Merge branch 'master' into GoogleDrive-issue-13619

test/test_YoutubeDL.py

@@ -527,6 +527,8 @@ class TestYoutubeDL(unittest.TestCase):
             'ext': 'mp4',
             'width': None,
             'height': 1080,
+            'title1': '$PATH',
+            'title2': '%PATH%',
         }
 
         def fname(templ):
@@ -545,10 +547,14 @@ class TestYoutubeDL(unittest.TestCase):
         self.assertEqual(fname('%(height)0 6d.%(ext)s'), ' 01080.mp4')
         self.assertEqual(fname('%(height)0   6d.%(ext)s'), ' 01080.mp4')
         self.assertEqual(fname('%(height)   0   6d.%(ext)s'), ' 01080.mp4')
+        self.assertEqual(fname('%%'), '%')
+        self.assertEqual(fname('%%%%'), '%%')
         self.assertEqual(fname('%%(height)06d.%(ext)s'), '%(height)06d.mp4')
         self.assertEqual(fname('%(width)06d.%(ext)s'), 'NA.mp4')
         self.assertEqual(fname('%(width)06d.%%(ext)s'), 'NA.%(ext)s')
         self.assertEqual(fname('%%(width)06d.%(ext)s'), '%(width)06d.mp4')
+        self.assertEqual(fname('Hello %(title1)s'), 'Hello $PATH')
+        self.assertEqual(fname('Hello %(title2)s'), 'Hello %PATH%')
 
     def test_format_note(self):
         ydl = YoutubeDL()

youtube_dl/YoutubeDL.py

@@ -20,6 +20,7 @@ import re
 import shutil
 import subprocess
 import socket
+import string
 import sys
 import time
 import tokenize
@@ -674,7 +675,19 @@ class YoutubeDL(object):
                         FORMAT_RE.format(numeric_field),
                         r'%({0})s'.format(numeric_field), outtmpl)
 
-            filename = expand_path(outtmpl % template_dict)
+            # expand_path translates '%%' into '%' and '$$' into '$'
+            # correspondingly that is not what we want since we need to keep
+            # '%%' intact for template dict substitution step. Working around
+            # with boundary-alike separator hack.
+            sep = ''.join([random.choice(string.ascii_letters) for _ in range(32)])
+            outtmpl = outtmpl.replace('%%', '%{0}%'.format(sep)).replace('$$', '${0}$'.format(sep))
+
+            # outtmpl should be expand_path'ed before template dict substitution
+            # because meta fields may contain env variables we don't want to
+            # be expanded. For example, for outtmpl "%(title)s.%(ext)s" and
+            # title "Hello $PATH", we don't want `$PATH` to be expanded.
+            filename = expand_path(outtmpl).replace(sep, '') % template_dict
+
             # Temporary fix for #4787
             # 'Treat' all problem characters by passing filename through preferredencoding
             # to workaround encoding issues with subprocess on python2 @ Windows

youtube_dl/downloader/ism.py

@@ -98,7 +98,7 @@ def write_piff_header(stream, params):
 
     if is_audio:
         smhd_payload = s88.pack(0)  # balance
-        smhd_payload = u16.pack(0)  # reserved
+        smhd_payload += u16.pack(0)  # reserved
         media_header_box = full_box(b'smhd', 0, 0, smhd_payload)  # Sound Media Header
     else:
         vmhd_payload = u16.pack(0)  # graphics mode
@@ -126,7 +126,6 @@ def write_piff_header(stream, params):
         if fourcc == 'AACL':
             sample_entry_box = box(b'mp4a', sample_entry_payload)
     else:
-        sample_entry_payload = sample_entry_payload
         sample_entry_payload += u16.pack(0)  # pre defined
         sample_entry_payload += u16.pack(0)  # reserved
         sample_entry_payload += u32.pack(0) * 3  # pre defined

youtube_dl/extractor/audioboom.py

@@ -43,7 +43,7 @@ class AudioBoomIE(InfoExtractor):
 
         def from_clip(field):
             if clip:
-                clip.get(field)
+                return clip.get(field)
 
