Merge 477b64b153 into 52c0ffe40a

Co-Authored-By: bashonly <88596187+bashonly@users.noreply.github.com>

yt_dlp/extractor/_extractors.py

@@ -747,6 +747,14 @@ from .gettr import (
     GettrStreamingIE,
 )
 from .giantbomb import GiantBombIE
+from .giphy import (
+    GiphyChannelIE,
+    GiphyChannelPageIE,
+    GiphyIE,
+    GiphySearchIE,
+    GiphySearchURLIE,
+    GiphyStoriesIE,
+)
 from .glide import GlideIE
 from .globalplayer import (
     GlobalPlayerAudioEpisodeIE,

yt_dlp/extractor/giphy.py

@@ -0,0 +1,529 @@
+import itertools
+import re
+
+from .common import InfoExtractor, SearchInfoExtractor
+from ..utils import (
+    determine_ext,
+    int_or_none,
+    merge_dicts,
+    str_or_none,
+    traverse_obj,
+    url_basename,
+    url_or_none,
+)
+
+
+class GiphyBaseIE(InfoExtractor):
+    _GIPHY_FE_WEB_API_KEY = 'Gc7131jiJuvI7IdN0HZ1D7nh0ow5BU6g'
+    _GIPHY_FE_STORIES_AND_GIPHY_TV_API_KEY = '3eFQvabDx69SMoOemSPiYfh9FY0nzO9x'
+    _GIPHY_MAX = 5000
+
+    def _extract_formats(self, media, is_still=False):
+        f = []
+        for format_id in media:
+            if ('_still' in format_id) == is_still:
+                for key in [k for k, v in media[format_id].items() if str(v)[:4] == 'http']:
+                    i = traverse_obj(media[format_id], {
+                        'width': ('width', {int_or_none}),
+                        'height': ('height', {int_or_none}),
+                        'url': (key, {url_or_none}),
+                    })
+                    f.append({
+                        'format_id': format_id,
+                        **i,
+                    })
+        return f
+
+    def _extract_info(self, gif_data, video_id):
+        formats, thumbnails, subtitles, uploader = [], [], {}, {}
+        # formats, thumbnails, subtitles
+        if data := gif_data.get('video'):
+            for lang in data.get('captions', {}):
+                for key in data['captions'][lang]:
+                    subtitles.setdefault(lang, []).append({'url': data['captions'][lang][key]})
+            for category in ['assets', 'previews']:
+                formats.extend(self._extract_formats(data.get(category, {})))
+            if data.get('hls_manifest_url'):
+                hls_fmts, hls_subs = self._extract_m3u8_formats_and_subtitles(
+                    data['hls_manifest_url'], video_id, 'mp4', m3u8_id='hls', fatal=False)
+                formats.extend(hls_fmts)
+                self._merge_subtitles(hls_subs, target=subtitles)
+            if data.get('dash_manifest_url'):
+                dash_fmts, dash_subs = self._extract_mpd_formats_and_subtitles(
+                    data['dash_manifest_url'], video_id, mpd_id='dash', fatal=False)
+                formats.extend(dash_fmts)
+                self._merge_subtitles(dash_subs, target=subtitles)
+        if data := gif_data.get('images'):
+            if data.get('looping'):
+                data['looping']['height'] = traverse_obj(data, ('original_mp4', 'height', {int_or_none}))
+                data['looping']['width'] = traverse_obj(data, ('original_mp4', 'width', {int_or_none}))
+            sorted_data = dict(sorted(data.items(), reverse=True))
+            formats.extend(self._extract_formats(sorted_data))
+            thumbnails.extend(self._extract_formats(data, is_still=True))
+        self._remove_duplicate_formats(formats)
+        for f in formats:
+            f.setdefault('http_headers', {})['Accept'] = 'video/*,image/*'
+        for l in subtitles:
+            for s in subtitles[l]:
+                s.setdefault('http_headers', {})['Accept'] = 'text/*'
+        for t in thumbnails:
+            t.setdefault('http_headers', {})['Accept'] = 'image/*'
+        # uploader
+        if data := gif_data.get('user'):
+            if isinstance(data, dict):
+                uploader = traverse_obj(data, {
+                    'uploader': (('display_name', 'name', 'attribution_display_name', 'username'),
+                                 {lambda x: x or gif_data.get('username')}),
+                    'uploader_id': ('username', {lambda x: x or gif_data.get('username')}),
+                    'uploader_url': (('profile_url', 'website_url'),
