]> Raphaƫl G. Git Repositories - youtubedl/blobdiff - youtube_dl/extractor/infoq.py
debian/watch: Make use of the signing key from upstream.
[youtubedl] / youtube_dl / extractor / infoq.py
index 9056742821a2fd99fadd5b1271df7dc6f5127749..e76dd222d1ee81dc0e0b2d5b1b3c28ef22e1bd83 100644 (file)
@@ -1,50 +1,67 @@
+from __future__ import unicode_literals
+
 import base64
 import re
 
 from .common import InfoExtractor
 from ..utils import (
     compat_urllib_parse,
-
-    ExtractorError,
 )
 
 
 class InfoQIE(InfoExtractor):
-    _VALID_URL = r'^(?:https?://)?(?:www\.)?infoq\.com/[^/]+/[^/]+$'
+    _VALID_URL = r'https?://(?:www\.)?infoq\.com/[^/]+/(?P<id>[^/]+)$'
+
+    _TEST = {
+        'url': 'http://www.infoq.com/presentations/A-Few-of-My-Favorite-Python-Things',
+        'md5': 'b5ca0e0a8c1fed93b0e65e48e462f9a2',
+        'info_dict': {
+            'id': '12-jan-pythonthings',
+            'ext': 'mp4',
+            'description': 'Mike Pirnat presents some tips and tricks, standard libraries and third party packages that make programming in Python a richer experience.',
+            'title': 'A Few of My Favorite [Python] Things',
+        },
+    }
 
     def _real_extract(self, url):
         mobj = re.match(self._VALID_URL, url)
+        video_id = mobj.group('id')
 
-        webpage = self._download_webpage(url, video_id=url)
-        self.report_extraction(url)
+        webpage = self._download_webpage(url, video_id)
 
-        # Extract video URL
-        mobj = re.search(r"jsclassref ?= ?'([^']*)'", webpage)
-        if mobj is None:
-            raise ExtractorError(u'Unable to extract video url')
-        real_id = compat_urllib_parse.unquote(base64.b64decode(mobj.group(1).encode('ascii')).decode('utf-8'))
-        video_url = 'rtmpe://video.infoq.com/cfx/st/' + real_id
+        video_title = self._html_search_regex(r'<title>(.*?)</title>', webpage, 'title')
+        video_description = self._html_search_meta('description', webpage, 'description')
 
-        # Extract title
-        video_title = self._search_regex(r'contentTitle = "(.*?)";',
-            webpage, u'title')
+        # The server URL is hardcoded
+        video_url = 'rtmpe://video.infoq.com/cfx/st/'
 
-        # Extract description
-        video_description = self._html_search_regex(r'<meta name="description" content="(.*)"(?:\s*/)?>',
-            webpage, u'description', fatal=False)
+        # Extract video URL
+        encoded_id = self._search_regex(
+            r"jsclassref\s*=\s*'([^']*)'", webpage, 'encoded id')
+        real_id = compat_urllib_parse.unquote(base64.b64decode(encoded_id.encode('ascii')).decode('utf-8'))
+        playpath = 'mp4:' + real_id
 
-        video_filename = video_url.split('/')[-1]
+        video_filename = playpath.split('/')[-1]
         video_id, extension = video_filename.split('.')
 
-        info = {
-            'id': video_id,
+        http_base = self._search_regex(
+            r'EXPRESSINSTALL_SWF\s*=\s*"(https?://[^/"]+/)', webpage,
+            'HTTP base URL')
+
+        formats = [{
+            'format_id': 'rtmp',
             'url': video_url,
-            'uploader': None,
-            'upload_date': None,
+            'ext': extension,
+            'play_path': playpath,
+        }, {
+            'format_id': 'http',
+            'url': http_base + real_id,
+        }]
+        self._sort_formats(formats)
+
+        return {
+            'id': video_id,
             'title': video_title,
-            'ext': extension, # Extension is always(?) mp4, but seems to be flv
-            'thumbnail': None,
             'description': video_description,
+            'formats': formats,
         }
-
-        return [info]
\ No newline at end of file