Merge branch 'cn-verification-proxy'

youtube_dl/YoutubeDL.py

@@ -54,6 +54,7 @@
     MaxDownloadsReached,
     PagedList,
     parse_filesize,
+    PerRequestProxyHandler,
     PostProcessingError,
     platform_name,
     preferredencoding,
@@ -183,6 +184,8 @@ class YoutubeDL(object):
     prefer_insecure:   Use HTTP instead of HTTPS to retrieve information.
                        At the moment, this is only supported by YouTube.
     proxy:             URL of the proxy server to use
+    cn_verification_proxy:  URL of the proxy to use for IP address verification
+                       on Chinese sites. (Experimental)
     socket_timeout:    Time to wait for unresponsive hosts, in seconds
     bidi_workaround:   Work around buggy terminals without bidirectional text
                        support, using fridibi
@@ -1762,13 +1765,14 @@ def _setup_opener(self):
             # Set HTTPS proxy to HTTP one if given (https://github.com/rg3/youtube-dl/issues/805)
             if 'http' in proxies and 'https' not in proxies:
                 proxies['https'] = proxies['http']
-        proxy_handler = compat_urllib_request.ProxyHandler(proxies)
+        proxy_handler = PerRequestProxyHandler(proxies)
 
         debuglevel = 1 if self.params.get('debug_printtraffic') else 0
         https_handler = make_HTTPS_handler(self.params, debuglevel=debuglevel)
         ydlh = YoutubeDLHandler(self.params, debuglevel=debuglevel)
         opener = compat_urllib_request.build_opener(
-            https_handler, proxy_handler, cookie_processor, ydlh)
+            proxy_handler, https_handler, cookie_processor, ydlh)
+
         # Delete the default user-agent header, which would otherwise apply in
         # cases where our custom HTTP handler doesn't come into play
         # (See https://github.com/rg3/youtube-dl/issues/1309 for details)

youtube_dl/__init__.py

@@ -364,6 +364,7 @@ def _real_main(argv=None):
         'ffmpeg_location': opts.ffmpeg_location,
         'hls_prefer_native': opts.hls_prefer_native,
         'external_downloader_args': external_downloader_args,
+        'cn_verification_proxy': opts.cn_verification_proxy,
     }
 
     with YoutubeDL(ydl_opts) as ydl:

youtube_dl/extractor/letv.py

@@ -7,8 +7,9 @@
 
 from .common import InfoExtractor
 from ..compat import (
-    compat_urlparse,
     compat_urllib_parse,
+    compat_urllib_request,
+    compat_urlparse,
 )
 from ..utils import (
     determine_ext,
@@ -39,12 +40,20 @@ class LetvIE(InfoExtractor):
             'title': '美人天下01',
             'description': 'md5:f88573d9d7225ada1359eaf0dbf8bcda',
         },
-        'expected_warnings': [
-            'publish time'
-        ]
+    }, {
+        'note': 'This video is available only in Mainland China, thus a proxy is needed',
+        'url': 'http://www.letv.com/ptv/vplay/1118082.html',
+        'md5': 'f80936fbe20fb2f58648e81386ff7927',
+        'info_dict': {
+            'id': '1118082',
+            'ext': 'mp4',
+            'title': '与龙共舞 完整版',
+            'description': 'md5:7506a5eeb1722bb9d4068f85024e3986',
+        },
+        'params': {
+            'cn_verification_proxy': 'http://proxy.uku.im:8888'
+        },
     }]
-    # http://www.letv.com/ptv/vplay/1118082.html
-    # This video is available only in Mainland China
 
     @staticmethod
     def urshift(val, n):
@@ -76,8 +85,14 @@ def _real_extract(self, url):
             'tkey': self.calc_time_key(int(time.time())),
             'domain': 'www.letv.com'
         }
+        play_json_req = compat_urllib_request.Request(
+            'http://api.letv.com/mms/out/video/playJson?' + compat_urllib_parse.urlencode(params)
+        )
+        play_json_req.add_header(
+            'Ytdl-request-proxy',
+            self._downloader.params.get('cn_verification_proxy'))
         play_json = self._download_json(
-            'http://api.letv.com/mms/out/video/playJson?' + compat_urllib_parse.urlencode(params),
+            play_json_req,
             media_id, 'playJson data')
 
         # Check for errors
@@ -114,7 +129,8 @@ def _real_extract(self, url):
 
                 url_info_dict = {
                     'url': media_url,
-                    'ext': determine_ext(dispatch[format_id][1])
+                    'ext': determine_ext(dispatch[format_id][1]),
+                    'format_id': format_id,
                 }
 
                 if format_id[-1:] == 'p':
@@ -123,7 +139,7 @@ def _real_extract(self, url):
                 urls.append(url_info_dict)
 
         publish_time = parse_iso8601(self._html_search_regex(
-            r'发布时间&nbsp;([^<>]+) ', page, 'publish time', fatal=False),
+            r'发布时间&nbsp;([^<>]+) ', page, 'publish time', default=None),
             delimiter=' ', timezone=datetime.timedelta(hours=8))
         description = self._html_search_meta('description', page, fatal=False)
 

youtube_dl/options.py

@@ -195,6 +195,12 @@ def _hide_login_info(opts):
         action='store_const', const='::', dest='source_address',
         help='Make all connections via IPv6 (experimental)',
     )
+    network.add_option(
+        '--cn-verification-proxy',
+        dest='cn_verification_proxy', default=None, metavar='URL',
+        help='Use this proxy to verify the IP address for some Chinese sites. '
+        'The default proxy specified by --proxy (or none, if the options is not present) is used for the actual downloading. (experimental)'
+    )
 
     selection = optparse.OptionGroup(parser, 'Video Selection')
     selection.add_option(

youtube_dl/utils.py

@@ -1768,3 +1768,24 @@ def _match_func(info_dict):
             video_title = info_dict.get('title', info_dict.get('id', 'video'))
             return '%s does not pass filter %s, skipping ..' % (video_title, filter_str)
     return _match_func
+
+
+class PerRequestProxyHandler(compat_urllib_request.ProxyHandler):
+    def __init__(self, proxies=None):
+        # Set default handlers
+        for type in ('http', 'https'):
+            setattr(self, '%s_open' % type,
+                    lambda r, proxy='__noproxy__', type=type, meth=self.proxy_open:
+                        meth(r, proxy, type))
+        return compat_urllib_request.ProxyHandler.__init__(self, proxies)
+
+    def proxy_open(self, req, proxy, type):
+        req_proxy = req.headers.get('Ytdl-request-proxy')
+        if req_proxy is not None:
+            proxy = req_proxy
+            del req.headers['Ytdl-request-proxy']
+
+        if proxy == '__noproxy__':
+            return None  # No Proxy
+        return compat_urllib_request.ProxyHandler.proxy_open(
+            self, req, proxy, type)