[InfoExtractor] Add search methods for Next/Nuxt.js from yt-dlp

* add _search_nextjs_data(), from https://github.com/yt-dlp/yt-dlp/pull/1386 thanks selfisekai * add _search_nuxt_data(), from https://github.com/yt-dlp/yt-dlp/pull/1921, thanks Lesmiscore, pukkandan * add tests for the above * also fix HTML5 type recognition and tests, from 222a230871, thanks Lesmiscore * update extractors in PR using above, fix tests.
2024-12-22 00:37:41 +00:00 · 2023-05-05 19:25:42 +01:00 · 2023-05-05 19:25:42 +01:00 · b2741f2654
commit b2741f2654
parent 8465222041
5 changed files with 168 additions and 58 deletions
--- a/test/test_InfoExtractor.py
+++ b/test/test_InfoExtractor.py
@ -7,15 +7,33 @@ import io
 import os
 import sys
 import unittest
+
 sys.path.insert(0, os.path.dirname(os.path.dirname(os.path.abspath(__file__))))

-from test.helper import FakeYDL, expect_dict, expect_value, http_server_port
-from youtube_dl.compat import compat_etree_fromstring, compat_http_server
-from youtube_dl.extractor.common import InfoExtractor
-from youtube_dl.extractor import YoutubeIE, get_info_extractor
-from youtube_dl.utils import encode_data_uri, strip_jsonp, ExtractorError, RegexNotFoundError
 import threading

+from test.helper import (
+    expect_dict,
+    expect_value,
+    FakeYDL,
+    http_server_port,
+)
+from youtube_dl.compat import (
+    compat_etree_fromstring,
+    compat_http_server,
+)
+from youtube_dl.extractor.common import InfoExtractor
+from youtube_dl.extractor import (
+    get_info_extractor,
+    YoutubeIE,
+)
+from youtube_dl.utils import (
+    encode_data_uri,
+    ExtractorError,
+    RegexNotFoundError,
+    strip_jsonp,
+)
+

 TEAPOT_RESPONSE_STATUS = 418
 TEAPOT_RESPONSE_BODY = "<h1>418 I'm a teapot</h1>"
@ -100,6 +118,71 @@ class TestInfoExtractor(unittest.TestCase):
        self.assertRaises(RegexNotFoundError, ie._html_search_meta, 'z', html, None, fatal=True)
        self.assertRaises(RegexNotFoundError, ie._html_search_meta, ('z', 'x'), html, None, fatal=True)

+    def test_search_nextjs_data(self):
+        html = '''
+<!DOCTYPE html>
+<html>
+<head>
+  <meta http-equiv="content-type" content=
+  "text/html; charset=utf-8">
+  <meta name="viewport" content="width=device-width">
+  <title>Test _search_nextjs_data()</title>
+</head>
+<body>
+  <div id="__next">
+    <div style="background-color:#17171E" class="FU" dir="ltr">
+      <div class="sc-93de261d-0 dyzzYE">
+        <div>
+          <header class="HD"></header>
+          <main class="MN">
+            <div style="height:0" class="HT0">
+              <div style="width:NaN%" data-testid=
+              "stream-container" class="WDN"></div>
+            </div>
+          </main>
+        </div>
+        <footer class="sc-6e5faf91-0 dEGaHS"></footer>
+      </div>
+    </div>
+  </div>
+  <script id="__NEXT_DATA__" type="application/json">
+  {"props":{"pageProps":{"video":{"id":"testid"}}}}
+  </script>
+</body>
+</html>
+'''
+        search = self.ie._search_nextjs_data(html, 'testID')
+        self.assertEqual(search['props']['pageProps']['video']['id'], 'testid')
+
+    def test_search_nuxt_data(self):
+        html = '''
+<!DOCTYPE html>
+<html>
+<head>
+  <meta http-equiv="content-type" content=
+  "text/html; charset=utf-8">
+  <title>Nuxt.js Test Page</title>
+  <meta name="viewport" content=
+  "width=device-width, initial-scale=1">
+  <meta data-hid="robots" name="robots" content="all">
+</head>
+<body class="BD">
+  <div id="__layout">
+    <h1 class="H1">Example heading</h1>
+    <div class="IN">
+      <p>Decoy text</p>
+    </div>
+  </div>
+  <script>
+  window.__NUXT__=(function(a,b,c,d,e,f,g,h){return {decoy:" default",data:[{track:{id:f,title:g}}]}}(null,null,"c",null,null,"testid","Nuxt.js title",null));
+  </script>
+  <script src="/_nuxt/a12345b.js" defer="defer"></script>
+</body>
+</html>
+'''
+        search = self.ie._search_nuxt_data(html, 'testID')
+        self.assertEqual(search['track']['id'], 'testid')
+
    def test_search_json_ld_realworld(self):
        # https://github.com/ytdl-org/youtube-dl/issues/23306
        expect_dict(
@ -348,6 +431,24 @@ class TestInfoExtractor(unittest.TestCase):
                }],
            })