+                                     {lambda x: f'https://giphy.com{x}' if x and x[0] == '/' else url_or_none(x)}),
+                }, get_all=False)
+        # basic info
+        info = {
+            **traverse_obj(gif_data, {
+                'id': ('id', {lambda x: x or video_id}),
+                'title': ('title', {lambda x: x.strip() if x else ''}),
+                'description': ((None, 'video'), ('alt_text', 'description'),
+                                {lambda x: x.strip() if x and not x.startswith('Discover & share') else None}),
+                'tags': ('tags', {list}),
+                'age_limit': ('rating', {lambda x: 18 if x in ['r', 'nc-17'] else None}),
+                'upload_date': (('import_datetime', 'create_datetime'),
+                                {lambda x: x[:10].replace('-', '') if x else None}),
+            }, get_all=False),
+        }
+        return {
+            **info,
+            **{k: v for k, v in uploader.items() if v is not None},
+            'thumbnails': thumbnails,
+            'formats': formats,
+            'subtitles': subtitles,
+        }
+
+    def _api_channel_feed(self, channel_id):
+        count, offset = 0, 0
+        query_url = f'https://giphy.com/api/v4/channels/{channel_id}/feed/?offset={offset}'
+        for _ in itertools.count(1):
+            search_results = self._download_json(query_url, channel_id, fatal=False,
+                                                 note=f'Downloading feed data {offset + 1}-{offset + 25}')
+            if not search_results or not search_results.get('results'):
+                return
+            for video in search_results['results']:
+                yield {
+                    **self._extract_info(video, video['id']),
+                    'webpage_url': video['url'],
+                    'webpage_url_basename': url_basename(video['url']),
+                }
+            count += len(search_results.get('results'))
+            if count >= (int_or_none(self.get_param('playlistend')) or (self._GIPHY_MAX + 1)):
+                return
+            query_url = url_or_none(search_results.get('next')) or ''
+            offset = int(self._search_regex(r'offset=(\d+)', query_url, 'offset', default=0))
+            # offset cannot exceed 5000
+            if not query_url or offset > self._GIPHY_MAX:
+                return
+
+
+class GiphyIE(GiphyBaseIE):
+    _VALID_URL = r'https?://giphy\.com/(?:clips|gifs|stickers|embed)/(?:.+[/-])?(?P<id>[^/]+)/?$'
+    _TESTS = [{
+        'url': 'https://giphy.com/gifs/l2JIcQ4UH5SoPtMJi',
+        'info_dict': {
+            'id': 'l2JIcQ4UH5SoPtMJi',
+            'ext': 'mp4',
+            'title': 'excited slow motion GIF by Cats the Musical',
+            'tags': ['giphyupload', 'excited', 'cats', 'musical', 'flip', 'slow motion', 'somersault', 'cats musical'],
+            'thumbnail': r're:^https?://.*',
+            'upload_date': '20160125',
+            'uploader': 'Cats the Musical',
+            'uploader_id': 'catsmusical',
+            'uploader_url': 'https://giphy.com/catsmusical/',
+        },
+    }, {
+        'url': 'http://giphy.com/gifs/l3vR8BKU0m8uX2mAg',
+        'info_dict': {
+            'id': 'l3vR8BKU0m8uX2mAg',
+            'ext': 'mp4',
+            'title': 'Giphy video #l3vR8BKU0m8uX2mAg',
+            'tags': ['giphyupload'],
+            'thumbnail': r're:^https?://.*',
+            'upload_date': '20161022',
+            'uploader': 'gus123',
+            'uploader_id': 'gus123',
+            'uploader_url': 'https://giphy.com/channel/gus123/',
+        },
+    }, {
+        'url': 'https://giphy.com/gifs/digitalpratik-digital-pratik-happy-fathers-day-dad-E1trcBzr59SGvmRDPY',
+        'info_dict': {
+            'id': 'E1trcBzr59SGvmRDPY',
+            'ext': 'mp4',
+            'title': 'Happy Fathers Day GIF by Digital Pratik',
+            'tags': 'count:14',
+            'thumbnail': r're:^https?://.*',
+            'upload_date': '20210619',
+            'uploader': 'Digital Pratik',
+            'uploader_id': 'digitalpratik',
+            'uploader_url': 'https://giphy.com/digitalpratik/',
+        },
+    }, {
