Merge tag 'upstream/2015.07.21'

[youtubedl] / youtube_dl / extractor / theplatform.py

diff --git a/youtube_dl/extractor/theplatform.py b/youtube_dl/extractor/theplatform.py
index e2653d62dc8c288ce8e58e5bfda52793aef7cfaf..83d833e30dbeb60caa43aa272bfd4d35f4507a53 100644 (file)
--- a/youtube_dl/extractor/theplatform.py
+++ b/youtube_dl/extractor/theplatform.py
@@ -2,13 +2,22 @@ from __future__ import unicode_literals
 
 import re
 import json
 
 import re
 import json

+import time
+import hmac
+import binascii
+import hashlib
+

 
 from .common import InfoExtractor
 
 from .common import InfoExtractor

-from ..utils import (
+from ..compat import (

     compat_str,
     compat_str,

+)
+from ..utils import (

     determine_ext,
     ExtractorError,
     xpath_with_ns,
     determine_ext,
     ExtractorError,
     xpath_with_ns,

+    unsmuggle_url,
+    int_or_none,

 )
 
 _x = lambda p: xpath_with_ns(p, {'smil': 'http://www.w3.org/2005/SMIL21/Language'})
 )
 
 _x = lambda p: xpath_with_ns(p, {'smil': 'http://www.w3.org/2005/SMIL21/Language'})


@@ -16,11 +25,11 @@ _x = lambda p: xpath_with_ns(p, {'smil': 'http://www.w3.org/2005/SMIL21/Language
 
 class ThePlatformIE(InfoExtractor):
     _VALID_URL = r'''(?x)
 
 class ThePlatformIE(InfoExtractor):
     _VALID_URL = r'''(?x)

-        (?:https?://(?:link|player)\.theplatform\.com/[sp]/[^/]+/
-           (?P<config>(?:[^/\?]+/(?:swf|config)|onsite)/select/)?
+        (?:https?://(?:link|player)\.theplatform\.com/[sp]/(?P<provider_id>[^/]+)/
+           (?:(?P<media>(?:[^/]+/)+select/media/)|(?P<config>(?:[^/\?]+/(?:swf|config)|onsite)/select/))?

          |theplatform:)(?P<id>[^/\?&]+)'''
 
          |theplatform:)(?P<id>[^/\?&]+)'''
 

-    _TEST = {
+    _TESTS = [{

         # from http://www.metacafe.com/watch/cb-e9I_cZgTgIPd/blackberrys_big_bold_z30/
         'url': 'http://link.theplatform.com/s/dJ5BDC/e9I_cZgTgIPd/meta.smil?format=smil&Tracking=true&mbr=true',
         'info_dict': {
         # from http://www.metacafe.com/watch/cb-e9I_cZgTgIPd/blackberrys_big_bold_z30/
         'url': 'http://link.theplatform.com/s/dJ5BDC/e9I_cZgTgIPd/meta.smil?format=smil&Tracking=true&mbr=true',
         'info_dict': {


@@ -34,40 +43,110 @@ class ThePlatformIE(InfoExtractor):
             # rtmp download
             'skip_download': True,
         },
             # rtmp download
             'skip_download': True,
         },

