[arte] Add support for multiple formats

[yt-dlp.git] / youtube_dl / extractor / arte.py

# encoding: utf-8
from __future__ import unicode_literals

import re
import json

from .common import InfoExtractor
from ..utils import (
    ExtractorError,
    find_xpath_attr,
    unified_strdate,
    determine_ext,
    get_element_by_id,
    compat_str,
    get_element_by_attribute,
)

# There are different sources of video in arte.tv, the extraction process 
# is different for each one. The videos usually expire in 7 days, so we can't
# add tests.


class ArteTvIE(InfoExtractor):
    _VALID_URL = r'http://videos\.arte\.tv/(?P<lang>fr|de)/.*-(?P<id>.*?)\.html'
    IE_NAME = 'arte.tv'

    def _real_extract(self, url):
        ref_xml_url = url.replace('/videos/', '/do_delegate/videos/')
        ref_xml_url = ref_xml_url.replace('.html', ',view,asPlayerXml.xml')
        ref_xml_doc = self._download_xml(
            ref_xml_url, video_id, note='Downloading metadata')
        config_node = find_xpath_attr(ref_xml_doc, './/video', 'lang', lang)
        config_xml_url = config_node.attrib['ref']
        config = self._download_xml(
            config_xml_url, video_id, note='Downloading configuration')

        formats = [{
            'forma_id': q.attrib['quality'],
            'url': q.text,
            'quality': 2 if q.attrib['quality'] == 'hd' else 1,
        } for q in config.findall('.//quality')]
        self._sort_formats(formats)

        title = config.find('.//name').text
        thumbnail = config.find('.//firstThumbnailUrl').text
        return {
            'id': video_id,
            'title': title,
            'thumbnail': thumbnail,
            'url': video_url,
            'ext': 'flv',
        }


class ArteTVPlus7IE(InfoExtractor):
    IE_NAME = 'arte.tv:+7'
    _VALID_URL = r'https?://(?:www\.)?arte\.tv/guide/(?P<lang>fr|de)/(?:(?:sendungen|emissions)/)?(?P<id>.*?)/(?P<name>.*?)(\?.*)?'

    @classmethod
    def _extract_url_info(cls, url):
        mobj = re.match(cls._VALID_URL, url)
        lang = mobj.group('lang')
        # This is not a real id, it can be for example AJT for the news
        # http://www.arte.tv/guide/fr/emissions/AJT/arte-journal
        video_id = mobj.group('id')
        return video_id, lang

    def _real_extract(self, url):
        video_id, lang = self._extract_url_info(url)
        webpage = self._download_webpage(url, video_id)
        return self._extract_from_webpage(webpage, video_id, lang)

    def _extract_from_webpage(self, webpage, video_id, lang):
        json_url = self._html_search_regex(r'arte_vp_url="(.*?)"', webpage, 'json url')
        return self._extract_from_json_url(json_url, video_id, lang)

    def _extract_from_json_url(self, json_url, video_id, lang):
        json_info = self._download_webpage(json_url, video_id, 'Downloading info json')
        self.report_extraction(video_id)
        info = json.loads(json_info)
        player_info = info['videoJsonPlayer']

        info_dict = {
            'id': player_info['VID'],
            'title': player_info['VTI'],
            'description': player_info.get('VDE'),
            'upload_date': unified_strdate(player_info.get('VDA', '').split(' ')[0]),
            'thumbnail': player_info.get('programImage') or player_info.get('VTU', {}).get('IUR'),
        }

        all_formats = player_info['VSR'].values()
        # Some formats use the m3u8 protocol
        all_formats = list(filter(lambda f: f.get('videoFormat') != 'M3U8', all_formats))
        def _match_lang(f):
            if f.get('versionCode') is None:
                return True
            # Return true if that format is in the language of the url
            if lang == 'fr':
                l = 'F'
            elif lang == 'de':
                l = 'A'
            regexes = [r'VO?%s' % l, r'VO?.-ST%s' % l]
            return any(re.match(r, f['versionCode']) for r in regexes)
        # Some formats may not be in the same language as the url
        formats = filter(_match_lang, all_formats)
        formats = list(formats) # in python3 filter returns an iterator
        if not formats:
            # Some videos are only available in the 'Originalversion'
            # they aren't tagged as being in French or German
            if all(f['versionCode'] == 'VO' for f in all_formats):
                formats = all_formats
            else:
                raise ExtractorError(u'The formats list is empty')

