Merge pull request #5588 from aajanki/encode_frag_filenames

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

diff --git a/youtube_dl/extractor/theplatform.py b/youtube_dl/extractor/theplatform.py
index f02e0f58d5752000ab0036be78c293cbc4b4ce62..25edc310008ef0da7b407c97db86ce5e49c78a50 100644 (file)
--- a/youtube_dl/extractor/theplatform.py
+++ b/youtube_dl/extractor/theplatform.py
@@ -1,7 +1,7 @@
+# -*- coding: utf-8 -*-
 from __future__ import unicode_literals
 
 import re
-import json
 import time
 import hmac
 import binascii
@@ -9,6 +9,10 @@ import hashlib
 
 
 from .common import InfoExtractor
+from ..compat import (
+    compat_parse_qs,
+    compat_urllib_parse_urlparse,
+)
 from ..utils import (
     determine_ext,
     ExtractorError,
@@ -24,7 +28,7 @@ _x = lambda p: xpath_with_ns(p, {'smil': default_ns})
 
 
 class ThePlatformBaseIE(InfoExtractor):
-    def _extract_theplatform_smil_formats(self, smil_url, video_id, note='Downloading SMIL data'):
+    def _extract_theplatform_smil(self, smil_url, video_id, note='Downloading SMIL data'):
         meta = self._download_xml(smil_url, video_id, note=note)
         try:
             error_msg = next(
@@ -50,12 +54,13 @@ class ThePlatformBaseIE(InfoExtractor):
 
         self._sort_formats(formats)
 
-        return formats
+        subtitles = self._parse_smil_subtitles(meta, default_ns)
+
+        return formats, subtitles
 
     def get_metadata(self, path, 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 = self._download_json(info_url, video_id)
 
         subtitles = {}
         captions = info.get('captions')
@@ -120,6 +125,20 @@ class ThePlatformIE(ThePlatformBaseIE):
     }, {
         'url': 'http://player.theplatform.com/p/NnzsPC/widget/select/media/4Y0TlYUr_ZT7',
         'only_matching': True,
+    }, {
+        'url': 'http://player.theplatform.com/p/2E2eJC/nbcNewsOffsite?guid=tdy_or_siri_150701',
+        'md5': '734f3790fb5fc4903da391beeebc4836',
+        'info_dict': {
+            'id': 'tdy_or_siri_150701',
+            'ext': 'mp4',
+            'title': 'iPhone Siri’s sassy response to a math question has people talking',
+            'description': 'md5:a565d1deadd5086f3331d57298ec6333',
+            'duration': 83.0,
+            'thumbnail': 're:^https?://.*\.jpg$',
+            'timestamp': 1435752600,
+            'upload_date': '20150701',
+            'categories': ['Today/Shows/Orange Room', 'Today/Sections/Money', 'Today/Topics/Tech', "Today/Topics/Editor's picks"],
+        },
     }]
 
     @staticmethod
@@ -154,6 +173,24 @@ class ThePlatformIE(ThePlatformBaseIE):
             path += '/media'
         path += '/' + video_id
 
+        qs_dict = compat_parse_qs(compat_urllib_parse_urlparse(url).query)
+        if 'guid' in qs_dict:
+            webpage = self._download_webpage(url, video_id)
+            scripts = re.findall(r'<script[^>]+src="([^"]+)"', webpage)
+            feed_id = None
+            # feed id usually locates in the last script.
+            # Seems there's no pattern for the interested script filename, so
+            # I try one by one
+            for script in reversed(scripts):
+                feed_script = self._download_webpage(script, video_id, 'Downloading feed script')
+                feed_id = self._search_regex(r'defaultFeedId\s*:\s*"([^"]+)"', feed_script, 'default feed id', default=None)
+                if feed_id is not None:
+                    break
+            if feed_id is None:
+                raise ExtractorError('Unable to find feed id')
+            return self.url_result('http://feed.theplatform.com/f/%s/%s?byGuid=%s' % (
+                provider_id, feed_id, qs_dict['guid'][0]))
+
         if smuggled_data.get('force_smil_url', False):
             smil_url = url
         elif mobj.group('config'):
@@ -173,12 +210,14 @@ class ThePlatformIE(ThePlatformBaseIE):
         if sig:
             smil_url = self._sign_url(smil_url, sig['key'], sig['secret'])
 
-        formats = self._extract_theplatform_smil_formats(smil_url, video_id)
+        formats, subtitles = self._extract_theplatform_smil(smil_url, video_id)
 
         ret = self.get_metadata(path, video_id)
+        combined_subtitles = self._merge_subtitles(ret.get('subtitles', {}), subtitles)
         ret.update({
             'id': video_id,
             'formats': formats,
+            'subtitles': combined_subtitles,
         })
 
         return ret
@@ -216,6 +255,7 @@ class ThePlatformFeedIE(ThePlatformBaseIE):
         entry = feed['entries'][0]
 
         formats = []
+        subtitles = {}
         first_video_id = None
         duration = None
         for item in entry['media$content']:
@@ -224,7 +264,9 @@ class ThePlatformFeedIE(ThePlatformBaseIE):
             if first_video_id is None:
                 first_video_id = cur_video_id
                 duration = float_or_none(item.get('plfile$duration'))
-            formats.extend(self._extract_theplatform_smil_formats(smil_url, video_id, 'Downloading SMIL data for %s' % cur_video_id))
+            cur_formats, cur_subtitles = self._extract_theplatform_smil(smil_url, video_id, 'Downloading SMIL data for %s' % cur_video_id)
+            formats.extend(cur_formats)
+            subtitles = self._merge_subtitles(subtitles, cur_subtitles)
 
         self._sort_formats(formats)
 
@@ -238,9 +280,11 @@ class ThePlatformFeedIE(ThePlatformBaseIE):
         categories = [item['media$name'] for item in entry.get('media$categories', [])]
 
         ret = self.get_metadata('%s/%s' % (provider_id, first_video_id), video_id)
+        subtitles = self._merge_subtitles(subtitles, ret['subtitles'])
         ret.update({
             'id': video_id,
             'formats': formats,
+            'subtitles': subtitles,
             'thumbnails': thumbnails,
             'duration': duration,
             'timestamp': timestamp,