]> Raphaël G. Git Repositories - youtubedl/blobdiff - youtube_dl/extractor/orf.py
Set upstream metadata fields: Bug-Database, Bug-Submit, Repository, Repository-Browse.
[youtubedl] / youtube_dl / extractor / orf.py
index 011e6be13e63562dad8def87ea264a7e1b6783af..d54b8ace6e5eeee6ee7188bd1cfe88a2471f3627 100644 (file)
 # coding: utf-8
 from __future__ import unicode_literals
 
-import json
 import re
-import calendar
-import datetime
 
 from .common import InfoExtractor
+from ..compat import compat_str
 from ..utils import (
+    clean_html,
+    determine_ext,
+    float_or_none,
     HEADRequest,
+    int_or_none,
+    orderedSet,
+    remove_end,
+    str_or_none,
+    strip_jsonp,
+    unescapeHTML,
     unified_strdate,
-    ExtractorError,
+    url_or_none,
 )
 
 
 class ORFTVthekIE(InfoExtractor):
     IE_NAME = 'orf:tvthek'
     IE_DESC = 'ORF TVthek'
-    _VALID_URL = r'https?://tvthek\.orf\.at/(?:programs/.+?/episodes|topics/.+?|program/[^/]+)/(?P<id>\d+)'
+    _VALID_URL = r'https?://tvthek\.orf\.at/(?:[^/]+/)+(?P<id>\d+)'
 
-    _TEST = {
-        'url': 'http://tvthek.orf.at/program/matinee-Was-Sie-schon-immer-ueber-Klassik-wissen-wollten/7317210/Was-Sie-schon-immer-ueber-Klassik-wissen-wollten/7319746/Was-Sie-schon-immer-ueber-Klassik-wissen-wollten/7319747',
-        'file': '7319747.mp4',
-        'md5': 'bd803c5d8c32d3c64a0ea4b4eeddf375',
+    _TESTS = [{
+        'url': 'http://tvthek.orf.at/program/Aufgetischt/2745173/Aufgetischt-Mit-der-Steirischen-Tafelrunde/8891389',
+        'playlist': [{
+            'md5': '2942210346ed779588f428a92db88712',
+            'info_dict': {
+                'id': '8896777',
+                'ext': 'mp4',
+                'title': 'Aufgetischt: Mit der Steirischen Tafelrunde',
+                'description': 'md5:c1272f0245537812d4e36419c207b67d',
+                'duration': 2668,
+                'upload_date': '20141208',
+            },
+        }],
+        'skip': 'Blocked outside of Austria / Germany',
+    }, {
+        'url': 'http://tvthek.orf.at/topic/Im-Wandel-der-Zeit/8002126/Best-of-Ingrid-Thurnher/7982256',
         'info_dict': {
-            'title': 'Was Sie schon immer über Klassik wissen wollten',
-            'description': 'md5:0ddf0d5f0060bd53f744edaa5c2e04a4',
-            'duration': 3508,
-            'upload_date': '20140105',
+            'id': '7982259',
+            'ext': 'mp4',
+            'title': 'Best of Ingrid Thurnher',
+            'upload_date': '20140527',
+            'description': 'Viele Jahre war Ingrid Thurnher das "Gesicht" der ZIB 2. Vor ihrem Wechsel zur ZIB 2 im Jahr 1995 moderierte sie unter anderem "Land und Leute", "Österreich-Bild" und "Niederösterreich heute".',
         },
-        'skip': 'Blocked outside of Austria',
-    }
+        'params': {
+            'skip_download': True,  # rtsp downloads
+        },
+        'skip': 'Blocked outside of Austria / Germany',
+    }, {
+        'url': 'http://tvthek.orf.at/topic/Fluechtlingskrise/10463081/Heimat-Fremde-Heimat/13879132/Senioren-betreuen-Migrantenkinder/13879141',
+        'only_matching': True,
+    }, {
+        'url': 'http://tvthek.orf.at/profile/Universum/35429',
+        'only_matching': True,
+    }]
 
     def _real_extract(self, url):
-        mobj = re.match(self._VALID_URL, url)
-        playlist_id = mobj.group('id')
+        playlist_id = self._match_id(url)
         webpage = self._download_webpage(url, playlist_id)
 