        if re.match(r'[A-Z]Q', formats[0]['quality']) is not None:
            def sort_key(f):
                return ['HQ', 'MQ', 'EQ', 'SQ'].index(f['quality'])
        else:
            def sort_key(f):
                return (
                    # Sort first by quality
                    int(f.get('height',-1)),
                    int(f.get('bitrate',-1)),
                    # The original version with subtitles has lower relevance
                    re.match(r'VO-ST(F|A)', f.get('versionCode', '')) is None,
                    # The version with sourds/mal subtitles has also lower relevance
                    re.match(r'VO?(F|A)-STM\1', f.get('versionCode', '')) is None,
                    # Prefer http downloads over m3u8
                    0 if f['url'].endswith('m3u8') else 1,
                )
        formats = sorted(formats, key=sort_key)
        def _format(format_info):
            quality = ''
            height = format_info.get('height')
            if height is not None:
                quality = compat_str(height)
            bitrate = format_info.get('bitrate')
            if bitrate is not None:
                quality += '-%d' % bitrate
            if format_info.get('versionCode') is not None:
                format_id = '%s-%s' % (quality, format_info['versionCode'])
            else:
                format_id = quality
            info = {
                'format_id': format_id,
                'format_note': format_info.get('versionLibelle'),
                'width': format_info.get('width'),
                'height': height,
            }
            if format_info['mediaType'] == 'rtmp':
                info['url'] = format_info['streamer']
                info['play_path'] = 'mp4:' + format_info['url']
                info['ext'] = 'flv'
            else:
                info['url'] = format_info['url']
                info['ext'] = determine_ext(info['url'])
            return info
        info_dict['formats'] = [_format(f) for f in formats]

        return info_dict


# It also uses the arte_vp_url url from the webpage to extract the information
class ArteTVCreativeIE(ArteTVPlus7IE):
    IE_NAME = 'arte.tv:creative'
    _VALID_URL = r'https?://creative\.arte\.tv/(?P<lang>fr|de)/magazine?/(?P<id>.+)'

    _TEST = {
        'url': 'http://creative.arte.tv/de/magazin/agentur-amateur-corporate-design',
        'info_dict': {
            'id': '050489-002',
            'ext': 'mp4',
            'title': 'Agentur Amateur / Agence Amateur #2 : Corporate Design',
        },
    }


class ArteTVFutureIE(ArteTVPlus7IE):
    IE_NAME = 'arte.tv:future'
    _VALID_URL = r'https?://future\.arte\.tv/(?P<lang>fr|de)/(thema|sujet)/.*?#article-anchor-(?P<id>\d+)'

    _TEST = {
        'url': 'http://future.arte.tv/fr/sujet/info-sciences#article-anchor-7081',
        'info_dict': {
            'id': '050940-003',
            'ext': 'mp4',
            'title': 'Les champignons au secours de la planète',
        },
    }

    def _real_extract(self, url):
        anchor_id, lang = self._extract_url_info(url)
        webpage = self._download_webpage(url, anchor_id)
        row = get_element_by_id(anchor_id, webpage)
        return self._extract_from_webpage(row, anchor_id, lang)


class ArteTVDDCIE(ArteTVPlus7IE):
    IE_NAME = 'arte.tv:ddc'
    _VALID_URL = r'https?://ddc\.arte\.tv/(?P<lang>emission|folge)/(?P<id>.+)'

    def _real_extract(self, url):
        video_id, lang = self._extract_url_info(url)
        if lang == 'folge':
            lang = 'de'
        elif lang == 'emission':
            lang = 'fr'
        webpage = self._download_webpage(url, video_id)
        scriptElement = get_element_by_attribute('class', 'visu_video_block', webpage)
        script_url = self._html_search_regex(r'src="(.*?)"', scriptElement, 'script url')
        javascriptPlayerGenerator = self._download_webpage(script_url, video_id, 'Download javascript player generator')
        json_url = self._search_regex(r"json_url=(.*)&rendering_place.*", javascriptPlayerGenerator, 'json url')
        return self._extract_from_json_url(json_url, video_id, lang)


class ArteTVConcertIE(ArteTVPlus7IE):
    IE_NAME = 'arte.tv:concert'
    _VALID_URL = r'https?://concert\.arte\.tv/(?P<lang>de|fr)/(?P<id>.+)'

    _TEST = {
        'url': 'http://concert.arte.tv/de/notwist-im-pariser-konzertclub-divan-du-monde',
        'md5': '9ea035b7bd69696b67aa2ccaaa218161',
        'info_dict': {
            'id': '186',
            'ext': 'mp4',
            'title': 'The Notwist im Pariser Konzertclub "Divan du Monde"',
            'upload_date': '20140128',
            'description': 'md5:486eb08f991552ade77439fe6d82c305',
        },
    }