[arcpublishing] Add new extractor
authorRemita Amine <remitamine@gmail.com>
Thu, 31 Dec 2020 22:31:07 +0000 (23:31 +0100)
committerRemita Amine <remitamine@gmail.com>
Thu, 31 Dec 2020 22:31:07 +0000 (23:31 +0100)
closes #2298
closes #9340
closes #17200

youtube_dl/extractor/arcpublishing.py [new file with mode: 0644]
youtube_dl/extractor/extractors.py
youtube_dl/extractor/generic.py
youtube_dl/extractor/washingtonpost.py

diff --git a/youtube_dl/extractor/arcpublishing.py b/youtube_dl/extractor/arcpublishing.py
new file mode 100644 (file)
index 0000000..d1fb1a0
--- /dev/null
@@ -0,0 +1,173 @@
+# coding: utf-8
+from __future__ import unicode_literals
+
+import re
+
+from .common import InfoExtractor
+from ..utils import (
+    extract_attributes,
+    int_or_none,
+    parse_iso8601,
+    try_get,
+)
+
+
+class ArcPublishingIE(InfoExtractor):
+    _UUID_REGEX = r'[\da-f]{8}-(?:[\da-f]{4}-){3}[\da-f]{12}'
+    _VALID_URL = r'arcpublishing:(?P<org>[a-z]+):(?P<id>%s)' % _UUID_REGEX
+    _TESTS = [{
+        # https://www.adn.com/politics/2020/11/02/video-senate-candidates-campaign-in-anchorage-on-eve-of-election-day/
+        'url': 'arcpublishing:adn:8c99cb6e-b29c-4bc9-9173-7bf9979225ab',
+        'only_matching': True,
+    }, {
+        # https://www.bostonglobe.com/video/2020/12/30/metro/footage-released-showing-officer-talking-about-striking-protesters-with-car/
+        'url': 'arcpublishing:bostonglobe:232b7ae6-7d73-432d-bc0a-85dbf0119ab1',
+        'only_matching': True,
+    }, {
+        # https://www.actionnewsjax.com/video/live-stream/
+        'url': 'arcpublishing:cmg:cfb1cf1b-3ab5-4d1b-86c5-a5515d311f2a',
+        'only_matching': True,
+    }, {
+        # https://elcomercio.pe/videos/deportes/deporte-total-futbol-peruano-seleccion-peruana-la-valorizacion-de-los-peruanos-en-el-exterior-tras-un-2020-atipico-nnav-vr-video-noticia/
+        'url': 'arcpublishing:elcomercio:27a7e1f8-2ec7-4177-874f-a4feed2885b3',
+        'only_matching': True,
+    }, {
+        # https://www.clickondetroit.com/video/community/2020/05/15/events-surrounding-woodward-dream-cruise-being-canceled/
+        'url': 'arcpublishing:gmg:c8793fb2-8d44-4242-881e-2db31da2d9fe',
+        'only_matching': True,
+    }, {
+        # https://www.wabi.tv/video/2020/12/30/trenton-company-making-equipment-pfizer-covid-vaccine/
+        'url': 'arcpublishing:gray:0b0ba30e-032a-4598-8810-901d70e6033e',
+        'only_matching': True,
+    }, {
+        # https://www.lateja.cr/el-mundo/video-china-aprueba-con-condiciones-su-primera/dfcbfa57-527f-45ff-a69b-35fe71054143/video/
+        'url': 'arcpublishing:gruponacion:dfcbfa57-527f-45ff-a69b-35fe71054143',
+        'only_matching': True,
+    }, {
+        # https://www.fifthdomain.com/video/2018/03/09/is-america-vulnerable-to-a-cyber-attack/
+        'url': 'arcpublishing:mco:aa0ca6fe-1127-46d4-b32c-be0d6fdb8055',
+        'only_matching': True,
+    }, {
+        # https://www.vl.no/kultur/2020/12/09/en-melding-fra-en-lytter-endret-julelista-til-lewi-bergrud/
+        'url': 'arcpublishing:mentormedier:47a12084-650b-4011-bfd0-3699b6947b2d',
