Merge pull request #7320 from remitamine/adobetv

[youtube-dl] / youtube_dl / extractor / bilibili.py

diff --git a/youtube_dl/extractor/bilibili.py b/youtube_dl/extractor/bilibili.py
index e00efb6eb0a1e4671f900e1e9bb4b0da144609aa..59beb11bce71bfc6ef9b036ad123dc44e872d0be 100644 (file)
--- a/youtube_dl/extractor/bilibili.py
+++ b/youtube_dl/extractor/bilibili.py
@@ -2,14 +2,14 @@
 from __future__ import unicode_literals
 
 import re
-import json
-import xml.etree.ElementTree as ET
 
 from .common import InfoExtractor
+from ..compat import compat_str
 from ..utils import (
     int_or_none,
     unescapeHTML,
     ExtractorError,
+    xpath_text,
 )
 
 
@@ -56,50 +56,46 @@ class BiliBiliIE(InfoExtractor):
         cid = view_data['cid']
         title = unescapeHTML(view_data['title'])
 
-        page = self._download_webpage(
+        doc = self._download_xml(
             'http://interface.bilibili.com/v_cdn_play?appkey=8e9fc618fbd41e28&cid=%s' % cid,
             cid,
             'Downloading page %s/%s' % (page_num, view_data['pages'])
         )
-        try:
-            err_info = json.loads(page)
-            raise ExtractorError(
-                'BiliBili said: ' + err_info['error_text'], expected=True)
-        except ValueError:
-            pass
 
-        doc = ET.fromstring(page)
-        durls = doc.findall('./durl')
+        if xpath_text(doc, './result') == 'error':
+            raise ExtractorError('%s said: %s' % (self.IE_NAME, xpath_text(doc, './message')), expected=True)
 
         entries = []
 
-        for durl in durls:
-            formats = []
-            backup_url = durl.find('./backup_url')
-            if backup_url is not None:
-                formats.append({'url': backup_url.find('./url').text})
-            size = durl.find('./filesize|./size')
-            formats.append({
+        for durl in doc.findall('./durl'):
+            size = xpath_text(durl, ['./filesize', './size'])
+            formats = [{
                 'url': durl.find('./url').text,
-                'filesize': int_or_none(size.text) if size else None,
+                'filesize': int_or_none(size),
                 'ext': 'flv',
-            })
+            }]
+            backup_urls = durl.find('./backup_url')
+            if backup_urls is not None:
+                for backup_url in backup_urls.findall('./url'):
+                    formats.append({'url': backup_url.text})
+            formats.reverse()
+
             entries.append({
-                'id': '%s_part%s' % (cid, durl.find('./order').text),
+                'id': '%s_part%s' % (cid, xpath_text(durl, './order')),
                 'title': title,
-                'duration': int_or_none(durl.find('./length').text) // 1000,
+                'duration': int_or_none(xpath_text(durl, './length'), 1000),
                 'formats': formats,
             })
 
         info = {
-            'id': str(cid),
+            'id': compat_str(cid),
             'title': title,
             'description': view_data.get('description'),
             'thumbnail': view_data.get('pic'),
             'uploader': view_data.get('author'),
             'timestamp': int_or_none(view_data.get('created')),
-            'view_count': view_data.get('play'),
-            'duration': int_or_none(doc.find('./timelength').text),
+            'view_count': int_or_none(view_data.get('play')),
+            'duration': int_or_none(xpath_text(doc, './timelength')),
         }
 
         if len(entries) == 1: