From 3d5f7a3947a8d304bc7ad46217f171996e95c475 Mon Sep 17 00:00:00 2001
From: Johannes Knoedtel <johannes@dev-baron.de>
Date: Mon, 12 Jan 2015 22:26:20 +0100
Subject: [PATCH 01/40] [utils] Prevent override of custom headers.

The dict of headers of request objects in urllib has its keys always
capitalized.

This causes the lookup to fail and overwrite the header. If for example
a Extractor tries to add a "User-Agent" header the internal
representation in the request object is "User-agent". The header is
therefore clobbered by the "User-Agent" in std_headers, because the
strings are not equal.
---
 youtube_dl/utils.py | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/youtube_dl/utils.py b/youtube_dl/utils.py
index 98732e8e9..daf94abd1 100644
--- a/youtube_dl/utils.py
+++ b/youtube_dl/utils.py
@@ -611,7 +611,9 @@ class YoutubeDLHandler(compat_urllib_request.HTTPHandler):
 
     def http_request(self, req):
         for h, v in std_headers.items():
-            if h not in req.headers:
+            # Capitalize is needed because of Python bug 2275: http://bugs.python.org/issue2275
+            # The dict keys are capitalized because of this bug by urllib
+            if h.capitalize() not in req.headers:
                 req.add_header(h, v)
         if 'Youtubedl-no-compression' in req.headers:
             if 'Accept-encoding' in req.headers:

From e205db3bcd7f7a9b0beaa13a4774bdcb26091e2f Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Jaime=20Marqui=CC=81nez=20Ferra=CC=81ndiz?=
 <jaime.marquinez.ferrandiz@gmail.com>
Date: Fri, 16 Jan 2015 13:29:01 +0100
Subject: [PATCH 02/40] FFmpegEmbedSubtitlePP: don't fail if the video doesn't
 have an audio stream (fixes #4718)

Instead of specifying which streams ffmpeg must copy, we tell it to copy all.
---
 youtube_dl/postprocessor/ffmpeg.py | 8 +++++++-
 1 file changed, 7 insertions(+), 1 deletion(-)

diff --git a/youtube_dl/postprocessor/ffmpeg.py b/youtube_dl/postprocessor/ffmpeg.py
index 26b99e43c..a75277778 100644
--- a/youtube_dl/postprocessor/ffmpeg.py
+++ b/youtube_dl/postprocessor/ffmpeg.py
@@ -475,7 +475,13 @@ class FFmpegEmbedSubtitlePP(FFmpegPostProcessor):
         filename = information['filepath']
         input_files = [filename] + [subtitles_filename(filename, lang, self._subformat) for lang in sub_langs]
 
-        opts = ['-map', '0:0', '-map', '0:1', '-c:v', 'copy', '-c:a', 'copy']
+        opts = [
+            '-map', '0',
+            '-c', 'copy',
+            # Don't copy the existing subtitles, we may be running the
+            # postprocessor a second time
+            '-map', '-0:s',
+        ]
         for (i, lang) in enumerate(sub_langs):
             opts.extend(['-map', '%d:0' % (i + 1), '-c:s:%d' % i, 'mov_text'])
             lang_code = self._conver_lang_code(lang)

From 2875cf01bb2229cb6cceeac72fb6dd770f7a96d2 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Jaime=20Marqui=CC=81nez=20Ferra=CC=81ndiz?=
 <jaime.marquinez.ferrandiz@gmail.com>
Date: Fri, 16 Jan 2015 13:37:37 +0100
Subject: [PATCH 03/40] FFmpegEmbedSubtitlePP: simplify command

---
 youtube_dl/postprocessor/ffmpeg.py | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/youtube_dl/postprocessor/ffmpeg.py b/youtube_dl/postprocessor/ffmpeg.py
index a75277778..5b0ff32b1 100644
--- a/youtube_dl/postprocessor/ffmpeg.py
+++ b/youtube_dl/postprocessor/ffmpeg.py
@@ -481,15 +481,15 @@ class FFmpegEmbedSubtitlePP(FFmpegPostProcessor):
             # Don't copy the existing subtitles, we may be running the
             # postprocessor a second time
             '-map', '-0:s',
+            '-c:s', 'mov_text',
         ]
         for (i, lang) in enumerate(sub_langs):
-            opts.extend(['-map', '%d:0' % (i + 1), '-c:s:%d' % i, 'mov_text'])
+            opts.extend(['-map', '%d:0' % (i + 1)])
             lang_code = self._conver_lang_code(lang)
             if lang_code is not None:
                 opts.extend(['-metadata:s:s:%d' % i, 'language=%s' % lang_code])
-        opts.extend(['-f', 'mp4'])
 
-        temp_filename = filename + '.temp'
+        temp_filename = prepend_extension(filename, 'temp')
         self._downloader.to_screen('[ffmpeg] Embedding subtitles in \'%s\'' % filename)
         self.run_ffmpeg_multiple_files(input_files, temp_filename, opts)
         os.remove(encodeFilename(filename))

From 4f4f6428224b641eb97ffd4b90be9d67034fc64a Mon Sep 17 00:00:00 2001
From: Philipp Hagemeister <phihag@phihag.de>
Date: Fri, 16 Jan 2015 13:44:36 +0100
Subject: [PATCH 04/40] [npo] Remove unused import

---
 youtube_dl/extractor/npo.py | 2 --
 1 file changed, 2 deletions(-)

diff --git a/youtube_dl/extractor/npo.py b/youtube_dl/extractor/npo.py
index 8d496488b..175b14583 100644
--- a/youtube_dl/extractor/npo.py
+++ b/youtube_dl/extractor/npo.py
@@ -1,7 +1,5 @@
 from __future__ import unicode_literals
 
-import re
-
 from .common import InfoExtractor
 from ..utils import (
     fix_xml_ampersands,

From 9d22a7dfb07ad6ff23680e11238744b2bb1735aa Mon Sep 17 00:00:00 2001
From: Philipp Hagemeister <phihag@phihag.de>
Date: Fri, 16 Jan 2015 13:44:44 +0100
Subject: [PATCH 05/40] [fourtube] Fix extraction

---
 youtube_dl/extractor/fourtube.py | 76 ++++++++++++++++++--------------
 youtube_dl/utils.py              |  2 +-
 2 files changed, 45 insertions(+), 33 deletions(-)

diff --git a/youtube_dl/extractor/fourtube.py b/youtube_dl/extractor/fourtube.py
index 7187e0752..b2284ab01 100644
--- a/youtube_dl/extractor/fourtube.py
+++ b/youtube_dl/extractor/fourtube.py
@@ -7,10 +7,9 @@ from ..compat import (
     compat_urllib_request,
 )
 from ..utils import (
-    clean_html,
     parse_duration,
+    parse_iso8601,
     str_to_int,
-    unified_strdate,
 )
 
 
@@ -28,68 +27,81 @@ class FourTubeIE(InfoExtractor):
             'uploader': 'WCP Club',
             'uploader_id': 'wcp-club',
             'upload_date': '20131031',
+            'timestamp': 1383263892,
             'duration': 583,
+            'view_count': int,
+            'like_count': int,
+            'categories': list,
         }
     }
 
     def _real_extract(self, url):
         video_id = self._match_id(url)
-        webpage_url = 'http://www.4tube.com/videos/' + video_id
-        webpage = self._download_webpage(webpage_url, video_id)
+        webpage = self._download_webpage(url, video_id)
 
-        self.report_extraction(video_id)
+        title = self._html_search_meta('name', webpage)
+        timestamp = parse_iso8601(self._html_search_meta(
+            'uploadDate', webpage))
+        thumbnail = self._html_search_meta('thumbnailUrl', webpage)
+        uploader_id = self._html_search_regex(
+            r'<a class="img-avatar" href="[^"]+/channels/([^/"]+)" title="Go to [^"]+ page">',
+            webpage, 'uploader id')
+        uploader = self._html_search_regex(
+            r'<a class="img-avatar" href="[^"]+/channels/[^/"]+" title="Go to ([^"]+) page">',
+            webpage, 'uploader')
 
-        playlist_json = self._html_search_regex(r'var playerConfigPlaylist\s+=\s+([^;]+)', webpage, 'Playlist')
-        media_id = self._search_regex(r'idMedia:\s*(\d+)', playlist_json, 'Media Id')
-        sources = self._search_regex(r'sources:\s*\[([^\]]*)\]', playlist_json, 'Sources').split(',')
-        title = self._search_regex(r'title:\s*"([^"]*)', playlist_json, 'Title')
-        thumbnail_url = self._search_regex(r'image:\s*"([^"]*)', playlist_json, 'Thumbnail', fatal=False)
+        categories_html = self._search_regex(
+            r'(?s)><i class="icon icon-tag"></i>\s*Categories / Tags\s*.*?<ul class="list">(.*?)</ul>',
+            webpage, 'categories', fatal=False)
+        categories = None
+        if categories_html:
+            categories = [
+                c.strip() for c in re.findall(
+                    r'(?s)<li><a.*?>(.*?)</a>', categories_html)]
 
