[sverigesradio] improve extraction(closes #18635)

1 files changed, 59 insertions, 49 deletions

diff --git a/youtube_dl/extractor/sverigesradio.py b/youtube_dl/extractor/sverigesradio.py
index 05de31a79..aa0691f0d 100644
--- a/youtube_dl/extractor/sverigesradio.py
+++ b/youtube_dl/extractor/sverigesradio.py
@@ -2,58 +2,70 @@
 from __future__ import unicode_literals
 
 from .common import InfoExtractor
-from ..utils import int_or_none
+from ..utils import (
+    determine_ext,
+    int_or_none,
+    str_or_none,
+)
 
 
 class SverigesRadioBaseIE(InfoExtractor):
-    _BASE_URL = 'https://sverigesradio.se/sida/playerajax'
-    _QUALITIES = ['high', 'medium', 'low']
-    _CODING_FORMATS = {
-        5: {'acodec': 'mp3', 'abr': 128},
-        11: {'acodec': 'aac', 'abr': 192},
-        12: {'acodec': 'aac', 'abr': 32},
-        13: {'acodec': 'aac', 'abr': 96},
+    _BASE_URL = 'https://sverigesradio.se/sida/playerajax/'
+    _QUALITIES = ['low', 'medium', 'high']
+    _EXT_TO_CODEC_MAP = {
+        'mp3': 'mp3',
+        'm4a': 'aac',
+    }
+    _CODING_FORMAT_TO_ABR_MAP = {
+        5: 128,
+        11: 192,
+        12: 32,
+        13: 96,
     }
 
-    def _extract_formats(self, query, audio_id, audio_type):
-        audiourls = {}
-        for quality in self._QUALITIES:
-            audiourl = self._download_json(
-                self._BASE_URL + '/getaudiourl', audio_id,
-                fatal=True,
-                query=dict(query, type=audio_type, quality=quality, format='iis'))
-            if audiourl is None:
-                continue
-
-            # for some reason url can be empty, skip if so
-            # also skip if url has already been seen (quality parameter is ignored?)
-            url = audiourl.get('audioUrl')
-            if url is None or url == "" or url in audiourls:
-                continue
-
-            audioformat = {'vcodec': 'none', 'url': url}
-            # add codec and bitrate if known coding format
-            codingformat = audiourl.get('codingFormat')
-            if codingformat:
-                audioformat.update(self._CODING_FORMATS.get(codingformat, {}))
-
-            audiourls[url] = audioformat
-
-        return audiourls.values()
-
-    def _extract_audio(self, audio_type, url):
+    def _real_extract(self, url):
         audio_id = self._match_id(url)
-        query = {'id': audio_id, 'type': audio_type}
+        query = {
+            'id': audio_id,
+            'type': self._AUDIO_TYPE,
+        }
 
-        metadata = self._download_json(self._BASE_URL + '/audiometadata', audio_id, query=query)
-        item = metadata['items'][0]
+        item = self._download_json(
+            self._BASE_URL + 'audiometadata', audio_id,
+            'Downloading audio JSON metadata', query=query)['items'][0]
+        title = item['subtitle']
 
-        formats = self._extract_formats(query, audio_id, audio_type)
+        query['format'] = 'iis'
+        urls = []
+        formats = []
+        for quality in self._QUALITIES:
+            query['quality'] = quality
+            audio_url_data = self._download_json(
+                self._BASE_URL + 'getaudiourl', audio_id,
+                'Downloading %s format JSON metadata' % quality,
+                fatal=False, query=query) or {}
+            audio_url = audio_url_data.get('audioUrl')
+            if not audio_url or audio_url in urls:
+                continue
+            urls.append(audio_url)
+            ext = determine_ext(audio_url)
+            coding_format = audio_url_data.get('codingFormat')
+            abr = int_or_none(self._search_regex(
+                r'_a(\d+)\.m4a', audio_url, 'audio bitrate',
+                default=None)) or self._CODING_FORMAT_TO_ABR_MAP.get(coding_format)
+            formats.append({
+                'abr': abr,
+                'acodec': self._EXT_TO_CODEC_MAP.get(ext),
+                'ext': ext,
+                'format_id': str_or_none(coding_format),
+                'vcodec': 'none',
+                'url': audio_url,
+            })
         self._sort_formats(formats)
 
         return {
             'id': audio_id,
-            'title': item['subtitle'],
+            'title': title,
             'formats': formats,
             'series': item.get('title'),
             'duration': int_or_none(item.get('duration')),
@@ -63,7 +75,8 @@ class SverigesRadioBaseIE(InfoExtractor):
 
 
 class SverigesRadioPublicationIE(SverigesRadioBaseIE):
-    _VALID_URL = r'https?://(?:www\.)?sverigesradio\.se/sida/(?:artikel|gruppsida)\.aspx\?.*artikel=(?P<id>[0-9]+)'
+    IE_NAME = 'sverigesradio:publication'
+    _VALID_URL = r'https?://(?:www\.)?sverigesradio\.se/sida/(?:artikel|gruppsida)\.aspx\?.*?\bartikel=(?P<id>[0-9]+)'
     _TESTS = [{
         'url': 'https://sverigesradio.se/sida/artikel.aspx?programid=83&artikel=7038546',
         'md5': '6a4917e1923fccb080e5a206a5afa542',
@@ -74,18 +87,17 @@ class SverigesRadioPublicationIE(SverigesRadioBaseIE):
             'series': 'Nyheter (Ekot)',
             'title': 'Esa Teittinen: Sanningen har inte kommit fram',
             'description': 'md5:daf7ce66a8f0a53d5465a5984d3839df',
-            'thumbnail': 're:^https://static-cdn.sr.se/sida/images/',
+            'thumbnail': r're:^https?://.*\.jpg',
         },
     }, {
         'url': 'https://sverigesradio.se/sida/gruppsida.aspx?programid=3304&grupp=6247&artikel=7146887',
         'only_matching': True,
     }]
-
-    def _real_extract(self, url):
-        return self._extract_audio('publication', url)
+    _AUDIO_TYPE = 'publication'
 
 
 class SverigesRadioEpisodeIE(SverigesRadioBaseIE):
+    IE_NAME = 'sverigesradio:episode'
     _VALID_URL = r'https?://(?:www\.)?sverigesradio\.se/(?:sida/)?avsnitt/(?P<id>[0-9]+)'
     _TEST = {
         'url': 'https://sverigesradio.se/avsnitt/1140922?programid=1300',
@@ -97,9 +109,7 @@ class SverigesRadioEpisodeIE(SverigesRadioBaseIE):
             'series': 'Konflikt',
             'title': 'Metoo och valen',
             'description': 'md5:fcb5c1f667f00badcc702b196f10a27e',
-            'thumbnail': 're:^https://static-cdn.sr.se/sida/images/'
+            'thumbnail': r're:^https?://.*\.jpg',
         }
     }
-
-    def _real_extract(self, url):
-        return self._extract_audio('episode', url)
+    _AUDIO_TYPE = 'episode'