# coding: utf-8
from __future__ import unicode_literals

import re
from .common import InfoExtractor
from ..utils import js_to_json


class GaskrankIE(InfoExtractor):
    """InfoExtractor for gaskrank.tv"""
    _VALID_URL = r'https?://(?:www\.)?gaskrank\.tv/tv/(?P<categories>[^/]+)/(?P<id>[^/]+)\.html?'
    _TESTS = [
        {
            'url': 'http://www.gaskrank.tv/tv/motorrad-fun/strike-einparken-durch-anfaenger-crash-mit-groesserem-flurschaden.htm',
            'md5': '1ae88dbac97887d85ebd1157a95fc4f9',
            'info_dict': {
                'id': '201601/26955',
                'ext': 'mp4',
                'title': 'Strike! Einparken können nur Männer - Flurschaden hält sich in Grenzen *lol*',
                'thumbnail': r're:^https?://.*\.jpg$',
                'categories': ['motorrad-fun'],
                'display_id': 'strike-einparken-durch-anfaenger-crash-mit-groesserem-flurschaden',
                'uploader_id': 'Bikefun',
                'upload_date': '20170110',
                'uploader_url': None,
                'tags': ['honkj', 'strike', 'idiot', 'depp', 'fahrschule', 'einparken', 'parken', 'autounfall', 'unfall', 'flurschaden', 'crash', 'accident', 'fail', 'rueckwaerts', 'umfaller'],
            }
        },
        {
            'url': 'http://www.gaskrank.tv/tv/racing/isle-of-man-tt-2011-michael-du-15920.htm',
            'md5': 'c33ee32c711bc6c8224bfcbe62b23095',
            'info_dict': {
                'id': '201106/15920',
                'ext': 'mp4',
                'title': 'Isle of Man - Michael Dunlop vs Guy Martin - schwindelig kucken',
                'thumbnail': r're:^https?://.*\.jpg$',
                'categories': ['racing'],
                'display_id': 'isle-of-man-tt-2011-michael-du-15920',
                'uploader_id': 'IOM',
                'upload_date': '20160506',
                'uploader_url': 'www.iomtt.com',
                'tags': ['schwindelig', 'iom', 'isle of man', 'guy martin', 'glen helen', 'michael dunlop', 'tt 2011', 'attacke'],
            }
        }
    ]

    def _real_extract(self, url):
        """extract information from gaskrank.tv"""
        def fix_json(code):
            """Removes trailing comma in json: {{},} --> {{}}"""
            return re.sub(r',[\s]*?}', r'}', js_to_json(code))

        display_id = self._match_id(url)
        webpage = self._download_webpage(url, display_id)
        categories = [re.match(self._VALID_URL, url).group('categories')]
        title = self._search_regex(r'movieName\s*:\s*\'([^\']*)\'', webpage, 'title')
        thumbnail = self._search_regex(r'poster\s*:\s*\'([^\']*)\'', webpage, 'thumbnail', default=None)

        mobj = re.search(r'Video von:\s*(?P<uploader_id>[^|]*?)\s*\|\s*vom:\s*(?P<upload_date_day>[0-9][0-9]).(?P<upload_date_month>[0-9][0-9]).(?P<upload_date_year>[0-9][0-9][0-9][0-9])', webpage)
        uploader_id = mobj.groupdict().get('uploader_id')
        upload_date = mobj.groupdict().get('upload_date_year') + mobj.groupdict().get('upload_date_month') + mobj.groupdict().get('upload_date_day')
        if len(upload_date) != 8:
            upload_date = None

        uploader_url = self._search_regex(r'Homepage:\s*<[^>]*?>(?P<uploader_url>[^<]*)', webpage, 'uploader_url', default=None)
        tags = re.findall(r'/tv/tags/[^/]*?/\"\s*>(?P<tag>[^<]*?)<', webpage)

        view_count = self._search_regex(r'class\s*=\s*\"gkRight\"(?:[^>]*>\s*<[^>]*)*icon-eye-open(?:[^>]*>\s*<[^>]*)*>\s*(?P<view_count>[0-9\.]*)', webpage, 'view_count')
        if view_count:
            view_count = int(view_count.replace('.', ''))

        average_rating = self._search_regex(r'itemprop\s*=\s*\"ratingValue\"\s*>\s*(?P<average_rating>[0-9,]*)', webpage, 'average_rating')
        if average_rating:
            average_rating = float(average_rating.replace(',', '.'))

        playlist = self._parse_json(
            self._search_regex(r'playlist\s*:\s*\[([^\]]*)\]', webpage, 'playlist', default='{}'),
            display_id, transform_source=fix_json, fatal=False)

        video_id = self._search_regex(r'https?://movies\.gaskrank\.tv/([^-]*?)(-[^\.]*)?\.mp4', playlist.get('0').get('src'), 'video id')

        formats = []
        for key in playlist:
            formats.append({
                'url': playlist[key]['src'],
                'format_id': key,
                'quality': playlist[key].get('quality')})
        self._sort_formats(formats, field_preference=['format_id'])

        return {
            'id': video_id,
            'title': title,
            'formats': formats,
            'thumbnail': thumbnail,
            'categories': categories,
            'display_id': display_id,
            'uploader_id': uploader_id,
            'upload_date': upload_date,
            'uploader_url': uploader_url,
            'tags': tags,
            'view_count': view_count,
            'average_rating': average_rating,
        }