-        uploader_str = self._search_regex(r'<span>Uploaded by</span>(.*?)<span>', webpage, 'uploader', fatal=False)
-        mobj = re.search(r'<a href="/sites/(?P<id>[^"]+)"><strong>(?P<name>[^<]+)</strong></a>', uploader_str)
-        (uploader, uploader_id) = (mobj.group('name'), mobj.group('id')) if mobj else (clean_html(uploader_str), None)
+        view_count = str_to_int(self._search_regex(
+            r'<meta itemprop="interactionCount" content="UserPlays:([0-9,]+)">',
+            webpage, 'view count', fatal=False))
+        like_count = str_to_int(self._search_regex(
+            r'<meta itemprop="interactionCount" content="UserLikes:([0-9,]+)">',
+            webpage, 'like count', fatal=False))
+        duration = parse_duration(self._html_search_meta('duration', webpage))
 
-        upload_date = None
-        view_count = None
-        duration = None
-        description = self._html_search_meta('description', webpage, 'description')
-        if description:
-            upload_date = self._search_regex(r'Published Date: (\d{2} [a-zA-Z]{3} \d{4})', description, 'upload date',
-                                             fatal=False)
-            if upload_date:
-                upload_date = unified_strdate(upload_date)
-            view_count = self._search_regex(r'Views: ([\d,\.]+)', description, 'view count', fatal=False)
-            if view_count:
-                view_count = str_to_int(view_count)
-            duration = parse_duration(self._search_regex(r'Length: (\d+m\d+s)', description, 'duration', fatal=False))
+        params_js = self._search_regex(
+            r'\$\.ajax\(url,\ opts\);\s*\}\s*\}\)\(([0-9,\[\] ]+)\)',
+            webpage, 'initialization parameters'
+        )
+        params = self._parse_json('[%s]' % params_js, video_id)
+        media_id = params[0]
+        sources = ['%s' % p for p in params[2]]
 
-        token_url = "http://tkn.4tube.com/{0}/desktop/{1}".format(media_id, "+".join(sources))
+        token_url = 'http://tkn.4tube.com/{0}/desktop/{1}'.format(
+            media_id, '+'.join(sources))
         headers = {
             b'Content-Type': b'application/x-www-form-urlencoded',
             b'Origin': b'http://www.4tube.com',
         }
         token_req = compat_urllib_request.Request(token_url, b'{}', headers)
         tokens = self._download_json(token_req, video_id)
-
         formats = [{
             'url': tokens[format]['token'],
             'format_id': format + 'p',
             'resolution': format + 'p',
             'quality': int(format),
         } for format in sources]
-
         self._sort_formats(formats)
 
         return {
             'id': video_id,
             'title': title,
             'formats': formats,
-            'thumbnail': thumbnail_url,
+            'categories': categories,
+            'thumbnail': thumbnail,
             'uploader': uploader,
             'uploader_id': uploader_id,
-            'upload_date': upload_date,
+            'timestamp': timestamp,
+            'like_count': like_count,
             'view_count': view_count,
             'duration': duration,
             'age_limit': 18,
-            'webpage_url': webpage_url,
         }
diff --git a/youtube_dl/utils.py b/youtube_dl/utils.py
index 2546fa45d..7832ed87f 100644
--- a/youtube_dl/utils.py
+++ b/youtube_dl/utils.py
@@ -1277,7 +1277,7 @@ def parse_duration(s):
     s = s.strip()
 
     m = re.match(
-        r'''(?ix)T?
+        r'''(?ix)(?:P?T)?
         (?:
             (?P<only_mins>[0-9.]+)\s*(?:mins?|minutes?)\s*|
             (?P<only_hours>[0-9.]+)\s*(?:hours?)|

From cce81f192c15f3766adb7340315a877bc4fc592b Mon Sep 17 00:00:00 2001
From: Philipp Hagemeister <phihag@phihag.de>
Date: Fri, 16 Jan 2015 14:20:25 +0100
Subject: [PATCH 06/40] [bandcamp:album] Fix title extraction (Fixes #4721)

---
 youtube_dl/extractor/bandcamp.py | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/youtube_dl/extractor/bandcamp.py b/youtube_dl/extractor/bandcamp.py
index b45d68a61..aea0263d6 100644
--- a/youtube_dl/extractor/bandcamp.py
+++ b/youtube_dl/extractor/bandcamp.py
@@ -161,7 +161,8 @@ class BandcampAlbumIE(InfoExtractor):
         entries = [
             self.url_result(compat_urlparse.urljoin(url, t_path), ie=BandcampIE.ie_key())
             for t_path in tracks_paths]
-        title = self._search_regex(r'album_title : "(.*?)"', webpage, 'title')
+        title = self._search_regex(
+            r'album_title\s*:\s*"(.*?)"', webpage, 'title', fatal=False)
         return {
             '_type': 'playlist',
             'id': playlist_id,

From 910c55205293097e79b2a134b69b572984e37846 Mon Sep 17 00:00:00 2001
From: Philipp Hagemeister <phihag@phihag.de>
Date: Fri, 16 Jan 2015 14:20:38 +0100
Subject: [PATCH 07/40] release 2015.01.16

---
 youtube_dl/version.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/youtube_dl/version.py b/youtube_dl/version.py
index 458813b2d..63a79a7ee 100644
--- a/youtube_dl/version.py
+++ b/youtube_dl/version.py
@@ -1,3 +1,3 @@
 from __future__ import unicode_literals
 
-__version__ = '2015.01.15.1'
+__version__ = '2015.01.16'

From ba319696a99f34342a321d23bb41d3d442a773cc Mon Sep 17 00:00:00 2001
From: Philipp Hagemeister <phihag@phihag.de>
Date: Sat, 17 Jan 2015 23:56:34 +0100
Subject: [PATCH 08/40] [options] Clarify that --password can be left out
 (#4723)

---
 youtube_dl/options.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/youtube_dl/options.py b/youtube_dl/options.py
index a30974efd..f25c12e52 100644
--- a/youtube_dl/options.py
+++ b/youtube_dl/options.py
@@ -264,7 +264,7 @@ def parseOpts(overrideArguments=None):
     authentication.add_option(
         '-p', '--password',
         dest='password', metavar='PASSWORD',
-        help='account password')
+        help='account password. If this option is left out, youtube-dl will ask interactively.')
     authentication.add_option(
         '-2', '--twofactor',
         dest='twofactor', metavar='TWOFACTOR',

From fdb2ed7455d509a5f0b17ad5c1d721d5484bde8f Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Sergey=20M=E2=80=A4?= <dstftw@gmail.com>
Date: Sun, 18 Jan 2015 08:09:18 +0600
Subject: [PATCH 09/40] [abc7news] Add extractor (Closes #4734)

---
 youtube_dl/extractor/__init__.py |  1 +
 youtube_dl/extractor/abc7news.py | 68 ++++++++++++++++++++++++++++++++
 2 files changed, 69 insertions(+)
 create mode 100644 youtube_dl/extractor/abc7news.py

diff --git a/youtube_dl/extractor/__init__.py b/youtube_dl/extractor/__init__.py
index 0902eb437..2b9d4455d 100644
--- a/youtube_dl/extractor/__init__.py
+++ b/youtube_dl/extractor/__init__.py
@@ -1,6 +1,7 @@
 from __future__ import unicode_literals
 
 from .abc import ABCIE
+from .abc7news import Abc7NewsIE
 from .academicearth import AcademicEarthCourseIE
 from .addanime import AddAnimeIE
 from .adobetv import AdobeTVIE
diff --git a/youtube_dl/extractor/abc7news.py b/youtube_dl/extractor/abc7news.py
new file mode 100644
index 000000000..c04949c21
--- /dev/null
+++ b/youtube_dl/extractor/abc7news.py
@@ -0,0 +1,68 @@
+from __future__ import unicode_literals
+
+import re
+
+from .common import InfoExtractor
+from ..utils import parse_iso8601
+
+
+class Abc7NewsIE(InfoExtractor):
+    _VALID_URL = r'https?://abc7news\.com(?:/[^/]+/(?P<display_id>[^/]+))?/(?P<id>\d+)'
+    _TESTS = [
+        {
+            'url': 'http://abc7news.com/entertainment/east-bay-museum-celebrates-vintage-synthesizers/472581/',
+            'info_dict': {
+                'id': '472581',
+                'display_id': 'east-bay-museum-celebrates-vintage-synthesizers',
+                'ext': 'mp4',
+                'title': 'East Bay museum celebrates history of synthesized music',
+                'description': 'md5:a4f10fb2f2a02565c1749d4adbab4b10',
+                'thumbnail': 're:^https?://.*\.jpg$',
+                'timestamp': 1421123075,
+                'upload_date': '20150113',
+                'uploader': 'Jonathan Bloom',
+            },
+            'params': {
+                # m3u8 download
+                'skip_download': True,
+            },
+        },
+        {
+            'url': 'http://abc7news.com/472581',
+            'only_matching': True,
+        },
+    ]
+
+    def _real_extract(self, url):
+        mobj = re.match(self._VALID_URL, url)
+        video_id = mobj.group('id')
+        display_id = mobj.group('display_id') or video_id
+
+        webpage = self._download_webpage(url, display_id)
+
+        m3u8 = self._html_search_meta(
+            'contentURL', webpage, 'm3u8 url', fatal=True)
+
+        formats = self._extract_m3u8_formats(m3u8, display_id, 'mp4')
+        self._sort_formats(formats)
+
+        title = self._og_search_title(webpage).strip()
+        description = self._og_search_description(webpage).strip()
+        thumbnail = self._og_search_thumbnail(webpage)
+        timestamp = parse_iso8601(self._search_regex(
+            r'<div class="meta">\s*<time class="timeago" datetime="([^"]+)">',
+            webpage, 'upload date', fatal=False))
+        uploader = self._search_regex(
+            r'rel="author">([^<]+)</a>',
+            webpage, 'uploader', default=None)
+
+        return {
+            'id': video_id,
+            'display_id': display_id,
+            'title': title,
+            'description': description,
+            'thumbnail': thumbnail,
+            'timestamp': timestamp,
+            'uploader': uploader,
+            'formats': formats,
+        }

