X-Git-Url: https://git.rapsys.eu/youtubedl/blobdiff_plain/c512650955de0b16d37e7fa7fb29ea0985e415bb..2b4515201a54c38e34434ea2817d606be134a7c9:/youtube_dl/extractor/infoq.py?ds=inline

diff --git a/youtube_dl/extractor/infoq.py b/youtube_dl/extractor/infoq.py
index e76dd22..9fb71e8 100644
--- a/youtube_dl/extractor/infoq.py
+++ b/youtube_dl/extractor/infoq.py
@@ -1,62 +1,130 @@
+# coding: utf-8
+
 from __future__ import unicode_literals
 
 import base64
-import re
 
-from .common import InfoExtractor
-from ..utils import (
-    compat_urllib_parse,
+from ..compat import (
+    compat_urllib_parse_unquote,
+    compat_urlparse,
 )
+from ..utils import determine_ext
+from .bokecc import BokeCCBaseIE
 
 
-class InfoQIE(InfoExtractor):
-    _VALID_URL = r'https?://(?:www\.)?infoq\.com/[^/]+/(?P<id>[^/]+)$'
+class InfoQIE(BokeCCBaseIE):
+    _VALID_URL = r'https?://(?:www\.)?infoq\.com/(?:[^/]+/)+(?P<id>[^/]+)'
 
-    _TEST = {
+    _TESTS = [{
         'url': 'http://www.infoq.com/presentations/A-Few-of-My-Favorite-Python-Things',
         'md5': 'b5ca0e0a8c1fed93b0e65e48e462f9a2',
         'info_dict': {
-            'id': '12-jan-pythonthings',
+            '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',
         },
-    }
-
-    def _real_extract(self, url):
-        mobj = re.match(self._VALID_URL, url)
-        video_id = mobj.group('id')
-
-        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')
+    }, {
+        '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',
+        },
+    }, {
+        'url': 'https://www.infoq.com/presentations/Simple-Made-Easy',
+        'md5': '0e34642d4d9ef44bf86f66f6399672db',
+        'info_dict': {
+            'id': 'Simple-Made-Easy',
+            'title': 'Simple Made Easy',
+            'ext': 'mp3',
+            'description': 'md5:3e0e213a8bbd074796ef89ea35ada25b',
+        },
+        'params': {
+            'format': 'bestaudio',
+        },
+    }]
 
+    def _extract_rtmp_video(self, webpage):
         # The server URL is hardcoded
         video_url = 'rtmpe://video.infoq.com/cfx/st/'
 
         # 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 = playpath.split('/')[-1]
-        video_id, extension = video_filename.split('.')
+            r"jsclassref\s*=\s*'([^']*)'", webpage, 'encoded id', default=None)
 
-        http_base = self._search_regex(
-            r'EXPRESSINSTALL_SWF\s*=\s*"(https?://[^/"]+/)', webpage,
-            'HTTP base URL')
+        real_id = compat_urllib_parse_unquote(base64.b64decode(encoded_id.encode('ascii')).decode('utf-8'))
+        playpath = 'mp4:' + real_id
 
-        formats = [{
-            'format_id': 'rtmp',
+        return [{
+            'format_id': 'rtmp_video',
             'url': video_url,
-            'ext': extension,
+            'ext': determine_ext(playpath),
             'play_path': playpath,
-        }, {
-            'format_id': 'http',
-            'url': http_base + real_id,
         }]
+
+    def _extract_cookies(self, webpage):
+        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 'CloudFront-Policy=%s; CloudFront-Signature=%s; CloudFront-Key-Pair-Id=%s' % (
+            policy, signature, key_pair_id)
+
+    def _extract_http_video(self, webpage):
+        http_video_url = self._search_regex(r'P\.s\s*=\s*\'([^\']+)\'', webpage, 'video URL')
+        return [{
+            'format_id': 'http_video',
+            'url': http_video_url,
+            'http_headers': {
+                'Cookie': self._extract_cookies(webpage)
+            },
+        }]
+
+    def _extract_http_audio(self, webpage, video_id):
+        fields = self._hidden_inputs(webpage)
+        http_audio_url = fields['filename']
+        if http_audio_url is None:
+            return []
+
+        cookies_header = {'Cookie': self._extract_cookies(webpage)}
+
+        # base URL is found in the Location header in the response returned by
+        # GET https://www.infoq.com/mp3download.action?filename=... when logged in.
+        http_audio_url = compat_urlparse.urljoin('http://res.infoq.com/downloads/mp3downloads/', http_audio_url)
+
+        # audio file seem to be missing some times even if there is a download link
+        # so probe URL to make sure
+        if not self._is_valid_url(http_audio_url, video_id, headers=cookies_header):
+            return []
+
+        return [{
+            'format_id': 'http_audio',
+            'url': http_audio_url,
+            'vcodec': 'none',
+            'http_headers': cookies_header,
+        }]
+
+    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_formats(webpage, video_id)
+        else:
+            formats = (
+                self._extract_rtmp_video(webpage) +
+                self._extract_http_video(webpage) +
+                self._extract_http_audio(webpage, video_id))
+
         self._sort_formats(formats)
 
         return {