[douyutv] Switch to the PC API to escape the 5-min limitation

Thanks @spacemeowx2 for the algo.

Ref: https://gist.github.com/spacemeowx2/629b1d131bd7e240a7d28742048e80fc

Closes #12316

diff --git a/ChangeLog b/ChangeLog
index e53fb7767de05ca2417b5a53a846c013ffa17d15..13ccb0f8fa74b24514b9518aeb156877eb746488 100644 (file)
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,3 +1,9 @@
+version <unreleased>
+
+Extractors
+* [douyutv] Switch to the PC API to escape the 5-min limitation (#12316)
+
+
 version 2017.03.02
 
 Core

diff --git a/youtube_dl/extractor/douyutv.py b/youtube_dl/extractor/douyutv.py
index 9a83fb31ab57dd57a49c09bee6c2a8d3be7583ca..82d8a042f58e738422a4c64b7fa90f8ac127f5fb 100644 (file)
--- a/youtube_dl/extractor/douyutv.py
+++ b/youtube_dl/extractor/douyutv.py
@@ -1,6 +1,9 @@
 # coding: utf-8
 from __future__ import unicode_literals
 
+import time
+import hashlib
+
 from .common import InfoExtractor
 from ..utils import (
     ExtractorError,
@@ -16,7 +19,7 @@ class DouyuTVIE(InfoExtractor):
         'info_dict': {
             'id': '17732',
             'display_id': 'iseven',
-            'ext': 'mp4',
+            'ext': 'flv',
             'title': 're:^清晨醒脑！T-ARA根本停不下来！ [0-9]{4}-[0-9]{2}-[0-9]{2} [0-9]{2}:[0-9]{2}$',
             'description': r're:.*m7show@163\.com.*',
             'thumbnail': r're:^https?://.*\.jpg$',
@@ -31,7 +34,7 @@ class DouyuTVIE(InfoExtractor):
         'info_dict': {
             'id': '85982',
             'display_id': '85982',
-            'ext': 'mp4',
+            'ext': 'flv',
             'title': 're:^小漠从零单排记！——CSOL2躲猫猫 [0-9]{4}-[0-9]{2}-[0-9]{2} [0-9]{2}:[0-9]{2}$',
             'description': 'md5:746a2f7a253966a06755a912f0acc0d2',
             'thumbnail': r're:^https?://.*\.jpg$',
@@ -47,7 +50,7 @@ class DouyuTVIE(InfoExtractor):
         'info_dict': {
             'id': '17732',
             'display_id': '17732',
-            'ext': 'mp4',
+            'ext': 'flv',
             'title': 're:^清晨醒脑！T-ARA根本停不下来！ [0-9]{4}-[0-9]{2}-[0-9]{2} [0-9]{2}:[0-9]{2}$',
             'description': r're:.*m7show@163\.com.*',
             'thumbnail': r're:^https?://.*\.jpg$',
@@ -66,10 +69,6 @@ class DouyuTVIE(InfoExtractor):
         'only_matching': True,
     }]
 
-    # Decompile core.swf in webpage by ffdec "Search SWFs in memory". core.swf
-    # is encrypted originally, but ffdec can dump memory to get the decrypted one.
-    _API_KEY = 'A12Svb&%1UUmf@hC'
-
     def _real_extract(self, url):
         video_id = self._match_id(url)
 
@@ -80,6 +79,7 @@ class DouyuTVIE(InfoExtractor):
             room_id = self._html_search_regex(
                 r'"room_id\\?"\s*:\s*(\d+),', page, 'room id')
 
+        # Grab metadata from mobile API
         room = self._download_json(
             'http://m.douyu.com/html5/live?roomId=%s' % room_id, video_id,
             note='Downloading room info')['data']
@@ -88,8 +88,19 @@ class DouyuTVIE(InfoExtractor):
         if room.get('show_status') == '2':
             raise ExtractorError('Live stream is offline', expected=True)
 
-        formats = self._extract_m3u8_formats(
-            room['hls_url'], video_id, ext='mp4')
+        # Grab the URL from PC client API
+        # The m3u8 url from mobile API requires re-authentication every 5 minutes
+        tt = int(time.time())
+        signContent = 'lapi/live/thirdPart/getPlay/%s?aid=pcclient&rate=0&time=%d9TUk5fjjUjg9qIMH3sdnh' % (room_id, tt)
+        sign = hashlib.md5(signContent.encode('ascii')).hexdigest()
+        video_url = self._download_json(
+            'http://coapi.douyucdn.cn/lapi/live/thirdPart/getPlay/' + room_id,
+            video_id, note='Downloading video URL info',
+            query={'rate': 0}, headers={
+                'auth': sign,
+                'time': str(tt),
+                'aid': 'pcclient'
+            })['data']['live_url']
 
         title = self._live_title(unescapeHTML(room['room_name']))
         description = room.get('show_details')
@@ -99,7 +110,7 @@ class DouyuTVIE(InfoExtractor):
         return {
             'id': room_id,
             'display_id': video_id,
-            'formats': formats,
+            'url': video_url,
             'title': title,
             'description': description,
             'thumbnail': thumbnail,