From bd3cbe0716c3cfc4c6ce7a5a338e881a06b19ef8 Mon Sep 17 00:00:00 2001
From: David Powell <bitlooter@users.noreply.github.com>
Date: Sun, 18 Jan 2015 15:01:05 -0800
Subject: [PATCH 10/40] Embed description and URL as MP4 tags

---
 youtube_dl/postprocessor/ffmpeg.py | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/youtube_dl/postprocessor/ffmpeg.py b/youtube_dl/postprocessor/ffmpeg.py
index 5b0ff32b1..faccdc43d 100644
--- a/youtube_dl/postprocessor/ffmpeg.py
+++ b/youtube_dl/postprocessor/ffmpeg.py
@@ -509,6 +509,10 @@ class FFmpegMetadataPP(FFmpegPostProcessor):
             metadata['artist'] = info['uploader']
         elif info.get('uploader_id') is not None:
             metadata['artist'] = info['uploader_id']
+        if info.get('description') is not None:
+            metadata['description'] = info['description']
+        if info.get('webpage_url') is not None:
+            metadata['comment'] = info['webpage_url']
 
         if not metadata:
             self._downloader.to_screen('[ffmpeg] There isn\'t any metadata to add')

From c511f13f223d843dd4affc30c76bf333a6863227 Mon Sep 17 00:00:00 2001
From: Philipp Hagemeister <phihag@phihag.de>
Date: Mon, 19 Jan 2015 10:10:05 +0100
Subject: [PATCH 11/40] [ndtv] Modernize

---
 youtube_dl/extractor/ndtv.py | 4 +---
 1 file changed, 1 insertion(+), 3 deletions(-)

diff --git a/youtube_dl/extractor/ndtv.py b/youtube_dl/extractor/ndtv.py
index 95e7d63aa..2a1ca80df 100644
--- a/youtube_dl/extractor/ndtv.py
+++ b/youtube_dl/extractor/ndtv.py
@@ -27,9 +27,7 @@ class NDTVIE(InfoExtractor):
     }
 
     def _real_extract(self, url):
-        mobj = re.match(self._VALID_URL, url)
-        video_id = mobj.group('id')
-
+        video_id = self._match_id(url)
         webpage = self._download_webpage(url, video_id)
 
         filename = self._search_regex(

From 030aa5d9e7f482ab7baba18a9ff45e41021d60e3 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Sergey=20M=E2=80=A4?= <dstftw@gmail.com>
Date: Mon, 19 Jan 2015 23:00:22 +0600
Subject: [PATCH 12/40] [tvp] Fix extraction

---
 youtube_dl/extractor/tvp.py | 37 +++++++++++++++++--------------------
 1 file changed, 17 insertions(+), 20 deletions(-)

diff --git a/youtube_dl/extractor/tvp.py b/youtube_dl/extractor/tvp.py
index cc26f417a..f57d609d4 100644
--- a/youtube_dl/extractor/tvp.py
+++ b/youtube_dl/extractor/tvp.py
@@ -12,61 +12,59 @@ class TvpIE(InfoExtractor):
 
     _TESTS = [{
         'url': 'http://vod.tvp.pl/filmy-fabularne/filmy-za-darmo/ogniem-i-mieczem/wideo/odc-2/4278035',
+        'md5': 'cdd98303338b8a7f7abab5cd14092bf2',
         'info_dict': {
             'id': '4278035',
             'ext': 'wmv',
             'title': 'Ogniem i mieczem, odc. 2',
-            'description': 'Bohun dowiaduje się o złamaniu przez kniahinię danego mu słowa i wyrusza do Rozłogów. Helenie w ostatniej chwili udaje się uciec dzięki pomocy Zagłoby.',
         },
     }, {
         'url': 'http://vod.tvp.pl/seriale/obyczajowe/czas-honoru/sezon-1-1-13/i-seria-odc-13/194536',
+        'md5': '8aa518c15e5cc32dfe8db400dc921fbb',
         'info_dict': {
             'id': '194536',
             'ext': 'mp4',
             'title': 'Czas honoru, I seria – odc. 13',
-            #  'description': 'WŁADEK\nCzesław prosi Marię o dostarczenie Władkowi zarazki tyfusu. Jeśli zachoruje zostanie przewieziony do szpitala skąd łatwiej będzie go odbić. Czy matka zdecyduje się zarazić syna? Karol odwiedza Wandę przyznaje się, że ją oszukiwał, ale ostrzega też, że grozi jej aresztowanie i nalega, żeby wyjechała z Warszawy. Czy dziewczyna zdecyduje się znów oddalić od ukochanego? Rozpoczyna się akcja odbicia Władka.',
         },
     }, {
         'url': 'http://www.tvp.pl/there-can-be-anything-so-i-shortened-it/17916176',
+        'md5': 'c3b15ed1af288131115ff17a17c19dda',
         'info_dict': {
             'id': '17916176',
             'ext': 'mp4',
             'title': 'TVP Gorzów pokaże filmy studentów z podroży dookoła świata',
         },
-        'params': {
-            # m3u8 download
-            'skip_download': 'true',
-        },
     }, {
         'url': 'http://vod.tvp.pl/seriale/obyczajowe/na-sygnale/sezon-2-27-/odc-39/17834272',
+        'md5': 'c3b15ed1af288131115ff17a17c19dda',
         'info_dict': {
             'id': '17834272',
             'ext': 'mp4',
             'title': 'Na sygnale, odc. 39',
-            'description': 'Ekipa Wiktora ratuje młodą matkę, która spadła ze schodów trzymając na rękach noworodka. Okazuje się, że dziewczyna jest surogatką, a biologiczni rodzice dziecka próbują zmusić ją do oddania synka…',
-        },
-        'params': {
-            # m3u8 download
-            'skip_download': 'true',
         },
     }]
 
     def _real_extract(self, url):
         video_id = self._match_id(url)
+
         webpage = self._download_webpage(
             'http://www.tvp.pl/sess/tvplayer.php?object_id=%s' % video_id, video_id)
 
-        title = self._og_search_title(webpage)
-        series = self._search_regex(
-            r'{name:\s*([\'"])SeriesTitle\1,\s*value:\s*\1(?P<series>.*?)\1},',
+        title = self._search_regex(
+            r'name\s*:\s*([\'"])Title\1\s*,\s*value\s*:\s*\1(?P<title>.+?)\1',
+            webpage, 'title', group='title')
+        series_title = self._search_regex(
+            r'name\s*:\s*([\'"])SeriesTitle\1\s*,\s*value\s*:\s*\1(?P<series>.+?)\1',
             webpage, 'series', group='series', default=None)
-        if series is not None and series not in title:
-            title = '%s, %s' % (series, title)
-        description = self._og_search_description(webpage, default=None)
+        if series_title:
+            title = '%s, %s' % (series_title, title)
+
+        thumbnail = self._search_regex(
+            r"poster\s*:\s*'([^']+)'", webpage, 'thumbnail', default=None)
 
         video_url = self._search_regex(
             r'0:{src:([\'"])(?P<url>.*?)\1', webpage, 'formats', group='url', default=None)
-        if video_url is None:
+        if not video_url:
             video_url = self._download_json(
                 'http://www.tvp.pl/pub/stat/videofileinfo?video_id=%s' % video_id,
                 video_id)['video_url']
@@ -89,8 +87,7 @@ class TvpIE(InfoExtractor):
         return {
             'id': video_id,
             'title': title,
-            'thumbnail': self._og_search_thumbnail(webpage),
-            'description': description,
+            'thumbnail': thumbnail,
             'formats': formats,
         }
 

From 2df54b4ba81a2615d8f5a9016061eed389ba5b6e Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Sergey=20M=E2=80=A4?= <dstftw@gmail.com>
Date: Tue, 20 Jan 2015 21:23:51 +0600
Subject: [PATCH 13/40] [nbcnews] Ignore HTTP errors while coping with
 playlists (Closes #4749)