+        # from https://0000.studio/
+        # with type attribute but without extension in URL
+        expect_dict(
+            self,
+            self.ie._parse_html5_media_entries(
+                'https://0000.studio',
+                r'''
+                <video src="https://d1ggyt9m8pwf3g.cloudfront.net/protected/ap-northeast-1:1864af40-28d5-492b-b739-b32314b1a527/archive/clip/838db6a7-8973-4cd6-840d-8517e4093c92"
+                    controls="controls" type="video/mp4" preload="metadata" autoplay="autoplay" playsinline class="object-contain">
+                </video>
+                ''', None)[0],
+            {
+                'formats': [{
+                    'url': 'https://d1ggyt9m8pwf3g.cloudfront.net/protected/ap-northeast-1:1864af40-28d5-492b-b739-b32314b1a527/archive/clip/838db6a7-8973-4cd6-840d-8517e4093c92',
+                    'ext': 'mp4',
+                }],
+            })
+
    def test_extract_jwplayer_data_realworld(self):
        # from http://www.suffolk.edu/sjc/
        expect_dict(
--- a/youtube_dl/extractor/clipchamp.py
+++ b/youtube_dl/extractor/clipchamp.py
@ -35,13 +35,6 @@ class ClipchampIE(InfoExtractor):
    _STREAM_URL_TMPL = 'https://%s.cloudflarestream.com/%s/manifest/video.%s'
    _STREAM_URL_QUERY = {'parentOrigin': 'https://clipchamp.com'}

-    def _search_nextjs_data(self, webpage, video_id, **kw):
-        return self._parse_json(
-            self._search_regex(
-                r'(?s)<script[^>]+id=[\'"]__NEXT_DATA__[\'"][^>]*>([^<]+)</script>',
-                webpage, 'next.js data', **kw),
-            video_id, **kw)
-
    def _real_extract(self, url):
        video_id = self._match_id(url)
        webpage = self._download_webpage(url, video_id)
--- a/youtube_dl/extractor/common.py
+++ b/youtube_dl/extractor/common.py
@ -3,6 +3,7 @@ from __future__ import unicode_literals

 import base64
 import datetime
+import functools
 import hashlib
 import json
 import netrc
@ -23,6 +24,7 @@ from ..compat import (
    compat_getpass,
    compat_integer_types,
    compat_http_client,
+    compat_map as map,
    compat_os_name,
    compat_str,
    compat_urllib_error,
@ -31,6 +33,7 @@ from ..compat import (
    compat_urllib_request,
    compat_urlparse,
    compat_xml_parse_error,
+    compat_zip as zip,
 )
 from ..downloader.f4m import (
    get_base_url,
@ -70,6 +73,7 @@ from ..utils import (
    str_or_none,
    str_to_int,
    strip_or_none,
+    traverse_obj,
    try_get,
    unescapeHTML,
    unified_strdate,
@ -1349,6 +1353,44 @@ class InfoExtractor(object):
                    break
        return dict((k, v) for k, v in info.items() if v is not None)

+    def _search_nextjs_data(self, webpage, video_id, **kw):
+        nkw = dict((k, v) for k, v in kw.items() if k in ('transform_source', 'fatal'))
+        kw.pop('transform_source', None)
+        next_data = self._search_regex(
+            r'''<script[^>]+\bid\s*=\s*('|")__NEXT_DATA__\1[^>]*>(?P<nd>[^<]+)</script>''',
+            webpage, 'next.js data', group='nd', **kw)
+        if not next_data:
+            return {}
+        return self._parse_json(next_data, video_id, **nkw)
+
+    def _search_nuxt_data(self, webpage, video_id, *args, **kwargs):
+        """Parses Nuxt.js metadata. This works as long as the function __NUXT__ invokes is a pure function"""
+
+        # self, webpage, video_id, context_name='__NUXT__', *, fatal=True, traverse=('data', 0)
+        context_name = args[0] if len(args) > 0 else kwargs.get('context_name', '__NUXT__')
+        fatal = kwargs.get('fatal', True)
+        traverse = kwargs.get('traverse', ('data', 0))
+
+        re_ctx = re.escape(context_name)
+
+        FUNCTION_RE = (r'\(\s*function\s*\((?P<arg_keys>[\s\S]*?)\)\s*\{\s*'
+                       r'return\s+(?P<js>\{[\s\S]*?})\s*;?\s*}\s*\((?P<arg_vals>[\s\S]*?)\)')
+
+        js, arg_keys, arg_vals = self._search_regex(
+            (p.format(re_ctx, FUNCTION_RE) for p in
+             (r'<script>\s*window\s*\.\s*{0}\s*=\s*{1}\s*\)\s*;?\s*</script>',
+              r'{0}\s*\([\s\S]*?{1}')),
+            webpage, context_name, group=('js', 'arg_keys', 'arg_vals'),
+            default=NO_DEFAULT if fatal else (None, None, None))
+        if js is None:
+            return {}
+
+        args = dict(zip(arg_keys.split(','), map(json.dumps, self._parse_json(
+            '[{0}]'.format(arg_vals), video_id, transform_source=js_to_json, fatal=fatal) or ())))
+
+        ret = self._parse_json(js, video_id, transform_source=functools.partial(js_to_json, vars=args), fatal=fatal)
+        return traverse_obj(ret, traverse) or {}
+
    @staticmethod
    def _hidden_inputs(html):
        html = re.sub(r'<!--(?:(?!<!--).)*-->', '', html)
@ -2496,7 +2538,8 @@ class InfoExtractor(object):
                return f
            return {}