+        'url': 'https://giphy.com/clips/southpark-south-park-episode-4-season-20-YyOPrvilA8FdiuSiQi',
+        'info_dict': {
+            'id': 'YyOPrvilA8FdiuSiQi',
+            'ext': 'mp4',
+            'title': 'You Can\'t Break Up With Me',
+            'description': 'South Park, Season 20, Episode 4, Wieners Out',
+            'tags': 'count:17',
+            'thumbnail': r're:^https?://.*',
+            'upload_date': '20220516',
+            'uploader': 'South Park',
+            'uploader_id': 'southpark',
+            'uploader_url': 'https://giphy.com/southpark/',
+        },
+    }, {
+        'url': 'https://giphy.com/stickers/mario-PFxFYEZNUavG8',
+        'info_dict': {
+            'id': 'PFxFYEZNUavG8',
+            'ext': 'mp4',
+            'title': 'nintendo mario STICKER',
+            'tags': ['transparent', 'gaming', 'nintendo', 'mario', 'giphynintendos'],
+            'thumbnail': r're:^https?://.*',
+            'upload_date': '20160908',
+        },
+    }, {
+        'url': 'https://giphy.com/embed/00xGP4zv8xENZ2tc3Y',
+        'info_dict': {
+            'id': '00xGP4zv8xENZ2tc3Y',
+            'ext': 'mp4',
+            'title': 'Love Is Blind Wow GIF by NETFLIX',
+            'description': 'md5:89445e21c848eef12af249faef4fcf9f',
+            'tags': 'count:24',
+            'thumbnail': r're:^https?://.*',
+            'upload_date': '20220214',
+            'uploader': 'NETFLIX',
+            'uploader_id': 'netflix',
+            'uploader_url': 'https://giphy.com/netflix/',
+        },
+    }]
+
+    def _real_extract(self, url):
+        video_id = self._match_id(url)
+        webpage = self._download_webpage(url.replace('/embed/', '/gifs/'), video_id)
+
+        title = (self._html_search_meta('twitter:title', webpage, default=None)
+                 or self._og_search_title(webpage).replace(' - Find & Share on GIPHY', '').strip())
+        description = (self._html_search_meta('twitter:description', webpage, default=None)
+                       or self._og_search_description(webpage))
+        description = description if not description.startswith('Discover & share') else None
+
+        gif_data = {}
+        # search for:  \"gif\":{\"type\":\"...},
+        if json_str := self._html_search_regex(r'\\"\w+\\":({\\"type\\":\\"(?!emoji).*?is_dynamic\\":\w+}),',
+                                               webpage, 'video_data', default=None):
+            gif_data = self._parse_json(json_str.replace(r'\"', '"'), video_id)
+        # search for:  gif: {"...},
+        elif json_str := self._html_search_regex(r'\s+\w+:\s*({".*?}),\n\s+', webpage, 'video_data', default='{}'):
+            gif_data = self._parse_json(json_str, video_id)
+
+        info = self._extract_info(gif_data, video_id)
+
+        if not info.get('formats'):
+            formats = []
+            if url := self._og_search_video_url(webpage, default=None):
+                formats.append({
+                    'format_id': determine_ext(url),
+                    'width': int_or_none(self._og_search_property('video:width', webpage)),
+                    'height': int_or_none(self._og_search_property('video:height', webpage)),
+                    'url': url,
+                })
+            if url := self._og_search_thumbnail(webpage, default=None):
+                formats.append({
+                    'format_id': determine_ext(url),
+                    'width': int_or_none(self._og_search_property('image:width', webpage)),
+                    'height': int_or_none(self._og_search_property('image:height', webpage)),
+                    'url': url,
+                })
+            if url := self._html_search_meta('twitter:image', webpage, default=None):
+                thumbnails = [{
+                    'width': int_or_none(self._html_search_meta('twitter:image:width', webpage, default=None)),
+                    'height': int_or_none(self._html_search_meta('twitter:image:height', webpage, default=None)),
+                    'url': url,
+                }]
+            info['formats'] = formats
+            if not info.get('thumbnails'):
+                info['thumbnails'] = thumbnails
+
+        if not info.get('uploader'):
+            uploader = {}