---
 youtube_dl/extractor/nbc.py | 24 ++++++++++++++++++++++--
 1 file changed, 22 insertions(+), 2 deletions(-)

diff --git a/youtube_dl/extractor/nbc.py b/youtube_dl/extractor/nbc.py
index 690c46b6a..7ae66616a 100644
--- a/youtube_dl/extractor/nbc.py
+++ b/youtube_dl/extractor/nbc.py
@@ -6,6 +6,7 @@ import json
 from .common import InfoExtractor
 from ..compat import (
     compat_str,
+    compat_HTTPError,
 )
 from ..utils import (
     ExtractorError,
@@ -78,6 +79,16 @@ class NBCNewsIE(InfoExtractor):
             },
             'add_ie': ['ThePlatform'],
         },
+        {
+            'url': 'http://www.nbcnews.com/feature/dateline-full-episodes/full-episode-family-business-n285156',
+            'md5': 'fdbf39ab73a72df5896b6234ff98518a',
+            'info_dict': {
+                'id': 'Wjf9EDR3A_60',
+                'ext': 'mp4',
+                'title': 'FULL EPISODE: Family Business',
+                'description': 'md5:757988edbaae9d7be1d585eb5d55cc04',
+            },
+        },
     ]
 
     def _real_extract(self, url):
@@ -115,10 +126,19 @@ class NBCNewsIE(InfoExtractor):
                 if not base_url:
                     continue
                 playlist_url = base_url + '?form=MPXNBCNewsAPI'
-                all_videos = self._download_json(playlist_url, title)['videos']
 
                 try:
-                    info = next(v for v in all_videos if v['mpxId'] == mpxid)
+                    all_videos = self._download_json(playlist_url, title)
+                except ExtractorError as ee:
+                    if isinstance(ee.cause, compat_HTTPError):
+                        continue
+                    raise
+
+                if not all_videos or not 'videos' in all_videos:
+                    continue
+
+                try:
+                    info = next(v for v in all_videos['videos'] if v['mpxId'] == mpxid)
                     break
                 except StopIteration:
                     continue

From 564bb5e964fc263012784baa579a10eee822a436 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Sergey=20M=E2=80=A4?= <dstftw@gmail.com>
Date: Wed, 21 Jan 2015 02:15:28 +0600
Subject: [PATCH 14/40] [tinypic] Tweak VALID_URL regex (Closes #4754)

---
 youtube_dl/extractor/tinypic.py | 24 +++++++++++++++---------
 1 file changed, 15 insertions(+), 9 deletions(-)

diff --git a/youtube_dl/extractor/tinypic.py b/youtube_dl/extractor/tinypic.py
index 4fe89dbe5..e036b8cdf 100644
--- a/youtube_dl/extractor/tinypic.py
+++ b/youtube_dl/extractor/tinypic.py
@@ -9,17 +9,23 @@ from ..utils import ExtractorError
 class TinyPicIE(InfoExtractor):
     IE_NAME = 'tinypic'
     IE_DESC = 'tinypic.com videos'
-    _VALID_URL = r'http://tinypic\.com/player\.php\?v=(?P<id>[^&]+)&s=\d+'
+    _VALID_URL = r'http://(?:.+?\.)?tinypic\.com/player\.php\?v=(?P<id>[^&]+)&s=\d+'
 
-    _TEST = {
-        'url': 'http://tinypic.com/player.php?v=6xw7tc%3E&s=5#.UtqZmbRFCM8',
-        'md5': '609b74432465364e72727ebc6203f044',
-        'info_dict': {
-            'id': '6xw7tc',
-            'ext': 'flv',
-            'title': 'shadow phenomenon weird',
+    _TESTS = [
+        {
+            'url': 'http://tinypic.com/player.php?v=6xw7tc%3E&s=5#.UtqZmbRFCM8',
+            'md5': '609b74432465364e72727ebc6203f044',
+            'info_dict': {
+                'id': '6xw7tc',
+                'ext': 'flv',
+                'title': 'shadow phenomenon weird',
+            },
+        },
+        {
+            'url': 'http://de.tinypic.com/player.php?v=dy90yh&s=8',
+            'only_matching': True,
         }
-    }
+    ]
 
     def _real_extract(self, url):
         mobj = re.match(self._VALID_URL, url)

From 5fe511258998666e7617c22a5db27506c16eb955 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Jaime=20Marqui=CC=81nez=20Ferra=CC=81ndiz?=
 <jaime.marquinez.ferrandiz@gmail.com>
Date: Wed, 21 Jan 2015 10:27:18 +0100
Subject: [PATCH 15/40] [CNNArticle] Update test

---
 youtube_dl/extractor/cnn.py | 10 +++++-----
 1 file changed, 5 insertions(+), 5 deletions(-)

diff --git a/youtube_dl/extractor/cnn.py b/youtube_dl/extractor/cnn.py
index 93e8d0de3..4bb58f25e 100644
--- a/youtube_dl/extractor/cnn.py
+++ b/youtube_dl/extractor/cnn.py
@@ -143,13 +143,13 @@ class CNNArticleIE(InfoExtractor):
     _VALID_URL = r'https?://(?:(?:edition|www)\.)?cnn\.com/(?!video/)'
     _TEST = {
         'url': 'http://www.cnn.com/2014/12/21/politics/obama-north-koreas-hack-not-war-but-cyber-vandalism/',
-        'md5': '275b326f85d80dff7592a9820f5dc887',
+        'md5': '689034c2a3d9c6dc4aa72d65a81efd01',
         'info_dict': {
-            'id': 'bestoftv/2014/12/21/sotu-crowley-president-obama-north-korea-not-going-to-be-intimidated.cnn',
+            'id': 'bestoftv/2014/12/21/ip-north-korea-obama.cnn',
             'ext': 'mp4',
-            'title': 'Obama: We\'re not going to be intimidated',
-            'description': 'md5:e735586f3dc936075fa654a4d91b21f9',
-            'upload_date': '20141220',
+            'title': 'Obama: Cyberattack not an act of war',
+            'description': 'md5:51ce6750450603795cad0cdfbd7d05c5',
+            'upload_date': '20141221',
         },
         'add_ie': ['CNN'],
     }

From efcddaebe9302eaaea1cbb5570c5c2bd25abf1fe Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Jaime=20Marqui=CC=81nez=20Ferra=CC=81ndiz?=
 <jaime.marquinez.ferrandiz@gmail.com>
Date: Wed, 21 Jan 2015 10:31:57 +0100
Subject: [PATCH 16/40] [cnn] Use edition.cnn.com for getting the information
 (fixes #4757)

Some videos (like http://edition.cnn.com/videos/us/2015/01/20/orig-yellowstone-oil-spill.cnn) will fail if we use cnn.com.
---
 youtube_dl/extractor/cnn.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/youtube_dl/extractor/cnn.py b/youtube_dl/extractor/cnn.py
index 4bb58f25e..90ea07438 100644
--- a/youtube_dl/extractor/cnn.py
+++ b/youtube_dl/extractor/cnn.py
@@ -51,7 +51,7 @@ class CNNIE(InfoExtractor):
         mobj = re.match(self._VALID_URL, url)
         path = mobj.group('path')
         page_title = mobj.group('title')
-        info_url = 'http://cnn.com/video/data/3.0/%s/index.xml' % path
+        info_url = 'http://edition.cnn.com/video/data/3.0/%s/index.xml' % path
         info = self._download_xml(info_url, page_title)
 
         formats = []

From 47e0e1e0e2306dd7119844bbbc3ef39ccb175ed1 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Jaime=20Marqui=CC=81nez=20Ferra=CC=81ndiz?=
 <jaime.marquinez.ferrandiz@gmail.com>
Date: Wed, 21 Jan 2015 10:36:15 +0100
Subject: [PATCH 17/40] [nbc] Fix pep8 issue

---
 youtube_dl/extractor/nbc.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/youtube_dl/extractor/nbc.py b/youtube_dl/extractor/nbc.py
index 7ae66616a..f840f6532 100644
--- a/youtube_dl/extractor/nbc.py
+++ b/youtube_dl/extractor/nbc.py
@@ -134,7 +134,7 @@ class NBCNewsIE(InfoExtractor):
                         continue
                     raise
 
-                if not all_videos or not 'videos' in all_videos:
+                if not all_videos or 'videos' not in all_videos:
                     continue
 
                 try:

From c8dc41a6e7eea2a1235238cfeadb4caf5619d0ae Mon Sep 17 00:00:00 2001
From: Yen Chi Hsuan <yan12125@gmail.com>
Date: Wed, 21 Jan 2015 23:05:47 +0800
Subject: [PATCH 18/40] [StreetVoice] Add new extractor