-        def _media_formats(src, cur_media_type, type_info={}):
+        def _media_formats(src, cur_media_type, type_info=None):
+            type_info = type_info or {}
            full_url = absolute_url(src)
            ext = type_info.get('ext') or determine_ext(full_url)
            if ext == 'm3u8':
@ -2514,6 +2557,7 @@ class InfoExtractor(object):
                formats = [{
                    'url': full_url,
                    'vcodec': 'none' if cur_media_type == 'audio' else None,
+                    'ext': ext,
                }]
            return is_plain_url, formats

@ -2522,7 +2566,7 @@ class InfoExtractor(object):
        # so we wll include them right here (see
        # https://www.ampproject.org/docs/reference/components/amp-video)
        # For dl8-* tags see https://delight-vr.com/documentation/dl8-video/
-        _MEDIA_TAG_NAME_RE = r'(?:(?:amp|dl8(?:-live)?)-)?(video|audio)'
+        _MEDIA_TAG_NAME_RE = r'(?:(?:amp|dl8(?:-live)?)-)?(video(?:-js)?|audio)'
        media_tags = [(media_tag, media_tag_name, media_type, '')
                      for media_tag, media_tag_name, media_type
                      in re.findall(r'(?s)(<(%s)[^>]*/>)' % _MEDIA_TAG_NAME_RE, webpage)]
@ -2540,7 +2584,8 @@ class InfoExtractor(object):
            media_attributes = extract_attributes(media_tag)
            src = strip_or_none(media_attributes.get('src'))
            if src:
-                _, formats = _media_formats(src, media_type)
+                f = parse_content_type(media_attributes.get('type'))
+                _, formats = _media_formats(src, media_type, f)
                media_info['formats'].extend(formats)
            media_info['thumbnail'] = absolute_url(media_attributes.get('poster'))
            if media_content:
--- a/youtube_dl/extractor/globalplayer.py
+++ b/youtube_dl/extractor/globalplayer.py
@ -24,13 +24,6 @@ class GlobalPlayerBaseIE(InfoExtractor):
    def _match_valid_url(cls, url):
        return cls.re.match(cls._VALID_URL, url)

-    def _search_nextjs_data(self, webpage, video_id, **kw):
-        return self._parse_json(
-            self._search_regex(
-                r'(?s)<script[^>]+id=[\'"]__NEXT_DATA__[\'"][^>]*>([^<]+)</script>',
-                webpage, 'next.js data', **kw),
-            video_id, **kw)
-
    def _get_page_props(self, url, video_id):
        webpage = self._download_webpage(url, video_id)
        return self._search_nextjs_data(webpage, video_id)['props']['pageProps']
@ -39,13 +32,14 @@ class GlobalPlayerBaseIE(InfoExtractor):
        return urlhandle_detect_ext(self._request_webpage(  # Server rejects HEAD requests
            url, video_id, note='Determining source extension'))

-    def _extract_audio(self, episode, series):
+    @staticmethod
+    def _clean_desc(x):
+        x = clean_html(x)
+        if x:
+            x = x.replace('\xa0', ' ')
+        return x