+            if data := gif_data.get('user'):
+                if isinstance(data, str):
+                    idx = data.replace('$', '')
+                    if json_str := self._html_search_regex(rf'\\n{idx}:({{.*?}})\\n\w+:',
+                                                           webpage, 'uploader_data', default=None):
+                        json_str = re.sub(r'"\]\)self\.__next_f\.push\(\[\d+,"', '', json_str).replace(r'\"', '"')
+                        data = self._parse_json(json_str, video_id, fatal=False)
+                    if isinstance(data, dict):
+                        uploader = traverse_obj(data, {
+                            'uploader': (('display_name', 'name', 'attribution_display_name', 'username'),
+                                         {lambda x: x or gif_data.get('username')}),
+                            'uploader_id': ('username', {str_or_none}),
+                            'uploader_url': (('profile_url', 'website_url'),
+                                             {lambda x: f'https://giphy.com{x}' if x and x[0] == '/' else url_or_none(x)}),
+                        }, get_all=False)
+            if not uploader:
+                up_id = (gif_data.get('username')
+                         or self._html_search_regex(r'<div>@(\w+)</div>', webpage, 'uploader_id', default=None)
+                         or self._html_search_regex(r'"woff2"/><link[^>]+\.giphy\.com/(?:channel_assets|avatars)/(.+?)/',
+                                                    webpage, 'uploader_id', default=None))
+                up_name = (title[(title.rfind(' by ') + 4):] if title.rfind(' by ') > 0 else None
+                           or self._html_search_regex(r'(?s)<h2\b[^>]*>([^<]+)</h2>', webpage, 'uploader', default=None)
+                           or self._html_search_regex(r'twitter:creator"[^>]+="((?!@giphy").*?)"', webpage, 'uploader', default=None)
+                           or up_id)
+                uploader = {
+                    'uploader': up_name,
+                    'uploader_id': up_id,
+                    'uploader_url': (f'https://giphy.com/channel/{up_id}/' if up_id else None),
+                }
+            info = merge_dicts(info, {**{k: v for k, v in uploader.items() if v is not None}})
+
+        return {
+            **merge_dicts(info, {
+                'title': title,
+                'description': description,
+            }),
+        }
+
+
+class GiphyChannelPageIE(GiphyBaseIE):
+    _VALID_URL = r'https?://giphy\.com/(?!(?:clips|gifs|stickers|stories|search|embed)/)(?:.+/)?(?P<id>[^/]+)/?$'
+    _TESTS = [{
+        'url': 'https://giphy.com/catsmusical/',
+        'playlist_count': 10,
+        'info_dict': {
+            'id': '8707',
+            'title': 'Cats the Musical',
+            'uploader_id': 'catsmusical',
+            'uploader_url': 'https://giphy.com/channel/catsmusical',
+        },
+    }, {
+        'url': 'https://giphy.com/channel/catsmusical',
+        'playlist_count': 10,
+        'info_dict': {
+            'id': '8707',
+            'title': 'Cats the Musical',
+            'uploader_id': 'catsmusical',
+            'uploader_url': 'https://giphy.com/channel/catsmusical',
+        },
+    }, {
+        'url': 'https://giphy.com/southpark/reactions/lol',
+        'playlist_count': 42,
+        'info_dict': {
+            'id': '1044',
+            'title': 'LOL',
+            'uploader_id': 'southpark',
+            'uploader_url': 'https://giphy.com/channel/southpark',
+        },
+    }, {
+        'url': 'https://giphy.com/corgiyolk/cute-and-wholesome-corgi',
+        'playlist_count': 14,
+        'info_dict': {
+            'id': '34458076',
+            'title': 'Cute and Wholesome corgi',
+            'uploader_id': 'corgiyolk',
+            'uploader_url': 'https://giphy.com/channel/corgiyolk',
+        },
+    }]
+
+    def _real_extract(self, url):
+        video_id = self._match_id(url)
+        webpage = self._download_webpage(url, video_id)
+
+        # search for:  {"channelId": ...} or {..., "channel_id": ...,
+        if channel_id := self._html_search_regex(r'\{[^\}\n]*"channel_?[iI]d":\s*"?([^",\}]+)[",\}]',
