[archiveorg] improve format extraction(closes #10219)

diff --git a/youtube_dl/extractor/archiveorg.py b/youtube_dl/extractor/archiveorg.py
index 8feb7cb7456ec4db8d6a8f28b411a58cb5ac47a1..2472e4cc6189a3fd47cc24f1a2e407a94b67d084 100644 (file)
--- a/youtube_dl/extractor/archiveorg.py
+++ b/youtube_dl/extractor/archiveorg.py
@@ -1,67 +1,65 @@
 from __future__ import unicode_literals
 
-from .common import InfoExtractor
-from ..utils import unified_strdate
+from .jwplatform import JWPlatformBaseIE
+from ..utils import (
+    unified_strdate,
+    clean_html,
+)
 
 
-class ArchiveOrgIE(InfoExtractor):
+class ArchiveOrgIE(JWPlatformBaseIE):
     IE_NAME = 'archive.org'
     IE_DESC = 'archive.org videos'
-    _VALID_URL = r'https?://(?:www\.)?archive\.org/details/(?P<id>[^?/]+)(?:[?].*)?$'
+    _VALID_URL = r'https?://(?:www\.)?archive\.org/(?:details|embed)/(?P<id>[^/?#]+)(?:[?].*)?$'
     _TESTS = [{
         'url': 'http://archive.org/details/XD300-23_68HighlightsAResearchCntAugHumanIntellect',
         'md5': '8af1d4cf447933ed3c7f4871162602db',
         'info_dict': {
             'id': 'XD300-23_68HighlightsAResearchCntAugHumanIntellect',
-            'ext': 'ogv',
+            'ext': 'ogg',
             'title': '1968 Demo - FJCC Conference Presentation Reel #1',
-            'description': 'md5:1780b464abaca9991d8968c877bb53ed',
+            'description': 'md5:da45c349df039f1cc8075268eb1b5c25',
             'upload_date': '19681210',
             'uploader': 'SRI International'
         }
     }, {
         'url': 'https://archive.org/details/Cops1922',
-        'md5': '18f2a19e6d89af8425671da1cf3d4e04',
+        'md5': 'bc73c8ab3838b5a8fc6c6651fa7b58ba',
         'info_dict': {
             'id': 'Cops1922',
-            'ext': 'ogv',
+            'ext': 'mp4',
             'title': 'Buster Keaton\'s "Cops" (1922)',
-            'description': 'md5:70f72ee70882f713d4578725461ffcc3',
+            'description': 'md5:b4544662605877edd99df22f9620d858',
         }
+    }, {
+        'url': 'http://archive.org/embed/XD300-23_68HighlightsAResearchCntAugHumanIntellect',
+        'only_matching': True,
     }]
 
     def _real_extract(self, url):
         video_id = self._match_id(url)
+        webpage = self._download_webpage(
+            'http://archive.org/embed/' + video_id, video_id)
+        jwplayer_playlist = self._parse_json(self._search_regex(
+            r"(?s)Play\('[^']+'\s*,\s*(\[.+\])\s*,\s*{.*?}\);",
+            webpage, 'jwplayer playlist'), video_id)
+        info = self._parse_jwplayer_data(
+            {'playlist': jwplayer_playlist}, video_id, base_url=url)
 
-        json_url = url + ('&' if '?' in url else '?') + 'output=json'
-        data = self._download_json(json_url, video_id)
-
-        def get_optional(data_dict, field):
-            return data_dict['metadata'].get(field, [None])[0]
-
-        title = get_optional(data, 'title')
-        description = get_optional(data, 'description')
-        uploader = get_optional(data, 'creator')
-        upload_date = unified_strdate(get_optional(data, 'date'))
+        def get_optional(metadata, field):
+            return metadata.get(field, [None])[0]
 
-        formats = [
-            {
-                'format': fdata['format'],
-                'url': 'http://' + data['server'] + data['dir'] + fn,
-                'file_size': int(fdata['size']),
-            }
-            for fn, fdata in data['files'].items()
-            if 'Video' in fdata['format']]
-
-        self._sort_formats(formats)
-
-        return {
-            '_type': 'video',
-            'id': video_id,
-            'title': title,
-            'formats': formats,
-            'description': description,
-            'uploader': uploader,
-            'upload_date': upload_date,
-            'thumbnail': data.get('misc', {}).get('image'),
-        }
+        metadata = self._download_json(
+            'http://archive.org/details/' + video_id, video_id, query={
+                'output': 'json',
+            })['metadata']
+        info.update({
+            'title': get_optional(metadata, 'title') or info.get('title'),
+            'description': clean_html(get_optional(metadata, 'description')),
+        })
+        if info.get('_type') != 'playlist':
+            info.update({
+                'uploader': get_optional(metadata, 'creator'),
+                'upload_date': unified_strdate(get_optional(metadata, 'date')),
+            })
+        return info
\ No newline at end of file