='%s'"
            % (tags_collection, new_field_string, metadata_id))

def map_tag_to_metadata(cursor, tag_id, metadata_item_id, untag=False):
    """Create new taggings entry if it doesn't already exist"""
    cursor.execute('SELECT id FROM taggings '
                'WHERE tag_id="%s" and metadata_item_id="%s"'
                % (tag_id, metadata_item_id))
    item = cursor.fetchone()
    if not item:
        db_exec(cursor, 'INSERT into taggings (metadata_item_id, tag_id,'
                'created_at) VALUES ("%s", "%s", datetime("now"))'
                % (metadata_item_id, tag_id))
    elif untag:
        db_exec(cursor, 'DELETE from taggings WHERE '
            'metadata_item_id=%s and tag_id=%s'
            % (metadata_item_id, tag_id))

def insert_new_tag(cursor, title):
    """Insert new tag"""
    db_exec(cursor, "INSERT into tags (tag, tag_type, created_at, updated_at) "
            "VALUES ('%s', '2', datetime('now'), datetime('now'))" % (title))
    return cursor.lastrowid

def create_tag(cursor, title):
    """Check if tag exists and if not then insert into db"""
    cursor.execute('SELECT id, tag, tag_type FROM tags '
        'WHERE tag_type = 2 and tag = "%s"' % (title))
    item = cursor.fetchone()
    if item:
        #tag already exists in database
        tag_id = item[0]
        #tag needs to be inserted
        tag_id = insert_new_tag(cursor, title)

    return tag_id

def findIMDBBytmdbId(id):
    ''' Find movie by TheMovieDB ID '''
    xml = getjson(id)
    	result = json.load(xml)
    	return result['imdb_id']
    	return 0
def getjson(id):
    url = "%s/%s?%s" % ('', id, 'api_key=xxxxxxxxxxxxxxxxxxx')
    data = urllib.urlopen(url)
    return data
def imdbwatchlist(url):
    data = urllib.urlopen(url)
    return data
def find_imdb_id(imdb_url):
    """Extract id from imdb_url"""
    re1 = '.*?'
    re2 = '(\\d+)'
    regex = re.compile(re1 + re2, re.IGNORECASE|re.DOTALL)
    match =
    if not match:
        return False
    imdb_id = "tt" +
    return imdb_id

def fetch_film_ids(cursor, csv_filename):
    """Get metadata ids for all films from Plex that appear in ICM CSV file"""
    csv_imdb_ids = []
    plex_imdb_ids = []
    metadata_ids = []
    title_lookup = {}
    if not csv_filename:
        return "Specify csv!"
        fhandle = open(csv_filename, 'rb')
        reader = csv.reader(fhandle)
        titles =
        reader = csv.DictReader(fhandle, titles)
        for row in reader:
            imdburl = row['URL']
            imdb_id = find_imdb_id(imdburl)
            if imdb_id:
                title_lookup[imdb_id] = row['Title'] + " (" + row['Year'] + ")"
                print "ERROR: Unable to find IMDB ID for %s" % (row['Title'])

    cursor.execute('SELECT id, title, user_fields, guid '
            'FROM metadata_items WHERE metadata_type=1')

    films = cursor.fetchall()
    #go through all films in database and try to match from CSV
    for item in films:
    	if 'com.plexapp.agents.themoviedb' in item[3]:
    		imdburl = item[3]
    		imdburl = imdburl.replace("?lang=en","")
    		imdburl = imdburl.replace("?lang=de","")
    		imdb_id = findIMDBBytmdbId(imdburl.replace("com.plexapp.agents.themoviedb://",""))
        	imdburl = item[3]
        	imdb_id = find_imdb_id(imdburl)
        if imdb_id in csv_imdb_ids:
            #Keep ID lookup for optional reporting
            #matched movie from csv->db, so save metadata id

    if REPORT:
        unowned_films = list(set(csv_imdb_ids) - set(plex_imdb_ids))
        if unowned_films:
            print "-------"
            print "%d of %d films owned"  % (len(plex_imdb_ids),
            print "%d unowned films:" % (len(unowned_films))
            for imdb_id in unowned_films:
                print "\t" + title_lookup[imdb_id]
            print "-------"
        print "%d films owned:" % (len(plex_imdb_ids))
        for imdb_id in plex_imdb_ids:
            print "\t" + title_lookup[imdb_id]
        print "-------"

    return metadata_ids

def display_usage():
    """ Display usage information and exit """
    script_name = os.path.basename(__file__)
    print """Usage: %s [options]
        -f CSVFILENAME, --file=CSVFILENAME        ICM top list CSV export [REQUIRED]
        -t "TAG TITLE", --tag="TAG TITLE"         Name of collection tag [REQUIRED]
        -x, --execute                             Commit changes to database, else dry run
        -r, --report                              Print list of films to tag and missing films
        -u, --untag                               Remove "TAG TITLE" from all listed films, essentially an undo
        -h, --help                                Show this help message and exit