---
 youtube_dl/extractor/__init__.py    |  1 +
 youtube_dl/extractor/streetvoice.py | 36 +++++++++++++++++++++++++++++
 2 files changed, 37 insertions(+)
 create mode 100644 youtube_dl/extractor/streetvoice.py

diff --git a/youtube_dl/extractor/__init__.py b/youtube_dl/extractor/__init__.py
index 2b9d4455d..9e942e6f9 100644
--- a/youtube_dl/extractor/__init__.py
+++ b/youtube_dl/extractor/__init__.py
@@ -409,6 +409,7 @@ from .stanfordoc import StanfordOpenClassroomIE
 from .steam import SteamIE
 from .streamcloud import StreamcloudIE
 from .streamcz import StreamCZIE
+from .streetvoice import StreetVoiceIE
 from .sunporno import SunPornoIE
 from .swrmediathek import SWRMediathekIE
 from .syfy import SyfyIE
diff --git a/youtube_dl/extractor/streetvoice.py b/youtube_dl/extractor/streetvoice.py
new file mode 100644
index 000000000..55ceea75b
--- /dev/null
+++ b/youtube_dl/extractor/streetvoice.py
@@ -0,0 +1,36 @@
+# coding: utf-8
+from __future__ import unicode_literals
+
+from .common import InfoExtractor
+
+
+class StreetVoiceIE(InfoExtractor):
+    _VALID_URL = r'http://tw.streetvoice.com/[^/]+/songs/(?P<id>[0-9]+)/'
+    _TESTS = [
+        {
+            'url': 'http://tw.streetvoice.com/skippylu/songs/94440/',
+            'md5': '15974627fc01a29e492c98593c2fd472',
+            'info_dict': {
+                'id': '94440',
+                'ext': 'mp3',
+                'title': '輸',
+                'description': '輸 - Crispy脆樂團'
+            }
+        }
+    ]
+
+    def _real_extract(self, url):
+        song_id = self._match_id(url)
+
+        api_url = 'http://tw.streetvoice.com/music/api/song/%s/' % song_id
+        info_dict = self._download_json(api_url, song_id)
+
+        author = info_dict['musician']['name']
+        title = info_dict['name']
+        return {
+            'id': song_id,
+            'ext': 'mp3',
+            'title': title,
+            'url': info_dict['file'],
+            'description': '%s - %s' % (title, author)
+        }

From c5db6bb32ba708bc5c92a68487b07be2374f416d Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Sergey=20M=E2=80=A4?= <dstftw@gmail.com>
Date: Wed, 21 Jan 2015 22:27:21 +0600
Subject: [PATCH 19/40] [twitch] Refactor and add support for past broadcasts

---
 youtube_dl/extractor/__init__.py |   8 +-
 youtube_dl/extractor/twitch.py   | 345 ++++++++++++++++---------------
 2 files changed, 184 insertions(+), 169 deletions(-)

diff --git a/youtube_dl/extractor/__init__.py b/youtube_dl/extractor/__init__.py
index 2b9d4455d..0b0f3be19 100644
--- a/youtube_dl/extractor/__init__.py
+++ b/youtube_dl/extractor/__init__.py
@@ -458,7 +458,13 @@ from .tvigle import TvigleIE
 from .tvp import TvpIE, TvpSeriesIE
 from .tvplay import TVPlayIE
 from .twentyfourvideo import TwentyFourVideoIE
-from .twitch import TwitchIE
+from .twitch import (
+    TwitchVideoIE,
+    TwitchChapterIE,
+    TwitchVodIE,
+    TwitchProfileIE,
+    TwitchPastBroadcastsIE,
+)
 from .ubu import UbuIE
 from .udemy import (
     UdemyIE,
diff --git a/youtube_dl/extractor/twitch.py b/youtube_dl/extractor/twitch.py
index b11a1d561..f70002f68 100644
--- a/youtube_dl/extractor/twitch.py
+++ b/youtube_dl/extractor/twitch.py
@@ -15,44 +15,11 @@ from ..utils import (
 )
 
 
-class TwitchIE(InfoExtractor):
-    # TODO: One broadcast may be split into multiple videos. The key
-    # 'broadcast_id' is the same for all parts, and 'broadcast_part'
-    # starts at 1 and increases. Can we treat all parts as one video?
-    _VALID_URL = r"""(?x)^(?:http://)?(?:www\.)?twitch\.tv/
-        (?:
-            (?P<channelid>[^/]+)|
-            (?:(?:[^/]+)/v/(?P<vodid>[^/]+))|
-            (?:(?:[^/]+)/b/(?P<videoid>[^/]+))|
-            (?:(?:[^/]+)/c/(?P<chapterid>[^/]+))
-        )
-        /?(?:\#.*)?$
-        """
-    _PAGE_LIMIT = 100
+class TwitchBaseIE(InfoExtractor):
+    _VALID_URL_BASE = r'http://(?:www\.)?twitch\.tv'
+
     _API_BASE = 'https://api.twitch.tv'
     _LOGIN_URL = 'https://secure.twitch.tv/user/login'
-    _TESTS = [{
-        'url': 'http://www.twitch.tv/riotgames/b/577357806',
-        'info_dict': {
-            'id': 'a577357806',
-            'title': 'Worlds Semifinals - Star Horn Royal Club vs. OMG',
-        },
-        'playlist_mincount': 12,
-    }, {
-        'url': 'http://www.twitch.tv/acracingleague/c/5285812',
-        'info_dict': {
-            'id': 'c5285812',
-            'title': 'ACRL Off Season - Sports Cars @ Nordschleife',
-        },
-        'playlist_mincount': 3,
-    }, {
-        'url': 'http://www.twitch.tv/vanillatv',
-        'info_dict': {
-            'id': 'vanillatv',
-            'title': 'VanillaTV',
-        },
-        'playlist_mincount': 412,
-    }]
 
     def _handle_error(self, response):
         if not isinstance(response, dict):
@@ -64,71 +31,10 @@ class TwitchIE(InfoExtractor):
                 expected=True)
 
     def _download_json(self, url, video_id, note='Downloading JSON metadata'):
-        response = super(TwitchIE, self)._download_json(url, video_id, note)
+        response = super(TwitchBaseIE, self)._download_json(url, video_id, note)
         self._handle_error(response)
         return response
 
-    def _extract_media(self, item, item_id):
-        ITEMS = {
-            'a': 'video',
-            'v': 'vod',
-            'c': 'chapter',
-        }
-        info = self._extract_info(self._download_json(
-            '%s/kraken/videos/%s%s' % (self._API_BASE, item, item_id), item_id,
-            'Downloading %s info JSON' % ITEMS[item]))
-
-        if item == 'v':
-            access_token = self._download_json(
-                '%s/api/vods/%s/access_token' % (self._API_BASE, item_id), item_id,
-                'Downloading %s access token' % ITEMS[item])
-            formats = self._extract_m3u8_formats(
-                'http://usher.twitch.tv/vod/%s?nauth=%s&nauthsig=%s'
-                % (item_id, access_token['token'], access_token['sig']),
-                item_id, 'mp4')
-            info['formats'] = formats
-            return info
-
-        response = self._download_json(
-            '%s/api/videos/%s%s' % (self._API_BASE, item, item_id), item_id,
-            'Downloading %s playlist JSON' % ITEMS[item])
-        entries = []
-        chunks = response['chunks']
-        qualities = list(chunks.keys())
-        for num, fragment in enumerate(zip(*chunks.values()), start=1):
-            formats = []
-            for fmt_num, fragment_fmt in enumerate(fragment):
-                format_id = qualities[fmt_num]
-                fmt = {
-                    'url': fragment_fmt['url'],
-                    'format_id': format_id,
-                    'quality': 1 if format_id == 'live' else 0,
-                }
-                m = re.search(r'^(?P<height>\d+)[Pp]', format_id)
-                if m:
-                    fmt['height'] = int(m.group('height'))
-                formats.append(fmt)
-            self._sort_formats(formats)
-            entry = dict(info)
-            entry['id'] = '%s_%d' % (entry['id'], num)
-            entry['title'] = '%s part %d' % (entry['title'], num)
-            entry['formats'] = formats
-            entries.append(entry)
-        return self.playlist_result(entries, info['id'], info['title'])
-
-    def _extract_info(self, info):
-        return {
-            'id': info['_id'],
-            'title': info['title'],
-            'description': info['description'],
-            'duration': info['length'],
-            'thumbnail': info['preview'],
-            'uploader': info['channel']['display_name'],
-            'uploader_id': info['channel']['name'],
-            'timestamp': parse_iso8601(info['recorded_at']),
-            'view_count': info['views'],
-        }
-
     def _real_initialize(self):
         self._login()
 
@@ -167,81 +73,184 @@ class TwitchIE(InfoExtractor):
             raise ExtractorError(
                 'Unable to login: %s' % m.group('msg').strip(), expected=True)
 