+        'only_matching': True,
+    }, {
+        # https://www.14news.com/2020/12/30/whiskey-theft-caught-camera-henderson-liquor-store/
+        'url': 'arcpublishing:raycom:b89f61f8-79fa-4c09-8255-e64237119bf7',
+        'only_matching': True,
+    }, {
+        # https://www.theglobeandmail.com/world/video-ethiopian-woman-who-became-symbol-of-integration-in-italy-killed-on/
+        'url': 'arcpublishing:tgam:411b34c1-8701-4036-9831-26964711664b',
+        'only_matching': True,
+    }, {
+        # https://www.pilotonline.com/460f2931-8130-4719-8ea1-ffcb2d7cb685-132.html
+        'url': 'arcpublishing:tronc:460f2931-8130-4719-8ea1-ffcb2d7cb685',
+        'only_matching': True,
+    }]
+    _POWA_DEFAULTS = [
+        (['cmg', 'prisa'], '%s-config-prod.api.cdn.arcpublishing.com/video'),
+        ([
+            'adn', 'advancelocal', 'answers', 'bonnier', 'bostonglobe', 'demo',
+            'gmg', 'gruponacion', 'infobae', 'mco', 'nzme', 'pmn', 'raycom',
+            'spectator', 'tbt', 'tgam', 'tronc', 'wapo', 'wweek',
+        ], 'video-api-cdn.%s.arcpublishing.com/api'),
+    ]
+
+    def _extract_urls(webpage):
+        entries = []
+        # https://arcpublishing.atlassian.net/wiki/spaces/POWA/overview
+        for powa_el in re.findall(r'(<div[^>]+class="[^"]*\bpowa\b[^"]*"[^>]+data-uuid="%s"[^>]*>)' % ArcPublishingIE._UUID_REGEX, webpage):
+            powa = extract_attributes(powa_el) or {}
+            org = powa.get('data-org')
+            uuid = powa.get('data-uuid')
+            if org and uuid:
+                entries.append('arcpublishing:%s:%s' % (org, uuid))
+        return entries
+
+    def _real_extract(self, url):
+        org, uuid = re.match(self._VALID_URL, url).groups()
+        for orgs, tmpl in self._POWA_DEFAULTS:
+            if org in orgs:
+                base_api_tmpl = tmpl
+                break
+        else:
+            base_api_tmpl = '%s-prod-cdn.video-api.arcpublishing.com/api'
+        if org == 'wapo':
+            org = 'washpost'
+        video = self._download_json(
+            'https://%s/v1/ansvideos/findByUuid' % (base_api_tmpl % org),
+            uuid, query={'uuid': uuid})[0]
+        title = video['headlines']['basic']
+        is_live = video.get('status') == 'live'
+
+        urls = []
+        formats = []
+        for s in video.get('streams', []):
+            s_url = s.get('url')
+            if not s_url or s_url in urls:
+                continue
+            urls.append(s_url)
+            stream_type = s.get('stream_type')
+            if stream_type == 'smil':
+                smil_formats = self._extract_smil_formats(
+                    s_url, uuid, fatal=False)
+                for f in smil_formats:
+                    if f['url'].endswith('/cfx/st'):
+                        f['app'] = 'cfx/st'
+                        if not f['play_path'].startswith('mp4:'):
+                            f['play_path'] = 'mp4:' + f['play_path']
+                        if isinstance(f['tbr'], float):
+                            f['vbr'] = f['tbr'] * 1000
+                            del f['tbr']