+                                                 webpage, 'channel_id', default=None):
+            uploader_id = self._html_search_meta('twitter:creator', webpage).replace('@', '').lower()
+            entries = []
+            for i in self._api_channel_feed(channel_id):
+                entries.append(i)
+
+            return {
+                'id': channel_id,
+                'title': (self._html_search_meta('twitter:title', webpage)
+                          or self._og_search_title(webpage)
+                          ).replace(' GIFs on GIPHY - Be Animated', '').strip(),
+                'uploader_id': uploader_id,
+                'uploader_url': f'https://giphy.com/channel/{uploader_id}' if uploader_id != 'giphy' else None,
+                '_type': 'playlist',
+                'entries': entries,
+            }
+
+
+class GiphyChannelIE(GiphyBaseIE, SearchInfoExtractor):
+    IE_NAME = 'giphy:channel'
+    IE_DESC = 'Giphy Channel'
+    _SEARCH_KEY = 'giphychannel'
+    _TESTS = [{
+        'url': 'giphychannel30:pbsnature',
+        'playlist_count': 30,
+        'info_dict': {
+            'id': 'pbsnature',
+            'title': 'pbsnature',
+        },
+    }]
+
+    def _search_results(self, query):
+        if webpage := self._download_webpage(f'https://giphy.com/channel/{query}', query):
+            if channel_id := self._html_search_regex(r'\{[^\}\n]*"channel_?[iI]d":\s*"?([^",\}]+)[",\}]',
+                                                     webpage, 'channel_id', default=None):
+                return self._api_channel_feed(channel_id)
+
+
+class GiphySearchIE(GiphyBaseIE, SearchInfoExtractor):
+    IE_NAME = 'giphy:search'
+    IE_DESC = 'Giphy Search'
+    _SEARCH_KEY = 'giphysearch'
+    _TESTS = [{
+        'url': 'giphysearch20:super mario',
+        'playlist_count': 20,
+        'info_dict': {
+            'id': 'super mario',
+            'title': 'super mario',
+        },
+    }, {
+        'url': 'giphysearch40:mickey&type=clips,stickers',
+        'playlist_count': 40,
+        'info_dict': {
+            'id': 'mickey&type=clips,stickers',
+            'title': 'mickey&type=clips,stickers',
+        },
+    }]
+
+    def _search_results(self, query):
+        def search_query(query, offset, limit, category):
+            # search api:
+            # https://api.giphy.com/v1/gifs/search?rating=pg-13&offset=40&limit=15&type=gifs&q={query}&excludeDynamicResults=undefined&api_key=Gc7131jiJuvI7IdN0HZ1D7nh0ow5BU6g&pingback_id=1904d6e524cee33d
+            return self._download_json(
+                f'https://api.giphy.com/v1/{category}/search', query,
+                note=f'Downloading {category} result {offset + 1}-{offset + limit}', query={
+                    'rating': 'r',      # MPA film rating
+                    'offset': offset,
+                    'limit': limit,
+                    'type': category,   # known types: 'clips', 'gifs', 'stickers', 'text', 'videos'
+                    'q': query,
+                    'excludeDynamicResults': 'undefined',
+                    'api_key': self._GIPHY_FE_WEB_API_KEY,
+                })
+
+        # type: comma delimited list
+        types = self._search_regex(r'&type=([^&]+)', query, 'type', default='gifs,stickers,videos')
+        types = [(f'{x}s' if x[-1] != 's' and any(x in t for t in ['clips', 'gifs', 'stickers', 'videos']) else x)
+                 for x in [x.strip() for x in types.lower().split(',')]]
+        query = query.split('&type=')[0]
+
+        offset, limit = 0, 50
+        types_done = []
+        for _ in itertools.count(1):
+            for t in types:
+                if t not in types_done:
+                    search_type = 'videos' if t == 'clips' else t       # clips use 'videos' type
+                    search_results = search_query(query, offset, limit, search_type)
+                    if not search_results.get('data'):
+                        self.to_screen(f'{query}: {offset} {t} found')
+                        types_done.append(t)
+                    else:
+                        for video in search_results['data']:
+                            yield {
+                                **self._extract_info(video, video['id']),