-        data_json = self._search_regex(
-            r'initializeAdworx\((.+?)\);\n', webpage, 'video info')
-        all_data = json.loads(data_json)
-
-        def get_segments(all_data):
-            for data in all_data:
-                if data['name'] == 'Tracker::EPISODE_DETAIL_PAGE_OVER_PROGRAM':
-                    return data['values']['segments']
-
-        sdata = get_segments(all_data)
-        if not sdata:
-            raise ExtractorError('Unable to extract segments')
-
-        def quality_to_int(s):
-            m = re.search('([0-9]+)', s)
-            if m is None:
-                return -1
-            return int(m.group(1))
+        data_jsb = self._parse_json(
+            self._search_regex(
+                r'<div[^>]+class=(["\']).*?VideoPlaylist.*?\1[^>]+data-jsb=(["\'])(?P<json>.+?)\2',
+                webpage, 'playlist', group='json'),
+            playlist_id, transform_source=unescapeHTML)['playlist']['videos']
 
         entries = []
-        for sd in sdata:
-            video_id = sd['id']
-            formats = [{
-                'preference': -10 if fd['delivery'] == 'hls' else None,
-                'format_id': '%s-%s-%s' % (
-                    fd['delivery'], fd['quality'], fd['quality_string']),
-                'url': fd['src'],
-                'protocol': fd['protocol'],
-                'quality': quality_to_int(fd['quality']),
-            } for fd in sd['playlist_item_array']['sources']]
+        for sd in data_jsb:
+            video_id, title = sd.get('id'), sd.get('title')
+            if not video_id or not title:
+                continue
+            video_id = compat_str(video_id)
+            formats = []
+            for fd in sd['sources']:
+                src = url_or_none(fd.get('src'))
+                if not src:
+                    continue
+                format_id_list = []
+                for key in ('delivery', 'quality', 'quality_string'):
+                    value = fd.get(key)
+                    if value:
+                        format_id_list.append(value)
+                format_id = '-'.join(format_id_list)
+                ext = determine_ext(src)
+                if ext == 'm3u8':
+                    m3u8_formats = self._extract_m3u8_formats(
+                        src, video_id, 'mp4', m3u8_id=format_id, fatal=False)
+                    if any('/geoprotection' in f['url'] for f in m3u8_formats):
+                        self.raise_geo_restricted()
+                    formats.extend(m3u8_formats)
+                elif ext == 'f4m':
+                    formats.extend(self._extract_f4m_formats(
+                        src, video_id, f4m_id=format_id, fatal=False))
+                else:
+                    formats.append({
+                        'format_id': format_id,
+                        'url': src,
+                        'protocol': fd.get('protocol'),
+                    })
 
             # Check for geoblocking.
             # There is a property is_geoprotection, but that's always false
@@ -90,16 +127,27 @@ class ORFTVthekIE(InfoExtractor):
                             % geo_str),
                         fatal=False)
 
+            self._check_formats(formats, video_id)
             self._sort_formats(formats)
 
-            upload_date = unified_strdate(sd['created_date'])
+            subtitles = {}
+            for sub in sd.get('subtitles', []):
+                sub_src = sub.get('src')
+                if not sub_src:
+                    continue
+                subtitles.setdefault(sub.get('lang', 'de-AT'), []).append({
+                    'url': sub_src,
+                })
+
+            upload_date = unified_strdate(sd.get('created_date'))
             entries.append({
                 '_type': 'video',
                 'id': video_id,
-                'title': sd['header'],
+                'title': title,
                 'formats': formats,
+                'subtitles': subtitles,
                 'description': sd.get('description'),
-                'duration': int(sd['duration_in_seconds']),
+                'duration': int_or_none(sd.get('duration_in_seconds')),
                 'upload_date': upload_date,
                 'thumbnail': sd.get('image_full_url'),
             })
@@ -111,71 +159,279 @@ class ORFTVthekIE(InfoExtractor):
         }
 
 
-# Audios on ORF radio are only available for 7 days, so we can't add tests.
+class ORFRadioIE(InfoExtractor):
+    def _real_extract(self, url):
+        mobj = re.match(self._VALID_URL, url)
+        station = mobj.group('station')
+        show_date = mobj.group('date')
+        show_id = mobj.group('show')
+
+        data = self._download_json(
+            'http://audioapi.orf.at/%s/api/json/current/broadcast/%s/%s'
+            % (station, show_id, show_date), show_id)
+
+        entries = []
+        for info in data['streams']:
+            loop_stream_id = str_or_none(info.get('loopStreamId'))
+            if not loop_stream_id:
+                continue
+            title = str_or_none(data.get('title'))
+            if not title:
+                continue
+            start = int_or_none(info.get('start'), scale=1000)
+            end = int_or_none(info.get('end'), scale=1000)
+            duration = end - start if end and start else None
+            entries.append({
+                'id': loop_stream_id.replace('.mp3', ''),
+                'url': 'http://loopstream01.apa.at/?channel=%s&id=%s' % (station, loop_stream_id),
+                'title': title,
+                'description': clean_html(data.get('subtitle')),
+                'duration': duration,
+                'timestamp': start,
+                'ext': 'mp3',
+                'series': data.get('programTitle'),
+            })
+
+        return {
+            '_type': 'playlist',
+            'id': show_id,
+            'title': data.get('title'),
+            'description': clean_html(data.get('subtitle')),
+            'entries': entries,
+        }
+
+
+class ORFFM4IE(ORFRadioIE):
+    IE_NAME = 'orf:fm4'
+    IE_DESC = 'radio FM4'
+    _VALID_URL = r'https?://(?P<station>fm4)\.orf\.at/player/(?P<date>[0-9]+)/(?P<show>4\w+)'
+
+    _TEST = {
+        'url': 'http://fm4.orf.at/player/20170107/4CC',
+        'md5': '2b0be47375432a7ef104453432a19212',
+        'info_dict': {
+            'id': '2017-01-07_2100_tl_54_7DaysSat18_31295',
+            'ext': 'mp3',
+            'title': 'Solid Steel Radioshow',
+            'description': 'Die Mixshow von Coldcut und Ninja Tune.',
+            'duration': 3599,
+            'timestamp': 1483819257,
+            'upload_date': '20170107',
+        },
+        'skip': 'Shows from ORF radios are only available for 7 days.',
+        'only_matching': True,
+    }
 
 
-class ORFOE1IE(InfoExtractor):
+class ORFOE1IE(ORFRadioIE):
     IE_NAME = 'orf:oe1'
     IE_DESC = 'Radio Österreich 1'
-    _VALID_URL = r'http://oe1\.orf\.at/programm/(?P<id>[0-9]+)'
+    _VALID_URL = r'https?://(?P<station>oe1)\.orf\.at/player/(?P<date>[0-9]+)/(?P<show>\w+)'
+
+    _TEST = {
+        'url': 'http://oe1.orf.at/player/20170108/456544',
+        'md5': '34d8a6e67ea888293741c86a099b745b',
+        'info_dict': {
+            'id': '2017-01-08_0759_tl_51_7DaysSun6_256141',
+            'ext': 'mp3',
+            'title': 'Morgenjournal',
+            'duration': 609,
+            'timestamp': 1483858796,
+            'upload_date': '20170108',
+        },
+        'skip': 'Shows from ORF radios are only available for 7 days.'
+    }
+
+
+class ORFIPTVIE(InfoExtractor):
+    IE_NAME = 'orf:iptv'
+    IE_DESC = 'iptv.ORF.at'
+    _VALID_URL = r'https?://iptv\.orf\.at/(?:#/)?stories/(?P<id>\d+)'
+
+    _TEST = {
+        'url': 'http://iptv.orf.at/stories/2275236/',
+        'md5': 'c8b22af4718a4b4af58342529453e3e5',
+        'info_dict': {
+            'id': '350612',
+            'ext': 'flv',
+            'title': 'Weitere Evakuierungen um Vulkan Calbuco',
+            'description': 'md5:d689c959bdbcf04efeddedbf2299d633',
+            'duration': 68.197,
+            'thumbnail': r're:^https?://.*\.jpg$',
+            'upload_date': '20150425',
+        },
+    }
 
