FlexGet Kinopoisk plugin

config.yml

templates:
  movies:
    set:
      path: /tmp/mnt/94E8B2B1E8B290CA/Torrents/download/DLNA/Movies
    transmission:
      host: ************
      port: ****
      username: *************
      password: *************
    email:
      from: **********@**********.ru
      to: **********@**********.ru
      title: FlexGet Notification
      smtp_host: *************
      smtp_port: 25
      smtp_username: *************
      smtp_password: '*************'

tasks:
  nnm-club:
    rss:
      url: http://nnm-club.me/forum/rss2.php?f=218&t=1&uk=*************
      other_fields: [description]
      ascii: yes
    accept_all: yes
    manipulate:
      - title:
          replace:
            regexp: '[^\x00-\x80]+'
            format: ''
      - kinopoisk_id:
          from: description
          replace:
            regexp: '(.|\n)*kinopoisk.ru/rating/([0-9]+)(.|\n)*'
            format: '\2'
    kinopoisk:
      min_score: 7.0
      min_votes: 1000
    template:
      - movies

kinopoisk.py

from __future__ import unicode_literals, division, absolute_import
import urllib2, xmltodict
import logging

from flexget import plugin
from flexget.event import event
from flexget.plugins.filter.seen import FilterSeen

log = logging.getLogger('kinopoisk')


class FilterKinopoisk(FilterSeen):
    """
       Kinopoisk filter

       Using:

         Add option to config.yml:
           kinopoisk:
             min_score: 7.0      # ignore movies with score below 7.0, default: 0
             min_votes: 1000     # ignore movies with votes below 1000, default: 0
             matching: 'strict'  # ignore entries without kinopoisk_id, values: 'strict' or 'loose', default: 'strict'
             scope: 'local'      # use global or local storage, values: 'global' or 'local', default: 'global'

         Please, replace values of '7.0' and '1000' to your
    """
    schema = {
        'oneOf': [
            {
                'type': 'object',
                'properties': {
                    'min_score': {'type': 'number'},
                    'min_votes': {'type': 'integer'},
                    'matching': {'type': 'string', 'enum': ['strict', 'loose']},
                    'scope': {'type': 'string', 'enum': ['global', 'local']}
                }
            }
        ]
    }

    def __init__(self):
        # remember and filter by these fields
        self.fields = ['kinopoisk_id']
        self.keyword = 'kinopoisk'

    def is_number(self, s):
        try:
            float(s)
            return True
        except ValueError:
            return False

    @plugin.priority(0) # Make filter run after other filters, but before exists_movies
    def on_task_filter(self, task, config):
        if config is False:
            return

        # Reject all entries without kinopoisk id
        if config.get('matching') != 'loose':
            for entry in task.entries:
                if 'kinopoisk_id' not in entry or not self.is_number(entry['kinopoisk_id']):
                    log.info('Rejecting %s because of missing movie kinopoisk id' % entry['title'])
                    entry.reject('Missing movie kinopoisk id')

        # call super
        super(FilterKinopoisk, self).on_task_filter(task, config.get('scope', True))

        # check that two copies of a movie have not been accepted this run
        kinopoisk_ids = set()

        for entry in task.accepted:
            if not self.is_number(entry['kinopoisk_id']):
                entry.reject('Kinopoisk ID not found')
                continue

            if entry['kinopoisk_id'] in kinopoisk_ids and self.is_number(entry['kinopoisk_id']):
                entry.reject('Already accepted once in task')
                continue
            else:
                kinopoisk_ids.add(entry['kinopoisk_id'])

            try:
                page = urllib2.urlopen('http://rating.kinopoisk.ru/' + entry['kinopoisk_id'] + '.xml')
            except urllib2.HTTPError:
                entry.reject('Rating is not found')
                continue

            doc = page.read()
            page.close()
            data = xmltodict.parse(doc)

            score = float(data['rating']['kp_rating']['#text'])
            votes = float(data['rating']['kp_rating']['@num_vote'])

            if 'min_score' in config:
                if score <= 0:
                    entry.reject('Score is zero')
                    continue

                if score < config['min_score']:
                    entry.reject('Score (%s) below minimum (%s)' % (score, config['min_score']))
                    continue

            if 'min_votes' in config:
                if votes < config['min_votes']:
                    entry.reject('Votes (%s) below minimum (%s)' % (votes, config['min_votes']))
                    continue
            log.debug('Accepting %s' % (entry['title']))


@event('plugin.register')
def register_plugin():
    plugin.register(FilterKinopoisk, 'kinopoisk', api_ver=2)

2018-04-22 21:21 CRITICAL task          nnm-club        BUG: Unhandled error in plugin kinopoisk: not well-formed (invalid token): line 1, column 0
Traceback (most recent call last):
  File "d:\program\python\python27\lib\site-packages\flexget\task.py", line 486, in __run_plugin
    return method(*args, **kwargs)
  File "d:\program\python\python27\lib\site-packages\flexget\event.py", line 23, in __call__
    return self.func(*args, **kwargs)
  File "C:\Users\222\flexget\plugins\kinopoisk.py", line 90, in on_task_filter
    data = xmltodict.parse(doc)
  File "d:\program\python\python27\lib\site-packages\xmltodict.py", line 330, in parse
    parser.Parse(xml_input, True)
ExpatError: not well-formed (invalid token): line 1, column 0

Как пофиксить? Спасибо.

Looks like sometimes it gives reponse in gzip format. It can be fixed like that:

doc = page.read()
page.close()
try:
    data = xmltodict.parse(doc)
except ExpatError:
    compressed_data = io.BytesIO(doc)
    for compressed_data_item in gzip.GzipFile(fileobj=compressed_data):
        data = xmltodict.parse(compressed_data_item)