-    }
+    }, {
+        # from http://www.cnet.com/videos/tesla-model-s-a-second-step-towards-a-cleaner-motoring-future/
+        'url': 'http://link.theplatform.com/s/kYEXFC/22d_qsQ6MIRT',
+        'info_dict': {
+            'id': '22d_qsQ6MIRT',
+            'ext': 'flv',
+            'description': 'md5:ac330c9258c04f9d7512cf26b9595409',
+            'title': 'Tesla Model S: A second step towards a cleaner motoring future',
+        },
+        'params': {
+            # rtmp download
+            'skip_download': True,
+        }
+    }, {
+        'url': 'https://player.theplatform.com/p/D6x-PC/pulse_preview/embed/select/media/yMBg9E8KFxZD',
+        'info_dict': {
+            'id': 'yMBg9E8KFxZD',
+            'ext': 'mp4',
+            'description': 'md5:644ad9188d655b742f942bf2e06b002d',
+            'title': 'HIGHLIGHTS: USA bag first ever series Cup win',
+        }
+    }, {
+        'url': 'http://player.theplatform.com/p/NnzsPC/widget/select/media/4Y0TlYUr_ZT7',
+        'only_matching': True,
+    }]
+
+    @staticmethod
+    def _sign_url(url, sig_key, sig_secret, life=600, include_qs=False):
+        flags = '10' if include_qs else '00'
+        expiration_date = '%x' % (int(time.time()) + life)
+
+        def str_to_hex(str):
+            return binascii.b2a_hex(str.encode('ascii')).decode('ascii')
+
+        def hex_to_str(hex):
+            return binascii.a2b_hex(hex)
+
+        relative_path = url.split('http://link.theplatform.com/s/')[1].split('?')[0]
+        clear_text = hex_to_str(flags + expiration_date + str_to_hex(relative_path))
+        checksum = hmac.new(sig_key.encode('ascii'), clear_text, hashlib.sha1).hexdigest()
+        sig = flags + expiration_date + checksum + str_to_hex(sig_secret)
+        return '%s&sig=%s' % (url, sig)

 
     def _real_extract(self, url):
 
     def _real_extract(self, url):

+        url, smuggled_data = unsmuggle_url(url, {})
+

         mobj = re.match(self._VALID_URL, url)
         mobj = re.match(self._VALID_URL, url)

+        provider_id = mobj.group('provider_id')

         video_id = mobj.group('id')
         video_id = mobj.group('id')

-        if mobj.group('config'):
+
+        if not provider_id:
+            provider_id = 'dJ5BDC'
+
+        path = provider_id
+        if mobj.group('media'):
+            path += '/media'
+        path += '/' + video_id
+
+        if smuggled_data.get('force_smil_url', False):
+            smil_url = url
+        elif mobj.group('config'):

             config_url = url + '&form=json'
             config_url = config_url.replace('swf/', 'config/')
             config_url = config_url.replace('onsite/', 'onsite/config/')
             config = self._download_json(config_url, video_id, 'Downloading config')
             smil_url = config['releaseUrl'] + '&format=SMIL&formats=MPEG4&manifest=f4m'
         else:
             config_url = url + '&form=json'
             config_url = config_url.replace('swf/', 'config/')
             config_url = config_url.replace('onsite/', 'onsite/config/')
             config = self._download_json(config_url, video_id, 'Downloading config')
             smil_url = config['releaseUrl'] + '&format=SMIL&formats=MPEG4&manifest=f4m'
         else:

-            smil_url = ('http://link.theplatform.com/s/dJ5BDC/{0}/meta.smil?'
-                        'format=smil&mbr=true'.format(video_id))
+            smil_url = 'http://link.theplatform.com/s/%s/meta.smil?format=smil&mbr=true' % path
+
+        sig = smuggled_data.get('sig')
+        if sig:
+            smil_url = self._sign_url(smil_url, sig['key'], sig['secret'])

 
         meta = self._download_xml(smil_url, video_id)
         try:
             error_msg = next(
                 n.attrib['abstract']
                 for n in meta.findall(_x('.//smil:ref'))
 
         meta = self._download_xml(smil_url, video_id)
         try:
             error_msg = next(
                 n.attrib['abstract']
                 for n in meta.findall(_x('.//smil:ref'))

-                if n.attrib.get('title') == 'Geographic Restriction')
+                if n.attrib.get('title') == 'Geographic Restriction' or n.attrib.get('title') == 'Expired')

         except StopIteration:
             pass
         else:
             raise ExtractorError(error_msg, expected=True)
 
         except StopIteration:
             pass
         else:
             raise ExtractorError(error_msg, expected=True)
 

-        info_url = 'http://link.theplatform.com/s/dJ5BDC/{0}?format=preview'.format(video_id)
+        info_url = 'http://link.theplatform.com/s/%s?format=preview' % path

         info_json = self._download_webpage(info_url, video_id)
         info = json.loads(info_json)
 
         info_json = self._download_webpage(info_url, video_id)
         info = json.loads(info_json)
 