     def _real_extract(self, url):
-        mobj = re.match(self._VALID_URL, url)
-        show_id = mobj.group('id')
+        story_id = self._match_id(url)
+
+        webpage = self._download_webpage(
+            'http://iptv.orf.at/stories/%s' % story_id, story_id)
+
+        video_id = self._search_regex(
+            r'data-video(?:id)?="(\d+)"', webpage, 'video id')
 
         data = self._download_json(
-            'http://oe1.orf.at/programm/%s/konsole' % show_id,
-            show_id
-        )
+            'http://bits.orf.at/filehandler/static-api/json/current/data.json?file=%s' % video_id,
+            video_id)[0]
+
+        duration = float_or_none(data['duration'], 1000)
+
+        video = data['sources']['default']
+        load_balancer_url = video['loadBalancerUrl']
+        abr = int_or_none(video.get('audioBitrate'))
+        vbr = int_or_none(video.get('bitrate'))
+        fps = int_or_none(video.get('videoFps'))
+        width = int_or_none(video.get('videoWidth'))
+        height = int_or_none(video.get('videoHeight'))
+        thumbnail = video.get('preview')
+
+        rendition = self._download_json(
+            load_balancer_url, video_id, transform_source=strip_jsonp)
 
-        timestamp = datetime.datetime.strptime('%s %s' % (
-            data['item']['day_label'],
-            data['item']['time']
-        ), '%d.%m.%Y %H:%M')
-        unix_timestamp = calendar.timegm(timestamp.utctimetuple())
+        f = {
+            'abr': abr,
+            'vbr': vbr,
+            'fps': fps,
+            'width': width,
+            'height': height,
+        }
+
+        formats = []
+        for format_id, format_url in rendition['redirect'].items():
+            if format_id == 'rtmp':
+                ff = f.copy()
+                ff.update({
+                    'url': format_url,
+                    'format_id': format_id,
+                })
+                formats.append(ff)
+            elif determine_ext(format_url) == 'f4m':
+                formats.extend(self._extract_f4m_formats(
+                    format_url, video_id, f4m_id=format_id))
+            elif determine_ext(format_url) == 'm3u8':
+                formats.extend(self._extract_m3u8_formats(
+                    format_url, video_id, 'mp4', m3u8_id=format_id))
+            else:
+                continue
+        self._sort_formats(formats)
+
+        title = remove_end(self._og_search_title(webpage), ' - iptv.ORF.at')
+        description = self._og_search_description(webpage)
+        upload_date = unified_strdate(self._html_search_meta(
+            'dc.date', webpage, 'upload date'))
 
         return {
-            'id': show_id,
-            'title': data['item']['title'],
-            'url': data['item']['url_stream'],
-            'ext': 'mp3',
-            'description': data['item'].get('info'),
-            'timestamp': unix_timestamp
+            'id': video_id,
+            'title': title,
+            'description': description,
+            'duration': duration,
+            'thumbnail': thumbnail,
+            'upload_date': upload_date,
+            'formats': formats,
         }
 
 
-class ORFFM4IE(InfoExtractor):
-    IE_DESC = 'orf:fm4'
-    IE_DESC = 'radio FM4'
-    _VALID_URL = r'http://fm4\.orf\.at/7tage/?#(?P<date>[0-9]+)/(?P<show>\w+)'
+class ORFFM4StoryIE(InfoExtractor):
+    IE_NAME = 'orf:fm4:story'
+    IE_DESC = 'fm4.orf.at stories'
+    _VALID_URL = r'https?://fm4\.orf\.at/stories/(?P<id>\d+)'
+
+    _TEST = {
+        'url': 'http://fm4.orf.at/stories/2865738/',
+        'playlist': [{
+            'md5': 'e1c2c706c45c7b34cf478bbf409907ca',
+            'info_dict': {
+                'id': '547792',
+                'ext': 'flv',
+                'title': 'Manu Delago und Inner Tongue live',
+                'description': 'Manu Delago und Inner Tongue haben bei der FM4 Soundpark Session live alles gegeben. Hier gibt es Fotos und die gesamte Session als Video.',
+                'duration': 1748.52,
+                'thumbnail': r're:^https?://.*\.jpg$',
+                'upload_date': '20170913',
+            },
+        }, {
+            'md5': 'c6dd2179731f86f4f55a7b49899d515f',
+            'info_dict': {
+                'id': '547798',
+                'ext': 'flv',
+                'title': 'Manu Delago und Inner Tongue live (2)',
+                'duration': 1504.08,
+                'thumbnail': r're:^https?://.*\.jpg$',
+                'upload_date': '20170913',
+                'description': 'Manu Delago und Inner Tongue haben bei der FM4 Soundpark Session live alles gegeben. Hier gibt es Fotos und die gesamte Session als Video.',
+            },
+        }],
+    }
 