+                            f['format_id'] = 'rtmp-%d' % f['vbr']
+                formats.extend(smil_formats)
+            elif stream_type in ('ts', 'hls'):
+                m3u8_formats = self._extract_m3u8_formats(
+                    s_url, uuid, 'mp4', 'm3u8' if is_live else 'm3u8_native',
+                    m3u8_id='hls', fatal=False)
+                if all([f.get('acodec') == 'none' for f in m3u8_formats]):
+                    continue
+                for f in m3u8_formats:
+                    if f.get('acodec') == 'none':
+                        f['preference'] = -40
+                    elif f.get('vcodec') == 'none':
+                        f['preference'] = -50
+                    height = f.get('height')
+                    if not height:
+                        continue
+                    vbr = self._search_regex(
+                        r'[_x]%d[_-](\d+)' % height, f['url'], 'vbr', default=None)
+                    if vbr:
+                        f['vbr'] = int(vbr)
+                formats.extend(m3u8_formats)
+            else:
+                vbr = int_or_none(s.get('bitrate'))
+                formats.append({
+                    'format_id': '%s-%d' % (stream_type, vbr) if vbr else stream_type,
+                    'vbr': vbr,
+                    'width': int_or_none(s.get('width')),
+                    'height': int_or_none(s.get('height')),
+                    'filesize': int_or_none(s.get('filesize')),
+                    'url': s_url,
+                    'preference': -1,
+                })
+        self._sort_formats(
+            formats, ('preference', 'width', 'height', 'vbr', 'filesize', 'tbr', 'ext', 'format_id'))
+
+        subtitles = {}
+        for subtitle in (try_get(video, lambda x: x['subtitles']['urls'], list) or []):
+            subtitle_url = subtitle.get('url')
+            if subtitle_url:
+                subtitles.setdefault('en', []).append({'url': subtitle_url})
+
+        return {
+            'id': uuid,
+            'title': self._live_title(title) if is_live else title,
+            'thumbnail': try_get(video, lambda x: x['promo_image']['url']),
+            'description': try_get(video, lambda x: x['subheadlines']['basic']),
+            'formats': formats,
+            'duration': int_or_none(video.get('duration'), 100),
+            'timestamp': parse_iso8601(video.get('created_date')),
+            'subtitles': subtitles,
+            'is_live': is_live,
+        }
index 54cbbe9d2363239656cbb06f358fe0ae059a962f..d9d3f4940a11dfaeacdaa36405ed9dadfb0c6531 100644 (file)
@@ -56,6 +56,7 @@ from .appletrailers import (
     AppleTrailersSectionIE,
 )
 from .archiveorg import ArchiveOrgIE
+from .arcpublishing import ArcPublishingIE
 from .arkena import ArkenaIE
 from .ard import (
     ARDBetaMediathekIE,
index 14c27c6daf049b8eeff3a3889df4dc646777405d..780971a92ecc0c5b249c8ee682f36f3f954b3612 100644 (file)
@@ -127,6 +127,7 @@ from .expressen import ExpressenIE
 from .zype import ZypeIE
 from .odnoklassniki import OdnoklassnikiIE
 from .kinja import KinjaEmbedIE
+from .arcpublishing import ArcPublishingIE
 