         audio_url = from_clip('clipURLPriorToLoading') or self._og_search_property(
             'audio', webpage, 'audio url')

youtube_dl/extractor/fivetv.py

@@ -43,7 +43,7 @@ class FiveTVIE(InfoExtractor):
         'info_dict': {
             'id': 'glavnoe',
             'ext': 'mp4',
-            'title': 'Итоги недели с 8 по 14 июня 2015 года',
+            'title': r're:^Итоги недели с \d+ по \d+ \w+ \d{4} года$',
             'thumbnail': r're:^https?://.*\.jpg$',
         },
     }, {
@@ -70,7 +70,8 @@ class FiveTVIE(InfoExtractor):
         webpage = self._download_webpage(url, video_id)
 
         video_url = self._search_regex(
-            r'<a[^>]+?href="([^"]+)"[^>]+?class="videoplayer"',
+            [r'<div[^>]+?class="flowplayer[^>]+?data-href="([^"]+)"',
+             r'<a[^>]+?href="([^"]+)"[^>]+?class="videoplayer"'],
             webpage, 'video url')
 
         title = self._og_search_title(webpage, default=None) or self._search_regex(

youtube_dl/extractor/generic.py

@@ -2740,7 +2740,7 @@ class GenericIE(InfoExtractor):
         rutube_urls = RutubeIE._extract_urls(webpage)
         if rutube_urls:
             return self.playlist_from_matches(
-                rutube_urls, ie=RutubeIE.ie_key())
+                rutube_urls, video_id, video_title, ie=RutubeIE.ie_key())
 
         # Look for WashingtonPost embeds
         wapo_urls = WashingtonPostIE._extract_urls(webpage)

youtube_dl/extractor/karrierevideos.py

@@ -48,7 +48,7 @@ class KarriereVideosIE(InfoExtractor):
         webpage = self._download_webpage(url, video_id)
 
         title = (self._html_search_meta('title', webpage, default=None) or
-                 self._search_regex(r'<h1 class="title">([^<]+)</h1>'))
+                 self._search_regex(r'<h1 class="title">([^<]+)</h1>', webpage, 'video title'))
 
         video_id = self._search_regex(
             r'/config/video/(.+?)\.xml', webpage, 'video id')

youtube_dl/extractor/slideshare.py

@@ -31,7 +31,7 @@ class SlideshareIE(InfoExtractor):
         page_title = mobj.group('title')
         webpage = self._download_webpage(url, page_title)
         slideshare_obj = self._search_regex(
-            r'\$\.extend\(slideshare_object,\s*(\{.*?\})\);',
+            r'\$\.extend\(.*?slideshare_object,\s*(\{.*?\})\);',
             webpage, 'slideshare object')
         info = json.loads(slideshare_obj)
         if info['slideshow']['type'] != 'video':

youtube_dl/extractor/youtube.py

@@ -673,6 +673,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
             },
         },
         # video_info is None (https://github.com/rg3/youtube-dl/issues/4421)
+        # YouTube Red ad is not captured for creator
         {
             'url': '__2ABJjxzNo',
             'info_dict': {
@@ -1649,7 +1650,21 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
             video_webpage, 'license', default=None)
 
         m_music = re.search(
-            r'<h4[^>]+class="title"[^>]*>\s*Music\s*</h4>\s*<ul[^>]*>\s*<li>(?P<title>.+?) by (?P<creator>.+?)(?:\(.+?\))?</li',
+            r'''(?x)
+                <h4[^>]+class="title"[^>]*>\s*Music\s*</h4>\s*
+                <ul[^>]*>\s*
+                <li>(?P<title>.+?)
+                by (?P<creator>.+?)
+                (?:
+                    \(.+?\)|
+                    <a[^>]*
+                        (?:
+                            \bhref=["\']/red[^>]*>|             # drop possible
+                            >\s*Listen ad-free with YouTube Red # YouTube Red ad 
+                        )
+                    .*?
+                )?</li
+            ''',
             video_webpage)
         if m_music:
             video_alt_title = remove_quotes(unescapeHTML(m_music.group('title')))