     def _real_extract(self, url):
-        mobj = re.match(self._VALID_URL, url)
-        show_date = mobj.group('date')
-        show_id = mobj.group('show')
+        story_id = self._match_id(url)
+        webpage = self._download_webpage(url, story_id)
 
-        data = self._download_json(
-            'http://audioapi.orf.at/fm4/json/2.0/broadcasts/%s/4%s' % (show_date, show_id),
-            show_id
-        )
-
-        def extract_entry_dict(info, title, subtitle):
-            return {
-                'id': info['loopStreamId'].replace('.mp3', ''),
-                'url': 'http://loopstream01.apa.at/?channel=fm4&id=%s' % info['loopStreamId'],
-                'title': title,
-                'description': subtitle,
-                'duration': (info['end'] - info['start']) / 1000,
-                'timestamp': info['start'] / 1000,
-                'ext': 'mp3'
+        entries = []
+        all_ids = orderedSet(re.findall(r'data-video(?:id)?="(\d+)"', webpage))
+        for idx, video_id in enumerate(all_ids):
+            data = self._download_json(
+                'http://bits.orf.at/filehandler/static-api/json/current/data.json?file=%s' % video_id,
+                video_id)[0]
+
+            duration = float_or_none(data['duration'], 1000)
+
+            video = data['sources']['q8c']
+            load_balancer_url = video['loadBalancerUrl']
+            abr = int_or_none(video.get('audioBitrate'))
+            vbr = int_or_none(video.get('bitrate'))
+            fps = int_or_none(video.get('videoFps'))
+            width = int_or_none(video.get('videoWidth'))
+            height = int_or_none(video.get('videoHeight'))
+            thumbnail = video.get('preview')
+
+            rendition = self._download_json(
+                load_balancer_url, video_id, transform_source=strip_jsonp)
+
+            f = {
+                'abr': abr,
+                'vbr': vbr,
+                'fps': fps,
+                'width': width,
+                'height': height,
             }
 
-        entries = [extract_entry_dict(t, data['title'], data['subtitle']) for t in data['streams']]
+            formats = []
+            for format_id, format_url in rendition['redirect'].items():
+                if format_id == 'rtmp':
+                    ff = f.copy()
+                    ff.update({
+                        'url': format_url,
+                        'format_id': format_id,
+                    })
+                    formats.append(ff)
+                elif determine_ext(format_url) == 'f4m':
+                    formats.extend(self._extract_f4m_formats(
+                        format_url, video_id, f4m_id=format_id))
+                elif determine_ext(format_url) == 'm3u8':
+                    formats.extend(self._extract_m3u8_formats(
+                        format_url, video_id, 'mp4', m3u8_id=format_id))
+                else:
+                    continue
+            self._sort_formats(formats)
 
-        return {
-            '_type': 'playlist',
-            'id': show_id,
-            'title': data['title'],
-            'description': data['subtitle'],
-            'entries': entries
-        }
\ No newline at end of file
+            title = remove_end(self._og_search_title(webpage), ' - fm4.ORF.at')
+            if idx >= 1:
+                # Titles are duplicates, make them unique
+                title += ' (' + str(idx + 1) + ')'
+            description = self._og_search_description(webpage)
+            upload_date = unified_strdate(self._html_search_meta(
+                'dc.date', webpage, 'upload date'))
+
+            entries.append({
+                'id': video_id,
+                'title': title,
+                'description': description,
+                'duration': duration,
+                'thumbnail': thumbnail,
+                'upload_date': upload_date,
+                'formats': formats,
+            })
+
+        return self.playlist_result(entries)