[extractor] Deprecate `_sort_formats`

[yt-dlp.git] / yt_dlp / extractor / la7.py

diff --git a/yt_dlp/extractor/la7.py b/yt_dlp/extractor/la7.py
index 363fbd6a51c7da49c8d37205868e1d93294ee678..68dc1d4df134a99f8c573c3992617344f415a377 100644 (file)
--- a/yt_dlp/extractor/la7.py
+++ b/yt_dlp/extractor/la7.py
@@ -1,14 +1,12 @@
-# coding: utf-8
-from __future__ import unicode_literals
-
 import re
 
 from .common import InfoExtractor
 from ..utils import (
     determine_ext,
     float_or_none,
+    HEADRequest,
+    int_or_none,
     parse_duration,
-    smuggle_url,
     unified_strdate,
 )
 
@@ -25,19 +23,38 @@ class LA7IE(InfoExtractor):
         'url': 'http://www.la7.it/crozza/video/inccool8-02-10-2015-163722',
         'md5': '8b613ffc0c4bf9b9e377169fc19c214c',
         'info_dict': {
-            'id': '0_42j6wd36',
+            'id': 'inccool8-02-10-2015-163722',
             'ext': 'mp4',
             'title': 'Inc.Cool8',
             'description': 'Benvenuti nell\'incredibile mondo della INC. COOL. 8. dove “INC.” sta per “Incorporated” “COOL” sta per “fashion” ed Eight sta per il gesto atletico',
             'thumbnail': 're:^https?://.*',
-            'uploader_id': 'kdla7pillole@iltrovatore.it',
-            'timestamp': 1443814869,
             'upload_date': '20151002',
         },
     }, {
         'url': 'http://www.la7.it/omnibus/rivedila7/omnibus-news-02-07-2016-189077',
         'only_matching': True,
     }]
+    _HOST = 'https://awsvodpkg.iltrovatore.it'
+
+    def _generate_mp4_url(self, quality, m3u8_formats):
+        for f in m3u8_formats:
+            if f['vcodec'] != 'none' and quality in f['url']:
+                http_url = '%s%s.mp4' % (self._HOST, quality)
+
+                urlh = self._request_webpage(
+                    HEADRequest(http_url), quality,
+                    note='Check filesize', fatal=False)
+                if urlh:
+                    http_f = f.copy()
+                    del http_f['manifest_url']
+                    http_f.update({
+                        'format_id': http_f['format_id'].replace('hls-', 'https-'),
+                        'url': http_url,
+                        'protocol': 'https',
+                        'filesize_approx': int_or_none(urlh.headers.get('Content-Length', None)),
+                    })
+                    return http_f
+                return None
 
     def _real_extract(self, url):
         video_id = self._match_id(url)
@@ -46,22 +63,28 @@ def _real_extract(self, url):
             url = '%s//%s' % (self.http_scheme(), url)
 
         webpage = self._download_webpage(url, video_id)
+        video_path = self._search_regex(r'(/content/.*?).mp4', webpage, 'video_path')
+
+        formats = self._extract_mpd_formats(
+            f'{self._HOST}/local/dash/,{video_path}.mp4.urlset/manifest.mpd',
+            video_id, mpd_id='dash', fatal=False)
+        m3u8_formats = self._extract_m3u8_formats(
+            f'{self._HOST}/local/hls/,{video_path}.mp4.urlset/master.m3u8',
+            video_id, 'mp4', m3u8_id='hls', fatal=False)
+        formats.extend(m3u8_formats)
 
-        player_data = self._search_regex(
-            [r'(?s)videoParams\s*=\s*({.+?});', r'videoLa7\(({[^;]+})\);'],
-            webpage, 'player data')
-        vid = self._search_regex(r'vid\s*:\s*"(.+?)",', player_data, 'vid')
+        for q in filter(None, video_path.split(',')):
+            http_f = self._generate_mp4_url(q, m3u8_formats)
+            if http_f:
+                formats.append(http_f)
 
         return {
-            '_type': 'url_transparent',
-            'url': smuggle_url('kaltura:103:%s' % vid, {
-                'service_url': 'http://nkdam.iltrovatore.it',
-            }),
             'id': video_id,
             'title': self._og_search_title(webpage, default=None),
             'description': self._og_search_description(webpage, default=None),
             'thumbnail': self._og_search_thumbnail(webpage, default=None),
-            'ie_key': 'Kaltura',
+            'formats': formats,
+            'upload_date': unified_strdate(self._search_regex(r'datetime="(.+?)"', webpage, 'upload_date', fatal=False))
         }
 
 
@@ -111,7 +134,6 @@ def _extract_info(self, webpage, video_id=None, ppn=None):
             'format_id': ext,
             'ext': ext,
         }]
-        self._sort_formats(formats)
 
         title = self._html_search_regex(
             (r'<div class="title">(?P<title>.+?)</',
@@ -169,7 +191,7 @@ def _real_extract(self, url):
         return self._extract_info(webpage, video_id)
 
 
-class LA7PodcastIE(LA7PodcastEpisodeIE):
+class LA7PodcastIE(LA7PodcastEpisodeIE):  # XXX: Do not subclass from concrete IE
     IE_NAME = 'la7.it:podcast'
     _VALID_URL = r'(https?://)?(www\.)?la7\.it/(?P<id>[^/]+)/podcast/?(?:$|[#?])'