+        subtitles = {}
+        captions = info.get('captions')
+        if isinstance(captions, list):
+            for caption in captions:
+                lang, src, mime = caption.get('lang', 'en'), caption.get('src'), caption.get('type')
+                subtitles[lang] = [{
+                    'ext': 'srt' if mime == 'text/srt' else 'ttml',
+                    'url': src,
+                }]
+

         head = meta.find(_x('smil:head'))
         body = meta.find(_x('smil:body'))
 
         f4m_node = body.find(_x('smil:seq//smil:video'))
         head = meta.find(_x('smil:head'))
         body = meta.find(_x('smil:body'))
 
         f4m_node = body.find(_x('smil:seq//smil:video'))

+        if f4m_node is None:
+            f4m_node = body.find(_x('smil:seq/smil:video'))

         if f4m_node is not None and '.f4m' in f4m_node.attrib['src']:
             f4m_url = f4m_node.attrib['src']
             if 'manifest.f4m?' not in f4m_url:
         if f4m_node is not None and '.f4m' in f4m_node.attrib['src']:
             f4m_url = f4m_node.attrib['src']
             if 'manifest.f4m?' not in f4m_url:


@@ -79,13 +158,19 @@ class ThePlatformIE(InfoExtractor):
         else:
             formats = []
             switch = body.find(_x('smil:switch'))
         else:
             formats = []
             switch = body.find(_x('smil:switch'))

+            if switch is None:
+                switch = body.find(_x('smil:par//smil:switch'))
+            if switch is None:
+                switch = body.find(_x('smil:par/smil:switch'))
+            if switch is None:
+                switch = body.find(_x('smil:par'))

             if switch is not None:
                 base_url = head.find(_x('smil:meta')).attrib['base']
                 for f in switch.findall(_x('smil:video')):
                     attr = f.attrib
             if switch is not None:
                 base_url = head.find(_x('smil:meta')).attrib['base']
                 for f in switch.findall(_x('smil:video')):
                     attr = f.attrib

-                    width = int(attr['width'])
-                    height = int(attr['height'])
-                    vbr = int(attr['system-bitrate']) // 1000
+                    width = int_or_none(attr.get('width'))
+                    height = int_or_none(attr.get('height'))
+                    vbr = int_or_none(attr.get('system-bitrate'), 1000)

                     format_id = '%dx%d_%dk' % (width, height, vbr)
                     formats.append({
                         'format_id': format_id,
                     format_id = '%dx%d_%dk' % (width, height, vbr)
                     formats.append({
                         'format_id': format_id,


@@ -98,9 +183,11 @@ class ThePlatformIE(InfoExtractor):
                     })
             else:
                 switch = body.find(_x('smil:seq//smil:switch'))
                     })
             else:
                 switch = body.find(_x('smil:seq//smil:switch'))

+                if switch is None:
+                    switch = body.find(_x('smil:seq/smil:switch'))

                 for f in switch.findall(_x('smil:video')):
                     attr = f.attrib
                 for f in switch.findall(_x('smil:video')):
                     attr = f.attrib

-                    vbr = int(attr['system-bitrate']) // 1000
+                    vbr = int_or_none(attr.get('system-bitrate'), 1000)

                     ext = determine_ext(attr['src'])
                     if ext == 'once':
                         ext = 'mp4'
                     ext = determine_ext(attr['src'])
                     if ext == 'once':
                         ext = 'mp4'


@@ -115,8 +202,9 @@ class ThePlatformIE(InfoExtractor):
         return {
             'id': video_id,
             'title': info['title'],
         return {
             'id': video_id,
             'title': info['title'],

+            'subtitles': subtitles,

             'formats': formats,
             'description': info['description'],
             'thumbnail': info['defaultThumbnailUrl'],
             'formats': formats,
             'description': info['description'],
             'thumbnail': info['defaultThumbnailUrl'],

-            'duration': info['duration'] // 1000,
+            'duration': int_or_none(info.get('duration'), 1000),

         }
         }