 
 class GenericIE(InfoExtractor):
@@ -2208,6 +2209,20 @@ class GenericIE(InfoExtractor):
                 'uploader': 'OTT Videos',
             },
         },
+        {
+            # ArcPublishing PoWa video player
+            'url': 'https://www.adn.com/politics/2020/11/02/video-senate-candidates-campaign-in-anchorage-on-eve-of-election-day/',
+            'md5': 'b03b2fac8680e1e5a7cc81a5c27e71b3',
+            'info_dict': {
+                'id': '8c99cb6e-b29c-4bc9-9173-7bf9979225ab',
+                'ext': 'mp4',
+                'title': 'Senate candidates wave to voters on Anchorage streets',
+                'description': 'md5:91f51a6511f090617353dc720318b20e',
+                'timestamp': 1604378735,
+                'upload_date': '20201103',
+                'duration': 1581,
+            },
+        },
     ]
 
     def report_following_redirect(self, new_url):
@@ -2574,6 +2589,10 @@ class GenericIE(InfoExtractor):
         if tp_urls:
             return self.playlist_from_matches(tp_urls, video_id, video_title, ie='ThePlatform')
 
+        arc_urls = ArcPublishingIE._extract_urls(webpage)
+        if arc_urls:
+            return self.playlist_from_matches(arc_urls, video_id, video_title, ie=ArcPublishingIE.ie_key())
+
         # Look for embedded rtl.nl player
         matches = re.findall(
             r'<iframe[^>]+?src="((?:https?:)?//(?:(?:www|static)\.)?rtl\.nl/(?:system/videoplayer/[^"]+(?:video_)?)?embed[^"]+)"',
index 625d0a1cc14a52604f46e264a0f93342056fd9df..8afb1af831aec44831d07d61adae269d796f10c9 100644 (file)
@@ -4,17 +4,13 @@ from __future__ import unicode_literals
 import re
 
 from .common import InfoExtractor
-from ..utils import (
-    int_or_none,
-    strip_jsonp,
-)
 
 
 class WashingtonPostIE(InfoExtractor):
     IE_NAME = 'washingtonpost'
-    _VALID_URL = r'(?:washingtonpost:|https?://(?:www\.)?washingtonpost\.com/video/(?:[^/]+/)*)(?P<id>[\da-f]{8}-[\da-f]{4}-[\da-f]{4}-[\da-f]{4}-[\da-f]{12})'
+    _VALID_URL = r'(?:washingtonpost:|https?://(?:www\.)?washingtonpost\.com/(?:video|posttv)/(?:[^/]+/)*)(?P<id>[\da-f]{8}-[\da-f]{4}-[\da-f]{4}-[\da-f]{4}-[\da-f]{12})'
     _EMBED_URL = r'https?://(?:www\.)?washingtonpost\.com/video/c/embed/[\da-f]{8}-[\da-f]{4}-[\da-f]{4}-[\da-f]{4}-[\da-f]{12}'
-    _TEST = {
+    _TESTS = [{
         'url': 'https://www.washingtonpost.com/video/c/video/480ba4ee-1ec7-11e6-82c2-a7dcb313287d',
         'md5': '6f537e1334b714eb15f9563bd4b9cdfa',
         'info_dict': {
@@ -23,10 +19,15 @@ class WashingtonPostIE(InfoExtractor):
             'title': 'Egypt finds belongings, debris from plane crash',
             'description': 'md5:a17ceee432f215a5371388c1f680bd86',
             'upload_date': '20160520',
-            'uploader': 'Reuters',
-            'timestamp': 1463778452,
+            'timestamp': 1463775187,
         },
-    }
+    }, {
+        'url': 'https://www.washingtonpost.com/video/world/egypt-finds-belongings-debris-from-plane-crash/2016/05/20/480ba4ee-1ec7-11e6-82c2-a7dcb313287d_video.html',
+        'only_matching': True,
+    }, {
+        'url': 'https://www.washingtonpost.com/posttv/world/iraq-to-track-down-antiquities-after-islamic-state-museum-rampage/2015/02/28/7c57e916-bf86-11e4-9dfb-03366e719af8_video.html',
+        'only_matching': True,
+    }]
 
     @classmethod
     def _extract_urls(cls, webpage):
@@ -35,73 +36,8 @@ class WashingtonPostIE(InfoExtractor):
 