        %s -f 500+essential+cult+movies.csv -t "500 Essential Cult Movies" -x -r
        The above will tag all owned movies from the .csv file with the tag "500 Essential Cult Movies", commit changes to database and print a report
        This new collection can then be filtered in Plex
    """ % (script_name, script_name)

def main():
    Why does main need a docstring?

    tag_title = ""
    csv_filename = ""
    untag = False

        opts, dummy = getopt.getopt(sys.argv[1:], "f:t:xruh",
                ["file=", "tag=", "execute", "report", "untag", "help"])
    except getopt.GetoptError:
    for opt, arg in opts:
        if opt in ("-f", "--file"):
            csv_filename = arg
        if opt in ("-t", "--tag"):
            tag_title = arg
        if opt in ("-x", "--execute"):
            global DRYRUN
            DRYRUN = False
        if opt in ("-r", "--report"):
            global REPORT
            REPORT = True
        if opt in ("-u", "--untag"):
            untag = True
        if opt in ("-h", "--help"):

    if not csv_filename or not tag_title:

    conn, cursor = connect_db()
    metadata_ids = fetch_film_ids(cursor, csv_filename)

    tag_id = create_tag(cursor, tag_title)

    for meta_id in metadata_ids:
		map_tag_to_metadata(cursor, tag_id, meta_id, untag)
			update_metadata_items(cursor, meta_id, tag_title, untag)
			print meta_id



Synology volume crashed


(English Version below)
dieses mal wieder einmal eine Problem-Behebung.
Ich habe ein Synology NAS und dieses ist gecrashed. Erst wollte es nicht wirklich starten, aber nach etwas Geduld war ich wieder im Webinterface der Synology aber habe keine Daten mehr gesehen. Der Storage-Manager der Synology sagte nur kurz und knapp “Status: Abgestürzt” hat aber keine Lösung angeboten. Die meisten Artikel im Internet sagten, dass dies nicht so ohne weiteres zu lösen ist und man den Synology-Support einschalten muss, falls man seine Daten wieder braucht.

Nun habe ich es aber geschaft die Synology wieder zu reparieren.

Zuerst schalte man sich per SSH auf die Synology, danach führ ihr folgende Statements aus:

syno_poweroff_task -d
vgchange -ay
fsck.ext4 -pvf -C 0 /dev/vg1000/lv

Die erste Zeile nimmt euer Synology Offline und alle Diesnte werden gestoppt und das Volume wird entfernt.
Die zweite Zeile aktiviert das Volume wieder
Die dritte führt eine Reparatur durch. – Diese kann jedoch die Meldung “unexpected inconsistency” bringen. Dann folgende Zeile als Alternative der dritten ausführen:

fsck.ext4 -yvf -C 0 /dev/vg1000/lv

Es kann aber wie es bei mir war sein, dass die Befehle mit einem simplen “killed” beendet werden. In diesem Fall hat euer Synology nicht genügend Power (Arbeitsspeicher – bei mir nur 2 GB). Ich hatte glücklicherweise Zugriff auf ein Synology mit mehr RAM (12GB) wodurch eine Reparatur möglich war. Ansonsten können auch Daten auf die Festplatte anstatt in den Arbeitsspeicher ausgelagert werden, aber dies war bei mir nicht so erfolgreich.

Auf jeden Fall kann ich jetzt wieder an die Daten und das Volume zeigt wieder “Status: Normal”

Hoffe dies hilft auch euch bei einem gecrashtem Volume.



this time once again a problem solving.
I have a Synology NAS and this one is crashed. At first it wouldn’t really start up, but after some patience, I went back to the Synology web interface but didn’t see any data. Synology’s storage manager only said “Status: Crashed” briefly but didn’t offer any solution. Most of the articles on the Internet said that this is not easily solved and you have to call Synology support if you need your data back.

But now I managed to fix the Synology.

First, you connect to the Synology via SSH, and then execute the following statements:

syno_poweroff_task -e
vgchange -ay
fsck.ext4 -pvf -C 0 /dev/vg1000/lv

The first line takes your Synology offline and all diestetics are stopped and the volume is removed.
The second line reactivates the volume.
The third one carries out a repair. – However, this can result in the message “unexpected inconsistency”. Then execute the following line as an alternative to the third:

fsck.ext4 -yvf -C 0 /dev/vg1000/lv

But it can be as it was with me, that the commands are ended with a simple “killed”. In this case, your Synology does not have enough power (memory – only 2GB for me). Fortunately, I had access to a synology with more RAM (12GB), which allowed me to repair it. Otherwise, data can be swapped out to the hard drive instead of RAM, but this was not so successful for me.

Anyway, I can now access the data again and the volume shows “Status: Normal” again.

Hope this helps you with a crashed volume.