-        def clean_desc(x):
-            x = clean_html(x)
-            if x:
-                x = x.replace('\xa0', ' ')
-            return x
+    def _extract_audio(self, episode, series):

        return merge_dicts({
            'vcodec': 'none',
@ -56,7 +50,7 @@ class GlobalPlayerBaseIE(InfoExtractor):
            'uploader': 'itunesAuthor',  # podcasts only
        }), traverse_obj(episode, {
            'id': 'id',
-            'description': ('description', T(clean_desc)),
+            'description': ('description', T(self._clean_desc)),
            'duration': ('duration', T(parse_duration)),
            'thumbnail': 'imageUrl',
            'url': 'streamUrl',
@ -141,9 +135,9 @@ class GlobalPlayerLivePlaylistIE(GlobalPlayerBaseIE):
            'ext': 'aac',
            # 'live_status': 'is_live',
            'is_live': True,
-            'description': 'md5:e10f5e10b01a7f2c14ba815509fbb38d',
+            'description': r're:(?s).+\bclassical\b.+\bClassic FM Hall [oO]f Fame\b',
            'thumbnail': 'https://images.globalplayer.com/images/551379?width=450&signature=oMLPZIoi5_dBSHnTMREW0Xg76mA=',
-            'title': 're:^Classic FM Hall of Fame.+$'
+            'title': 're:Classic FM Hall of Fame.+$'
        },
    }]

@ -160,7 +154,7 @@ class GlobalPlayerLivePlaylistIE(GlobalPlayerBaseIE):
            'is_live': True,
        }, traverse_obj(station, {
            'title': 'title',
-            'description': 'description',
+            'description': ('description', T(self._clean_desc)),
            'thumbnail': 'image',
        }), rev=True)

@ -177,7 +171,7 @@ class GlobalPlayerAudioIE(GlobalPlayerBaseIE):
            'thumbnail': 'md5:60286e7d12d795bd1bbc9efc6cee643e',
            'categories': ['Society & Culture', 'True Crime'],
            'uploader': 'Global',
-            'description': 'md5:da5b918eac9ae319454a10a563afacf9',
+            'description': r're:(?s).+\bscam\b.+?\bseries available now\b',
        },
    }, {
        # radio catchup
@ -203,7 +197,7 @@ class GlobalPlayerAudioIE(GlobalPlayerBaseIE):
                        series, ('episodes', lambda _, v: v['id'] and v['streamUrl']))],
            'categories': traverse_obj(series, ('categories', Ellipsis, 'name')) or None,
        }, traverse_obj(series, {
-            'description': 'description',
+            'description': ('description', T(self._clean_desc)),
            'thumbnail': 'imageUrl',
            'title': 'title',
            'uploader': 'itunesAuthor',  # podcasts only
--- a/youtube_dl/extractor/whyp.py
+++ b/youtube_dl/extractor/whyp.py
@ -21,7 +21,7 @@ class WhypIE(InfoExtractor):
            'url': 'https://cdn.whyp.it/50eb17cc-e9ff-4e18-b89b-dc9206a95cb1.mp3',
            'id': '18337',
            'title': 'Home Page Example Track',
-            'description': 'md5:bd758000fb93f3159339c852b5b9133c',
+            'description': r're:(?s).+\bexample track\b',
            'ext': 'mp3',
            'duration': 52.82,
            'uploader': 'Brad',
@ -33,29 +33,6 @@ class WhypIE(InfoExtractor):
        'only_matching': True,
    }]

-    def _search_nuxt_data(self, webpage, video_id, context_name='__NUXT__', fatal=True, traverse=('data', 0)):
-        """Parses Nuxt.js metadata. This works as long as the function __NUXT__ invokes is a pure function"""
-
-        import functools
-        import json
-        import re
-        from ..utils import (js_to_json, NO_DEFAULT)
-
-        re_ctx = re.escape(context_name)
-        FUNCTION_RE = r'\(function\((?P<arg_keys>.*?)\){return\s+(?P<js>{.*?})\s*;?\s*}\((?P<arg_vals>.*?)\)'
-        js, arg_keys, arg_vals = self._search_regex(
-            (p.format(re_ctx, FUNCTION_RE) for p in (r'<script>\s*window\.{0}={1}\s*\)\s*;?\s*</script>', r'{0}\(.*?{1}')),
-            webpage, context_name, group=('js', 'arg_keys', 'arg_vals'),
-            default=NO_DEFAULT if fatal else (None, None, None))
-        if js is None:
-            return {}
-
-        args = dict(zip(arg_keys.split(','), map(json.dumps, self._parse_json(
-            '[{0}]'.format(arg_vals), video_id, transform_source=js_to_json, fatal=fatal) or ())))
-
-        ret = self._parse_json(js, video_id, transform_source=functools.partial(js_to_json, vars=args), fatal=fatal)
-        return traverse_obj(ret, traverse) or {}
-
    def _real_extract(self, url):
        unique_id = self._match_id(url)
        webpage = self._download_webpage(url, unique_id)