Merge tag 'upstream/2016.02.22'

[youtubedl] / youtube_dl / extractor / infoq.py

diff --git a/youtube_dl/extractor/infoq.py b/youtube_dl/extractor/infoq.py
index c79c589c72cac1fef1240aefd6f90a677ad3bdaf..016af2084b3301bccf6f6191df66b2eec84968a4 100644 (file)
--- a/youtube_dl/extractor/infoq.py
+++ b/youtube_dl/extractor/infoq.py
@@ -1,62 +1,115 @@
+# coding: utf-8
+
+from __future__ import unicode_literals
+

 import base64
 import base64

-import re

 
 from .common import InfoExtractor
 
 from .common import InfoExtractor

-from ..utils import (
-    compat_urllib_parse,
-
-    ExtractorError,
+from ..compat import (
+    compat_urllib_parse_unquote,
+    compat_parse_qs,

 )
 )

+from ..utils import determine_ext

 
 
 class InfoQIE(InfoExtractor):
 
 
 class InfoQIE(InfoExtractor):

-    _VALID_URL = r'^(?:https?://)?(?:www\.)?infoq\.com/[^/]+/[^/]+$'
-    _TEST = {
-        u"name": u"InfoQ",
-        u"url": u"http://www.infoq.com/presentations/A-Few-of-My-Favorite-Python-Things",
-        u"file": u"12-jan-pythonthings.mp4",
-        u"info_dict": {
-            u"description": u"Mike Pirnat presents some tips and tricks, standard libraries and third party packages that make programming in Python a richer experience.",
-            u"title": u"A Few of My Favorite [Python] Things"
+    _VALID_URL = r'https?://(?:www\.)?infoq\.com/(?:[^/]+/)+(?P<id>[^/]+)'
+
+    _TESTS = [{
+        'url': 'http://www.infoq.com/presentations/A-Few-of-My-Favorite-Python-Things',
+        'md5': 'b5ca0e0a8c1fed93b0e65e48e462f9a2',
+        'info_dict': {
+            'id': 'A-Few-of-My-Favorite-Python-Things',
+            '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',

         },
         },

-        u"params": {
-            u"skip_download": True
-        }
-    }
+    }, {
+        'url': 'http://www.infoq.com/fr/presentations/changez-avis-sur-javascript',
+        'only_matching': True,
+    }, {
+        'url': 'http://www.infoq.com/cn/presentations/openstack-continued-delivery',
+        'md5': '4918d0cca1497f2244572caf626687ef',
+        'info_dict': {
+            'id': 'openstack-continued-delivery',
+            'title': 'OpenStack持续交付之路',
+            'ext': 'flv',
+            'description': 'md5:308d981fb28fa42f49f9568322c683ff',
+        },
+    }]

 
 

-    def _real_extract(self, url):
-        mobj = re.match(self._VALID_URL, url)
+    def _extract_bokecc_videos(self, webpage, video_id):
+        # TODO: bokecc.com is a Chinese video cloud platform
+        # It should have an independent extractor but I don't have other
+        # examples using bokecc
+        player_params_str = self._html_search_regex(
+            r'<script[^>]+src="http://p\.bokecc\.com/player\?([^"]+)',
+            webpage, 'player params', default=None)
+
+        player_params = compat_parse_qs(player_params_str)
+
+        info_xml = self._download_xml(
+            'http://p.bokecc.com/servlet/playinfo?uid=%s&vid=%s&m=1' % (
+                player_params['siteid'][0], player_params['vid'][0]), video_id)
+
+        return [{
+            'format_id': 'bokecc',
+            'url': quality.find('./copy').attrib['playurl'],
+            'preference': int(quality.attrib['value']),
+        } for quality in info_xml.findall('./video/quality')]

 
 

-        webpage = self._download_webpage(url, video_id=url)
-        self.report_extraction(url)
+    def _extract_rtmp_videos(self, webpage):
+        # The server URL is hardcoded
+        video_url = 'rtmpe://video.infoq.com/cfx/st/'

 
         # Extract video URL
 
         # 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
+        encoded_id = self._search_regex(
+            r"jsclassref\s*=\s*'([^']*)'", webpage, 'encoded id', default=None)

 
 

-        # Extract title
-        video_title = self._search_regex(r'contentTitle = "(.*?)";',
-            webpage, u'title')
+        real_id = compat_urllib_parse_unquote(base64.b64decode(encoded_id.encode('ascii')).decode('utf-8'))
+        playpath = 'mp4:' + real_id

 
 

-        # Extract description
-        video_description = self._html_search_regex(r'<meta name="description" content="(.*)"(?:\s*/)?>',
-            webpage, u'description', fatal=False)
+        return [{
+            'format_id': 'rtmp',
+            'url': video_url,
+            'ext': determine_ext(playpath),
+            'play_path': playpath,
+        }]
+
+    def _extract_http_videos(self, webpage):
+        http_video_url = self._search_regex(r'P\.s\s*=\s*\'([^\']+)\'', webpage, 'video URL')
+
+        policy = self._search_regex(r'InfoQConstants.scp\s*=\s*\'([^\']+)\'', webpage, 'policy')
+        signature = self._search_regex(r'InfoQConstants.scs\s*=\s*\'([^\']+)\'', webpage, 'signature')
+        key_pair_id = self._search_regex(r'InfoQConstants.sck\s*=\s*\'([^\']+)\'', webpage, 'key-pair-id')
+
+        return [{
+            'format_id': 'http',
+            'url': http_video_url,
+            'http_headers': {
+                'Cookie': 'CloudFront-Policy=%s; CloudFront-Signature=%s; CloudFront-Key-Pair-Id=%s' % (
+                    policy, signature, key_pair_id),
+            },
+        }]
+
+    def _real_extract(self, url):
+        video_id = self._match_id(url)
+        webpage = self._download_webpage(url, video_id)
+
+        video_title = self._html_search_regex(r'<title>(.*?)</title>', webpage, 'title')
+        video_description = self._html_search_meta('description', webpage, 'description')
+
+        if '/cn/' in url:
+            # for China videos, HTTP video URL exists but always fails with 403
+            formats = self._extract_bokecc_videos(webpage, video_id)
+        else:
+            formats = self._extract_rtmp_videos(webpage) + self._extract_http_videos(webpage)

 
 

-        video_filename = video_url.split('/')[-1]
-        video_id, extension = video_filename.split('.')
+        self._sort_formats(formats)

 
 

-        info = {
+        return {

             'id': video_id,
             'id': video_id,

-            'url': video_url,
-            'uploader': None,
-            'upload_date': None,

             'title': video_title,
             'title': video_title,

-            'ext': extension, # Extension is always(?) mp4, but seems to be flv
-            'thumbnail': None,

             'description': video_description,
             'description': video_description,

+            'formats': formats,

         }
         }

-
-        return [info]
\ No newline at end of file