youtube-dl

commit: aa9118a373a6e9cfb9fda24533df86286eccc468
parent 36abc16c3cf050a3368367038d40cce27504c28a
Author: Sergey M․ <dstftw@gmail.com>
Date:   Mon, 22 Feb 2021 02:29:50 +0700

[apa] Improve extraction (closes #27750)

Diffstat:
M
youtube_dl/extractor/apa.py
38
+++++++++++++++++++++++++-------------

1 file changed, 25 insertions(+), 13 deletions(-)
diff --git a/youtube_dl/extractor/apa.py b/youtube_dl/extractor/apa.py
@@ -6,12 +6,13 @@ import re
 from .common import InfoExtractor
 from ..utils import (
     determine_ext,
-    js_to_json,
+    int_or_none,
+    url_or_none,
 )
 
 
 class APAIE(InfoExtractor):
-    _VALID_URL = r'https?://[^/]+\.apa\.at/embed/(?P<id>[\da-f]{8}-[\da-f]{4}-[\da-f]{4}-[\da-f]{4}-[\da-f]{12})'
+    _VALID_URL = r'(?P<base_url>https?://[^/]+\.apa\.at)/embed/(?P<id>[\da-f]{8}-[\da-f]{4}-[\da-f]{4}-[\da-f]{4}-[\da-f]{12})'
     _TESTS = [{
         'url': 'http://uvp.apa.at/embed/293f6d17-692a-44e3-9fd5-7b178f3a1029',
         'md5': '2b12292faeb0a7d930c778c7a5b4759b',
@@ -41,9 +42,11 @@ class APAIE(InfoExtractor):
                 webpage)]
 
     def _real_extract(self, url):
-        video_id = self._match_id(url)
+        mobj = re.match(self._VALID_URL, url)
+        video_id, base_url = mobj.group('id', 'base_url')
 
-        webpage = self._download_webpage('https://uvp.apa.at/player/%s' % video_id, video_id)
+        webpage = self._download_webpage(
+            '%s/player/%s' % (base_url, video_id), video_id)
 
         jwplatform_id = self._search_regex(
             r'media[iI]d\s*:\s*["\'](?P<id>[a-zA-Z0-9]{8})', webpage,
@@ -54,30 +57,39 @@ class APAIE(InfoExtractor):
                 'jwplatform:' + jwplatform_id, ie='JWPlatform',
                 video_id=video_id)
 
-        sources = self._parse_json("{" + self._search_regex(
-            r'("hls"\s*:\s*"[^"]+"\s*,\s*"progressive"\s*:\s*"[^"]+")', webpage, 'sources')
-            + "}", video_id, transform_source=js_to_json)
+        def extract(field, name=None):
+            return self._search_regex(
+                r'\b%s["\']\s*:\s*(["\'])(?P<value>(?:(?!\1).)+)\1' % field,
+                webpage, name or field, default=None, group='value')
+
+        title = extract('title') or video_id
+        description = extract('description')
+        thumbnail = extract('poster', 'thumbnail')
 
         formats = []
-        for (format, source_url) in sources.items():
+        for format_id in ('hls', 'progressive'):
+            source_url = url_or_none(extract(format_id))
+            if not source_url:
+                continue
             ext = determine_ext(source_url)
             if ext == 'm3u8':
                 formats.extend(self._extract_m3u8_formats(
                     source_url, video_id, 'mp4', entry_protocol='m3u8_native',
                     m3u8_id='hls', fatal=False))
             else:
+                height = int_or_none(self._search_regex(
+                    r'(\d+)\.mp4', source_url, 'height', default=None))
                 formats.append({
                     'url': source_url,
+                    'format_id': format_id,
+                    'height': height,
                 })
         self._sort_formats(formats)
 
-        thumbnail = self._search_regex(
-            r'"poster"\s*:\s*(["\'])(?P<url>(?:(?!\1).)+)\1', webpage,
-            'thumbnail', fatal=False, group='url')
-
         return {
             'id': video_id,
-            'title': video_id,
+            'title': title,
+            'description': description,
             'thumbnail': thumbnail,
             'formats': formats,
         }