+
+class TwitchItemBaseIE(TwitchBaseIE):
+    def _download_info(self, item, item_id):
+        return self._extract_info(self._download_json(
+            '%s/kraken/videos/%s%s' % (self._API_BASE, item, item_id), item_id,
+            'Downloading %s info JSON' % self._ITEM_TYPE))
+
+    def _extract_media(self, item_id):
+        info = self._download_info(self._ITEM_SHORTCUT, item_id)
+        response = self._download_json(
+            '%s/api/videos/%s%s' % (self._API_BASE, self._ITEM_SHORTCUT, item_id), item_id,
+            'Downloading %s playlist JSON' % self._ITEM_TYPE)
+        entries = []
+        chunks = response['chunks']
+        qualities = list(chunks.keys())
+        for num, fragment in enumerate(zip(*chunks.values()), start=1):
+            formats = []
+            for fmt_num, fragment_fmt in enumerate(fragment):
+                format_id = qualities[fmt_num]
+                fmt = {
+                    'url': fragment_fmt['url'],
+                    'format_id': format_id,
+                    'quality': 1 if format_id == 'live' else 0,
+                }
+                m = re.search(r'^(?P<height>\d+)[Pp]', format_id)
+                if m:
+                    fmt['height'] = int(m.group('height'))
+                formats.append(fmt)
+            self._sort_formats(formats)
+            entry = dict(info)
+            entry['id'] = '%s_%d' % (entry['id'], num)
+            entry['title'] = '%s part %d' % (entry['title'], num)
+            entry['formats'] = formats
+            entries.append(entry)
+        return self.playlist_result(entries, info['id'], info['title'])
+
+    def _extract_info(self, info):
+        return {
+            'id': info['_id'],
+            'title': info['title'],
+            'description': info['description'],
+            'duration': info['length'],
+            'thumbnail': info['preview'],
+            'uploader': info['channel']['display_name'],
+            'uploader_id': info['channel']['name'],
+            'timestamp': parse_iso8601(info['recorded_at']),
+            'view_count': info['views'],
+        }
+
     def _real_extract(self, url):
-        mobj = re.match(self._VALID_URL, url)
-        if mobj.group('chapterid'):
-            return self._extract_media('c', mobj.group('chapterid'))
+        return self._extract_media(self._match_id(url))
 
-            """
-            webpage = self._download_webpage(url, chapter_id)
-            m = re.search(r'PP\.archive_id = "([0-9]+)";', webpage)
-            if not m:
-                raise ExtractorError('Cannot find archive of a chapter')
-            archive_id = m.group(1)
 
-            api = api_base + '/broadcast/by_chapter/%s.xml' % chapter_id
-            doc = self._download_xml(
-                api, chapter_id,
-                note='Downloading chapter information',
-                errnote='Chapter information download failed')
-            for a in doc.findall('.//archive'):
-                if archive_id == a.find('./id').text:
-                    break
-            else:
-                raise ExtractorError('Could not find chapter in chapter information')
+class TwitchVideoIE(TwitchItemBaseIE):
+    IE_NAME = 'twitch:video'
+    _VALID_URL = r'%s/[^/]+/b/(?P<id>[^/]+)' % TwitchBaseIE._VALID_URL_BASE
+    _ITEM_TYPE = 'video'
+    _ITEM_SHORTCUT = 'a'
 
-            video_url = a.find('./video_file_url').text
-            video_ext = video_url.rpartition('.')[2] or 'flv'
+    _TEST = {
+        'url': 'http://www.twitch.tv/riotgames/b/577357806',
+        'info_dict': {
+            'id': 'a577357806',
+            'title': 'Worlds Semifinals - Star Horn Royal Club vs. OMG',
+        },
+        'playlist_mincount': 12,
+    }
 
-            chapter_api_url = 'https://api.twitch.tv/kraken/videos/c' + chapter_id
-            chapter_info = self._download_json(
-                chapter_api_url, 'c' + chapter_id,
-                note='Downloading chapter metadata',
-                errnote='Download of chapter metadata failed')
 
-            bracket_start = int(doc.find('.//bracket_start').text)
-            bracket_end = int(doc.find('.//bracket_end').text)
+class TwitchChapterIE(TwitchItemBaseIE):
+    IE_NAME = 'twitch:chapter'
+    _VALID_URL = r'%s/[^/]+/c/(?P<id>[^/]+)' % TwitchBaseIE._VALID_URL_BASE
+    _ITEM_TYPE = 'chapter'
+    _ITEM_SHORTCUT = 'c'
 
-            # TODO determine start (and probably fix up file)
-            #  youtube-dl -v http://www.twitch.tv/firmbelief/c/1757457
-            #video_url += '?start=' + TODO:start_timestamp
-            # bracket_start is 13290, but we want 51670615
-            self._downloader.report_warning('Chapter detected, but we can just download the whole file. '
-                                            'Chapter starts at %s and ends at %s' % (formatSeconds(bracket_start), formatSeconds(bracket_end)))
+    _TEST = {
+        'url': 'http://www.twitch.tv/acracingleague/c/5285812',
+        'info_dict': {
+            'id': 'c5285812',
+            'title': 'ACRL Off Season - Sports Cars @ Nordschleife',
+        },
+        'playlist_mincount': 3,
+    }
 
-            info = {
-                'id': 'c' + chapter_id,
-                'url': video_url,
-                'ext': video_ext,
-                'title': chapter_info['title'],
-                'thumbnail': chapter_info['preview'],
-                'description': chapter_info['description'],
-                'uploader': chapter_info['channel']['display_name'],
-                'uploader_id': chapter_info['channel']['name'],
-            }
-            return info
-            """
-        elif mobj.group('videoid'):
-            return self._extract_media('a', mobj.group('videoid'))
-        elif mobj.group('vodid'):
-            return self._extract_media('v', mobj.group('vodid'))
-        elif mobj.group('channelid'):
-            channel_id = mobj.group('channelid')
-            info = self._download_json(
-                '%s/kraken/channels/%s' % (self._API_BASE, channel_id),
-                channel_id, 'Downloading channel info JSON')
-            channel_name = info.get('display_name') or info.get('name')
-            entries = []
-            offset = 0
-            limit = self._PAGE_LIMIT
-            for counter in itertools.count(1):
-                response = self._download_json(
-                    '%s/kraken/channels/%s/videos/?offset=%d&limit=%d'
-                    % (self._API_BASE, channel_id, offset, limit),
-                    channel_id, 'Downloading channel videos JSON page %d' % counter)
-                videos = response['videos']
-                if not videos:
-                    break
-                entries.extend([self.url_result(video['url'], 'Twitch') for video in videos])
-                offset += limit
-            return self.playlist_result(entries, channel_id, channel_name)
+
+class TwitchVodIE(TwitchItemBaseIE):
+    IE_NAME = 'twitch:vod'
+    _VALID_URL = r'%s/[^/]+/v/(?P<id>[^/]+)' % TwitchBaseIE._VALID_URL_BASE
+    _ITEM_TYPE = 'vod'
+    _ITEM_SHORTCUT = 'v'
+
+    _TEST = {
+        'url': 'http://www.twitch.tv/ksptv/v/3622000',
+        'info_dict': {
+            'id': 'v3622000',
+            'ext': 'mp4',
+            'title': '''KSPTV: Squadcast: "Everyone's on vacation so here's Dahud" Edition!''',
+            'thumbnail': 're:^https?://.*\.jpg$',
+            'duration': 6951,
+            'timestamp': 1419028564,
+            'upload_date': '20141219',
+            'uploader': 'KSPTV',
+            'uploader_id': 'ksptv',
+            'view_count': int,
+        },
+        'params': {
+            # m3u8 download
+            'skip_download': True,
+        },
+    }
+
+    def _real_extract(self, url):
+        item_id = self._match_id(url)
+        info = self._download_info(self._ITEM_SHORTCUT, item_id)
+        access_token = self._download_json(
+            '%s/api/vods/%s/access_token' % (self._API_BASE, item_id), item_id,
+            'Downloading %s access token' % self._ITEM_TYPE)
+        formats = self._extract_m3u8_formats(
+            'http://usher.twitch.tv/vod/%s?nauth=%s&nauthsig=%s'
+            % (item_id, access_token['token'], access_token['sig']),
+            item_id, 'mp4')
+        info['formats'] = formats
+        return info
+
+
+class TwitchPlaylistBaseIE(TwitchBaseIE):
+    _PLAYLIST_URL = '%s/kraken/channels/%%s/videos/?offset=%%d&limit=%%d' % TwitchBaseIE._API_BASE
+    _PAGE_LIMIT = 100
+
+    def _extract_playlist(self, channel_id):
+        info = self._download_json(
+            '%s/kraken/channels/%s' % (self._API_BASE, channel_id),
+            channel_id, 'Downloading channel info JSON')
+        channel_name = info.get('display_name') or info.get('name')
+        entries = []
+        offset = 0
+        limit = self._PAGE_LIMIT
+        for counter in itertools.count(1):
+            response = self._download_json(
+                self._PLAYLIST_URL % (channel_id, offset, limit),
+                channel_id, 'Downloading %s videos JSON page %d' % (self._PLAYLIST_TYPE, counter))
+            videos = response['videos']
+            if not videos:
+                break
+            entries.extend([self.url_result(video['url']) for video in videos])
+            offset += limit
+        return self.playlist_result(entries, channel_id, channel_name)
+
+    def _real_extract(self, url):
+        return self._extract_playlist(self._match_id(url))
+
+
+class TwitchProfileIE(TwitchPlaylistBaseIE):
+    IE_NAME = 'twitch:profile'
+    _VALID_URL = r'%s/(?P<id>[^/]+)/profile/?(?:\#.*)?$' % TwitchBaseIE._VALID_URL_BASE
+    _PLAYLIST_TYPE = 'profile'
+
+    _TEST = {
+        'url': 'http://www.twitch.tv/vanillatv/profile',
+        'info_dict': {
+            'id': 'vanillatv',
+            'title': 'VanillaTV',
+        },
+        'playlist_mincount': 412,
+    }
+
+
+class TwitchPastBroadcastsIE(TwitchPlaylistBaseIE):
+    IE_NAME = 'twitch:profile'
+    _VALID_URL = r'%s/(?P<id>[^/]+)/profile/past_broadcasts/?(?:\#.*)?$' % TwitchBaseIE._VALID_URL_BASE
+    _PLAYLIST_URL = TwitchPlaylistBaseIE._PLAYLIST_URL + '&broadcasts=true'
+    _PLAYLIST_TYPE = 'past broadcasts'
+
+    _TEST = {
+        'url': 'http://www.twitch.tv/spamfish/profile/past_broadcasts',
+        'info_dict': {
+            'id': 'spamfish',
+            'title': 'Spamfish',
+        },
+        'playlist_mincount': 54,
+    }

