[Nuvid] Fix extractor (#1022)

Fixes: https://github.com/ytdl-org/youtube-dl/issues/29886
Authored by: u-spec-png

yt_dlp/extractor/nuvid.py

@@ -1,71 +1,73 @@
+# coding: utf-8
 from __future__ import unicode_literals
 
-import re
-
 from .common import InfoExtractor
 from ..utils import (
     parse_duration,
+    int_or_none,
+    try_get,
 )
 
 
 class NuvidIE(InfoExtractor):
     _VALID_URL = r'https?://(?:www|m)\.nuvid\.com/video/(?P<id>[0-9]+)'
-    _TEST = {
-        'url': 'http://m.nuvid.com/video/1310741/',
-        'md5': 'eab207b7ac4fccfb4e23c86201f11277',
+    _TESTS = [{
+        'url': 'https://www.nuvid.com/video/6513023/italian-babe',
+        'md5': '772d2f8288f3d3c5c45f7a41761c7844',
         'info_dict': {
-            'id': '1310741',
+            'id': '6513023',
             'ext': 'mp4',
-            'title': 'Horny babes show their awesome bodeis and',
-            'duration': 129,
+            'title': 'italian babe',
+            'duration': 321.0,
             'age_limit': 18,
         }
+    }, {
+        'url': 'https://m.nuvid.com/video/6523263',
+        'info_dict': {
+            'id': '6523263',
+            'ext': 'mp4',
+            'age_limit': 18,
+            'title': 'Slut brunette college student anal dorm',
         }
+    }]
 
     def _real_extract(self, url):
         video_id = self._match_id(url)
 
-        page_url = 'http://m.nuvid.com/video/%s' % video_id
-        webpage = self._download_webpage(
-            page_url, video_id, 'Downloading video page')
-        # When dwnld_speed exists and has a value larger than the MP4 file's
-        # bitrate, Nuvid returns the MP4 URL
-        # It's unit is 100bytes/millisecond, see mobile-nuvid-min.js for the algorithm
-        self._set_cookie('nuvid.com', 'dwnld_speed', '10.0')
-        mp4_webpage = self._download_webpage(
-            page_url, video_id, 'Downloading video page for MP4 format')
+        qualities = {
+            'lq': '360p',
+            'hq': '720p',
+        }
 
-        html5_video_re = r'(?s)<(?:video|audio)[^<]*(?:>.*?<source[^>]*)?\s+src=["\'](.*?)["\']',
-        video_url = self._html_search_regex(html5_video_re, webpage, video_id)
-        mp4_video_url = self._html_search_regex(html5_video_re, mp4_webpage, video_id)
-        formats = [{
-            'url': video_url,
-        }]
-        if mp4_video_url != video_url:
-            formats.append({
-                'url': mp4_video_url,
+        json_url = f'https://www.nuvid.com/player_config_json/?vid={video_id}&aid=0&domain_id=0&embed=0&check_speed=0'
+        video_data = self._download_json(
+            json_url, video_id, headers={
+                'Accept': 'application/json, text/javascript, */*; q = 0.01',
+                'Content-Type': 'application/x-www-form-urlencoded; charset=utf-8',
             })
 
-        title = self._html_search_regex(
-            [r'<span title="([^"]+)">',
-             r'<div class="thumb-holder video">\s*<h5[^>]*>([^<]+)</h5>',
-             r'<span[^>]+class="title_thumb">([^<]+)</span>'], webpage, 'title').strip()
-        thumbnails = [
-            {
-                'url': thumb_url,
-            } for thumb_url in re.findall(r'<img src="([^"]+)" alt="" />', webpage)
-        ]
-        thumbnail = thumbnails[0]['url'] if thumbnails else None
-        duration = parse_duration(self._html_search_regex(
-            [r'<i class="fa fa-clock-o"></i>\s*(\d{2}:\d{2})',
-             r'<span[^>]+class="view_time">([^<]+)</span>'], webpage, 'duration', fatal=False))
+        formats = [{
+            'url': source,
+            'format_id': qualities.get(quality),
+            'height': int_or_none(qualities.get(quality)[:-1]),
+        } for quality, source in video_data.get('files').items() if source]
+
+        self._check_formats(formats, video_id)
+        self._sort_formats(formats)
+
+        title = video_data.get('title')
+        thumbnail_base_url = try_get(video_data, lambda x: x['thumbs']['url'])
+        thumbnail_extension = try_get(video_data, lambda x: x['thumbs']['extension'])
+        thumbnail_id = self._search_regex(
+            r'/media/videos/tmb/6523263/preview/(/d+)' + thumbnail_extension, video_data.get('poster', ''), 'thumbnail id', default=19)
+        thumbnail = f'{thumbnail_base_url}player/{thumbnail_id}{thumbnail_extension}'
+        duration = parse_duration(video_data.get('duration') or video_data.get('duration_format'))
 
         return {
             'id': video_id,
+            'formats': formats,
             'title': title,
-            'thumbnails': thumbnails,
             'thumbnail': thumbnail,
             'duration': duration,
             'age_limit': 18,
-            'formats': formats,
         }