     def _real_extract(self, url):
         video_id = self._match_id(url)
-        video_data = self._download_json(
-            'http://www.washingtonpost.com/posttv/c/videojson/%s?resType=jsonp' % video_id,
-            video_id, transform_source=strip_jsonp)[0]['contentConfig']
-        title = video_data['title']
-
-        urls = []
-        formats = []
-        for s in video_data.get('streams', []):
-            s_url = s.get('url')
-            if not s_url or s_url in urls:
-                continue
-            urls.append(s_url)
-            video_type = s.get('type')
-            if video_type == 'smil':
-                continue
-            elif video_type in ('ts', 'hls') and ('_master.m3u8' in s_url or '_mobile.m3u8' in s_url):
-                m3u8_formats = self._extract_m3u8_formats(
-                    s_url, video_id, 'mp4', 'm3u8_native', m3u8_id='hls', fatal=False)
-                for m3u8_format in m3u8_formats:
-                    width = m3u8_format.get('width')
-                    if not width:
-                        continue
-                    vbr = self._search_regex(
-                        r'%d_%d_(\d+)' % (width, m3u8_format['height']), m3u8_format['url'], 'vbr', default=None)
-                    if vbr:
-                        m3u8_format.update({
-                            'vbr': int_or_none(vbr),
-                        })
-                formats.extend(m3u8_formats)
-            else:
-                width = int_or_none(s.get('width'))
-                vbr = int_or_none(s.get('bitrate'))
-                has_width = width != 0
-                formats.append({
-                    'format_id': (
-                        '%s-%d-%d' % (video_type, width, vbr)
-                        if width
-                        else video_type),
-                    'vbr': vbr if has_width else None,
-                    'width': width,
-                    'height': int_or_none(s.get('height')),
-                    'acodec': s.get('audioCodec'),
-                    'vcodec': s.get('videoCodec') if has_width else 'none',
-                    'filesize': int_or_none(s.get('fileSize')),
-                    'url': s_url,
-                    'ext': 'mp4',
-                    'protocol': 'm3u8_native' if video_type in ('ts', 'hls') else None,
-                })
-        source_media_url = video_data.get('sourceMediaURL')
-        if source_media_url:
-            formats.append({
-                'format_id': 'source_media',
-                'url': source_media_url,
-            })
-        self._sort_formats(
-            formats, ('width', 'height', 'vbr', 'filesize', 'tbr', 'format_id'))
-
-        return {
-            'id': video_id,
-            'title': title,
-            'description': video_data.get('blurb'),
-            'uploader': video_data.get('credits', {}).get('source'),
-            'formats': formats,
-            'duration': int_or_none(video_data.get('videoDuration'), 100),
-            'timestamp': int_or_none(
-                video_data.get('dateConfig', {}).get('dateFirstPublished'), 1000),
-        }
+        return self.url_result(
+            'arcpublishing:wapo:' + video_id, 'ArcPublishing', video_id)
 
 
 class WashingtonPostArticleIE(InfoExtractor):
@@ -121,9 +57,8 @@ class WashingtonPostArticleIE(InfoExtractor):
                 'title': 'Breaking Points: The Paper Mine',
                 'duration': 1290,
                 'description': 'Overly complicated paper pushing is nothing new to government bureaucracy. But the way federal retirement applications are filed may be the most outdated. David Fahrenthold explains.',
-                'uploader': 'The Washington Post',
-                'timestamp': 1395527908,
-                'upload_date': '20140322',
+                'timestamp': 1395440416,
+                'upload_date': '20140321',
             },
         }, {
             'md5': '1fff6a689d8770966df78c8cb6c8c17c',
@@ -133,9 +68,8 @@ class WashingtonPostArticleIE(InfoExtractor):
                 'title': 'The town bureaucracy sustains',
                 'description': 'Underneath the friendly town of Boyers is a sea of government paperwork. In a disused limestone mine, hundreds of locals now track, file and process retirement applications for the federal government. We set out to find out what it\'s like to do paperwork 230 feet underground.',
                 'duration': 2220,
-                'timestamp': 1395528005,
-                'upload_date': '20140322',
-                'uploader': 'The Washington Post',
+                'timestamp': 1395441819,
+                'upload_date': '20140321',
             },
         }],
     }, {
@@ -151,8 +85,7 @@ class WashingtonPostArticleIE(InfoExtractor):
                 'ext': 'mp4',
                 'description': 'Washington Post transportation reporter Ashley Halsey III explains why a plane\'s black box needs to be recovered from a crash site instead of having its information streamed in real time throughout the flight.',
                 'upload_date': '20141230',
-                'uploader': 'The Washington Post',
-                'timestamp': 1419974765,
+                'timestamp': 1419972442,
                 'title': 'Why black boxes don’t transmit data in real time',
             }
         }]