From bb0aa4cb3ca1b21084e6e6229e4a69014e61ccff Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Sergey=20M=E2=80=A4?= <dstftw@gmail.com>
Date: Wed, 21 Jan 2015 22:53:51 +0600
Subject: [PATCH 20/40] [streetvoice] Improve

---
 youtube_dl/extractor/streetvoice.py | 53 ++++++++++++++++++-----------
 1 file changed, 34 insertions(+), 19 deletions(-)

diff --git a/youtube_dl/extractor/streetvoice.py b/youtube_dl/extractor/streetvoice.py
index 55ceea75b..6a57fa60a 100644
--- a/youtube_dl/extractor/streetvoice.py
+++ b/youtube_dl/extractor/streetvoice.py
@@ -2,35 +2,50 @@
 from __future__ import unicode_literals
 
 from .common import InfoExtractor
+from ..compat import compat_str
+from ..utils import unified_strdate
 
 
 class StreetVoiceIE(InfoExtractor):
-    _VALID_URL = r'http://tw.streetvoice.com/[^/]+/songs/(?P<id>[0-9]+)/'
-    _TESTS = [
-        {
-            'url': 'http://tw.streetvoice.com/skippylu/songs/94440/',
-            'md5': '15974627fc01a29e492c98593c2fd472',
-            'info_dict': {
-                'id': '94440',
-                'ext': 'mp3',
-                'title': '輸',
-                'description': '輸 - Crispy脆樂團'
-            }
+    _VALID_URL = r'https?://(?:.+?\.)?streetvoice\.com/[^/]+/songs/(?P<id>[0-9]+)'
+    _TESTS = [{
+        'url': 'http://streetvoice.com/skippylu/songs/94440/',
+        'md5': '15974627fc01a29e492c98593c2fd472',
+        'info_dict': {
+            'id': '94440',
+            'ext': 'mp3',
+            'filesize': 4167053,
+            'title': '輸',
+            'description': 'Crispy脆樂團 - 輸',
+            'thumbnail': 're:^https?://.*\.jpg$',
+            'duration': 260,
+            'upload_date': '20091018',
+            'uploader': 'Crispy脆樂團',
+            'uploader_id': '627810',
         }
-    ]
+    }, {
+        'url': 'http://tw.streetvoice.com/skippylu/songs/94440/',
+        'only_matching': True,
+    }]
 
     def _real_extract(self, url):
         song_id = self._match_id(url)
 
-        api_url = 'http://tw.streetvoice.com/music/api/song/%s/' % song_id
-        info_dict = self._download_json(api_url, song_id)
+        song = self._download_json(
+            'http://streetvoice.com/music/api/song/%s' % song_id, song_id)
+
+        title = song['name']
+        author = song['musician']['name']
 
-        author = info_dict['musician']['name']
-        title = info_dict['name']
         return {
             'id': song_id,
-            'ext': 'mp3',
+            'url': song['file'],
+            'filesize': song.get('size'),
             'title': title,
-            'url': info_dict['file'],
-            'description': '%s - %s' % (title, author)
+            'description': '%s - %s' % (author, title),
+            'thumbnail': self._proto_relative_url(song.get('image'), 'http:'),
+            'duration': song.get('length'),
+            'upload_date': unified_strdate(song.get('created_at')),
+            'uploader': author,
+            'uploader_id': compat_str(song['musician']['id']),
         }

From c62b44976545401af60fd35f0188fd09a93e449b Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Sergey=20M=E2=80=A4?= <dstftw@gmail.com>
Date: Wed, 21 Jan 2015 22:56:28 +0600
Subject: [PATCH 21/40] Credit @yan12125 for streetvoice (#4758)

---
 AUTHORS | 1 +
 1 file changed, 1 insertion(+)

diff --git a/AUTHORS b/AUTHORS
index 35e256d49..b8bf3cb6f 100644
--- a/AUTHORS
+++ b/AUTHORS
@@ -103,3 +103,4 @@ Christopher Krooss
 Ondřej Caletka
 Dinesh S
 Johan K. Jensen
+Yen Chi Hsuan

From 8bb1bdfae91246e1306ee31674c06a2ef7e42a08 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Sergey=20M=E2=80=A4?= <dstftw@gmail.com>
Date: Wed, 21 Jan 2015 23:06:16 +0600
Subject: [PATCH 22/40] [twitch:past_broadcasts] Fix IE_NAME

---
 youtube_dl/extractor/twitch.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/youtube_dl/extractor/twitch.py b/youtube_dl/extractor/twitch.py
index f70002f68..8866e5182 100644
--- a/youtube_dl/extractor/twitch.py
+++ b/youtube_dl/extractor/twitch.py
@@ -241,7 +241,7 @@ class TwitchProfileIE(TwitchPlaylistBaseIE):
 
 
 class TwitchPastBroadcastsIE(TwitchPlaylistBaseIE):
-    IE_NAME = 'twitch:profile'
+    IE_NAME = 'twitch:past_broadcasts'
     _VALID_URL = r'%s/(?P<id>[^/]+)/profile/past_broadcasts/?(?:\#.*)?$' % TwitchBaseIE._VALID_URL_BASE
     _PLAYLIST_URL = TwitchPlaylistBaseIE._PLAYLIST_URL + '&broadcasts=true'
     _PLAYLIST_TYPE = 'past broadcasts'

From e5763a7a7e630cfd94b8993a9592cb3243890e0c Mon Sep 17 00:00:00 2001
From: Naglis Jonaitis <njonaitis@gmail.com>
Date: Wed, 21 Jan 2015 21:47:55 +0200
Subject: [PATCH 23/40] [hearthisat] Add new extractor (Closes #4743)

---
 youtube_dl/extractor/__init__.py   |  1 +
 youtube_dl/extractor/hearthisat.py | 92 ++++++++++++++++++++++++++++++
 2 files changed, 93 insertions(+)
 create mode 100644 youtube_dl/extractor/hearthisat.py