+                                'webpage_url': video['url'],
+                                'webpage_url_basename': url_basename(video['url']),
+                            }
+                        if len(search_results['data']) < limit:
+                            total = offset + len(search_results['data'])
+                            self.to_screen(f'{query}: {total} {t} found')
+                            types_done.append(t)
+            if len(types) > len(types_done):
+                offset += limit
+                if offset >= self._GIPHY_MAX:
+                    return
+            else:
+                return
+
+
+class GiphySearchURLIE(GiphyBaseIE):
+    _VALID_URL = r'https?://giphy\.com/search/(?P<id>[^/]+)'
+    _TESTS = [{
+        'url': 'https://giphy.com/search/carcinoma',
+        'playlist_count': 13,
+        'info_dict': {
+            'id': 'carcinoma',
+            'title': 'carcinoma',
+        },
+    }]
+
+    def _real_extract(self, url):
+        if query := self._match_id(url):
+            query = query.replace('-', ' ')
+            return self.url_result(url=f'giphysearchall:{query}', url_transparent=True)
+
+
+class GiphyStoriesIE(GiphyBaseIE):
+    _VALID_URL = r'https?://giphy\.com/stories/(?P<id>[^/]+)'
+    _TESTS = [{
+        'url': 'https://giphy.com/stories/indian-garyvee-de56ee78-adb5',
+        'playlist_count': 6,
+        'info_dict': {
+            'id': 'de56ee78-adb5-428e-afb6-765bb489c4fc',
+            'title': 'Indian Garyvee',
+            'description': 'md5:468a8e2c898225baac05e32dcafe313b',
+            'tags': [],
+            'upload_date': '20190710',
+            'uploader': 'Digital Pratik',
+            'uploader_id': 'digitalpratik',
+            'uploader_url': 'https://giphy.com/digitalpratik',
+        },
+    }, {
+        'url': 'https://giphy.com/stories/a-2021-oscars-wrap-up-161cda9a-4517',
+        'playlist_count': 10,
+        'info_dict': {
+            'id': '161cda9a-4517-473b-9fea-6731af6c0d49',
+            'title': 'A 2021 Oscars Wrap-Up',
+            'description': 'What were the best moments in GIFs? Glad you asked!',
+            'tags': [],
+            'upload_date': '20210426',
+            'uploader': 'Entertainment GIFs',
+            'uploader_id': 'entertainment',
+            'uploader_url': 'https://giphy.com/entertainment',
+        },
+    }]
+
+    def _real_extract(self, url):
+        slug = self._match_id(url)
+        # https://x.giphy.com/v1/stories/slug/{slug}?api_key=3eFQvabDx69SMoOemSPiYfh9FY0nzO9x
+        story = self._download_json(f'https://x.giphy.com/v1/stories/slug/{slug}?api_key={self._GIPHY_FE_STORIES_AND_GIPHY_TV_API_KEY}', slug)
+
+        if data := story.get('data'):
+            entries = []
+            for video in data.get('gifs'):
+                entries.append({
+                    **self._extract_info(video['gif'], video['gif']['id']),
+                    'description': video['caption'],
+                    'webpage_url': video['gif']['url'],
+                    'webpage_url_basename': url_basename(video['gif']['url']),
+                })
+            info = traverse_obj(data, {
+                'id': ('story_id', {lambda x: x or slug}),
+                'title': ('title', {str_or_none}),
+                'description': ('description', {str_or_none}),
+                'tags': ('tags', {list}),
+                'thumbnails': ('cover_gif', 'gif', 'images', {dict}, {lambda x: self._extract_formats(x, is_still=True)}),
+                'upload_date': (('create_datetime', 'publish_datetime'),
+                                {lambda x: x[:10].replace('-', '') if x else None}),
+                'uploader': ('user', ('display_name', 'username'), {str_or_none}),
+                'uploader_id': ('user', 'username', {str_or_none}),
+                'uploader_url': ('user', ('profile_url', 'website_url'),
+                                 {lambda x: f'https://giphy.com{x}' if x and x[0] == '/' else url_or_none(x)}),
+            }, get_all=False)
+
+            return {
+                **info,
+                '_type': 'playlist',
+                'entries': entries,
+            }