diff --git a/youtube_dl/extractor/__init__.py b/youtube_dl/extractor/__init__.py
index c41f2de0f..be3fecd0d 100644
--- a/youtube_dl/extractor/__init__.py
+++ b/youtube_dl/extractor/__init__.py
@@ -176,6 +176,7 @@ from .goshgay import GoshgayIE
 from .grooveshark import GroovesharkIE
 from .groupon import GrouponIE
 from .hark import HarkIE
+from .hearthisat import HearThisAtIE
 from .heise import HeiseIE
 from .hellporno import HellPornoIE
 from .helsinki import HelsinkiIE
diff --git a/youtube_dl/extractor/hearthisat.py b/youtube_dl/extractor/hearthisat.py
new file mode 100644
index 000000000..c1177a9c5
--- /dev/null
+++ b/youtube_dl/extractor/hearthisat.py
@@ -0,0 +1,92 @@
+# coding: utf-8
+from __future__ import unicode_literals
+
+import re
+
+from .common import InfoExtractor
+from ..compat import compat_urllib_request
+from ..utils import (
+    str_to_int,
+    urlencode_postdata,
+)
+
+
+class HearThisAtIE(InfoExtractor):
+    _VALID_URL = r'https?://(?:www\.)?hearthis\.at/(?P<artist>[^/]+)/(?P<title>[A-Za-z0-9\-]+)/?$'
+    _PLAYLIST_URL = 'https://hearthis.at/playlist.php'
+    _TEST = {
+        'url': 'https://hearthis.at/moofi/dr-kreep',
+        'md5': 'd594c573227a89f4256f0b03e68c80cc',
+        'info_dict': {
+            'id': '150939',
+            'ext': 'mp3',
+            'title': 'Moofi - Dr. Kreep',
+            'thumbnail': 're:^https?://.*\.jpg$',
+            'timestamp': 1421564134,
+            'description': 'Creepy Patch. Mutable Instruments Braids Vowel + Formant Mode.',
+            'upload_date': '20150118',
+            'comment_count': int,
+            'view_count': int,
+            'like_count': int,
+            'duration': 71,
+            'categories': ['Experimental'],
+        }
+    }
+
+    def _real_extract(self, url):
+        m = re.match(self._VALID_URL, url)
+        display_id = '{artist:s} - {title:s}'.format(**m.groupdict())
+
+        webpage = self._download_webpage(url, display_id)
+        track_id = self._search_regex(
+            r'intTrackId\s*=\s*(\d+)', webpage, 'track ID')
+
+        payload = urlencode_postdata({'tracks[]': track_id})
+        req = compat_urllib_request.Request(self._PLAYLIST_URL, payload)
+        req.add_header('Content-type', 'application/x-www-form-urlencoded')
+
+        track = self._download_json(req, track_id, 'Downloading playlist')[0]
+        title = '{artist:s} - {title:s}'.format(**track)
+
+        categories = None
+        if track.get('category'):
+            categories = [track['category']]
+
+        description = self._og_search_description(webpage)
+        thumbnail = self._og_search_thumbnail(webpage)
+
+        meta_span = r'<span[^>]+class="%s".*?</i>([^<]+)</span>'
+        view_count = str_to_int(self._search_regex(
+            meta_span % 'plays_count', webpage, 'view count', fatal=False))
+        like_count = str_to_int(self._search_regex(
+            meta_span % 'likes_count', webpage, 'like count', fatal=False))
+        comment_count = str_to_int(self._search_regex(
+            meta_span % 'comment_count', webpage, 'comment count', fatal=False))
+        duration = str_to_int(self._search_regex(
+            r'data-length="(\d+)', webpage, 'duration', fatal=False))
+        timestamp = str_to_int(self._search_regex(
+            r'<span[^>]+class="calctime"[^>]+data-time="(\d+)', webpage, 'timestamp', fatal=False))
+
+        track_url = self._search_regex(
+            r'<a[^>]+data-mp3="([^"]+)"', webpage, 'track URL')
+
+        formats = [{
+            'format_id': 'mp3',
+            'url': track_url,
+            'vcodec': 'none',
+        }]
+
+        return {
+            'id': track_id,
+            'display-id': display_id,
+            'title': title,
+            'formats': formats,
+            'thumbnail': thumbnail,
+            'description': description,
+            'duration': duration,
+            'timestamp': timestamp,
+            'view_count': view_count,
+            'comment_count': comment_count,
+            'like_count': like_count,
+            'categories': categories,
+        }

From b55ee18ff3a9642fe25a977e1152472877294493 Mon Sep 17 00:00:00 2001
From: Philipp Hagemeister <phihag@phihag.de>
Date: Thu, 22 Jan 2015 12:04:07 +0100
Subject: [PATCH 24/40] [hearthisat] Add support for more high-quality download
 links

---
 youtube_dl/extractor/hearthisat.py | 49 ++++++++++++++++++++++--------
 youtube_dl/utils.py                |  8 +++++
 2 files changed, 45 insertions(+), 12 deletions(-)

diff --git a/youtube_dl/extractor/hearthisat.py b/youtube_dl/extractor/hearthisat.py
index c1177a9c5..d83072c1a 100644
--- a/youtube_dl/extractor/hearthisat.py
+++ b/youtube_dl/extractor/hearthisat.py
@@ -4,10 +4,15 @@ from __future__ import unicode_literals
 import re
 
 from .common import InfoExtractor
-from ..compat import compat_urllib_request
+from ..compat import (
+    compat_urllib_request,
+    compat_urlparse,
+)
 from ..utils import (
+    HEADRequest,
     str_to_int,
     urlencode_postdata,
+    urlhandle_detect_ext,
 )
 
 
@@ -16,10 +21,10 @@ class HearThisAtIE(InfoExtractor):
     _PLAYLIST_URL = 'https://hearthis.at/playlist.php'
     _TEST = {
         'url': 'https://hearthis.at/moofi/dr-kreep',
-        'md5': 'd594c573227a89f4256f0b03e68c80cc',
+        'md5': 'ab6ec33c8fed6556029337c7885eb4e0',
         'info_dict': {
             'id': '150939',
-            'ext': 'mp3',
+            'ext': 'wav',
             'title': 'Moofi - Dr. Kreep',
             'thumbnail': 're:^https?://.*\.jpg$',
             'timestamp': 1421564134,
@@ -67,18 +72,38 @@ class HearThisAtIE(InfoExtractor):
         timestamp = str_to_int(self._search_regex(
             r'<span[^>]+class="calctime"[^>]+data-time="(\d+)', webpage, 'timestamp', fatal=False))
 
-        track_url = self._search_regex(
-            r'<a[^>]+data-mp3="([^"]+)"', webpage, 'track URL')
-
-        formats = [{
-            'format_id': 'mp3',
-            'url': track_url,
-            'vcodec': 'none',
-        }]
+        formats = []
+        mp3_url = self._search_regex(
+            r'(?s)<a class="player-link"\s+(?:[a-zA-Z0-9_:-]+="[^"]+"\s+)*?data-mp3="([^"]+)"',
+            webpage, 'title', fatal=False)
+        if mp3_url:
+            formats.append({
+                'format_id': 'mp3',
+                'vcodec': 'none',
+                'acodec': 'mp3',
+                'url': mp3_url,
+            })
+        download_path = self._search_regex(
+            r'<a class="[^"]*download_fct[^"]*"\s+href="([^"]+)"',
+            webpage, 'download URL', default=None)
+        if download_path:
+            download_url = compat_urlparse.urljoin(url, download_path)
+            ext_req = HEADRequest(download_url)
+            ext_handle = self._request_webpage(
+                ext_req, display_id, note='Determining extension')
+            ext = urlhandle_detect_ext(ext_handle)
+            formats.append({
+                'format_id': 'download',
+                'vcodec': 'none',
+                'ext': ext,
+                'url': download_url,
+                'preference': 2,  # Usually better quality
+            })
+        self._sort_formats(formats)
 
         return {
             'id': track_id,
-            'display-id': display_id,
+            'display_id': display_id,
             'title': title,
             'formats': formats,
             'thumbnail': thumbnail,
diff --git a/youtube_dl/utils.py b/youtube_dl/utils.py
index 7832ed87f..764474c33 100644
--- a/youtube_dl/utils.py
+++ b/youtube_dl/utils.py
@@ -1612,6 +1612,14 @@ def urlhandle_detect_ext(url_handle):
     except AttributeError:  # Python < 3
         getheader = url_handle.info().getheader
 
+    cd = getheader('Content-Disposition')
+    if cd:
+        m = re.match(r'attachment;\s*filename="(?P<filename>[^"]+)"', cd)
+        if m:
+            e = determine_ext(m.group('filename'), default_ext=None)
+            if e:
+                return e
+
     return getheader('Content-Type').split("/")[1]
 
 

From 898c23c03f634d14fa414f0e0693d5c4033b2671 Mon Sep 17 00:00:00 2001
From: Philipp Hagemeister <phihag@phihag.de>
Date: Thu, 22 Jan 2015 12:04:26 +0100
Subject: [PATCH 25/40] release 2015.01.22

---
 README.md             | 3 ++-
 youtube_dl/version.py | 2 +-
 2 files changed, 3 insertions(+), 2 deletions(-)

diff --git a/README.md b/README.md
index 078e9df82..60bd7e2b6 100644
--- a/README.md
+++ b/README.md
@@ -304,7 +304,8 @@ which means you can modify it, redistribute it or use it however you like.
 
 ## Authentication Options:
     -u, --username USERNAME          login with this account ID
-    -p, --password PASSWORD          account password
+    -p, --password PASSWORD          account password. If this option is left
+                                     out, youtube-dl will ask interactively.
     -2, --twofactor TWOFACTOR        two-factor auth code
     -n, --netrc                      use .netrc authentication data
     --video-password PASSWORD        video password (vimeo, smotri)
diff --git a/youtube_dl/version.py b/youtube_dl/version.py
index 63a79a7ee..f5df7be31 100644
--- a/youtube_dl/version.py
+++ b/youtube_dl/version.py
@@ -1,3 +1,3 @@
 from __future__ import unicode_literals
 
-__version__ = '2015.01.16'
+__version__ = '2015.01.22'

From bb6e38787dcc3593a4d9cb66bcf9f736570e944a Mon Sep 17 00:00:00 2001
From: Naglis Jonaitis <njonaitis@gmail.com>
Date: Thu, 22 Jan 2015 18:35:53 +0200
Subject: [PATCH 26/40] [videomega] Fix extraction (Closes #4763)

---
 youtube_dl/extractor/videomega.py | 10 ++++++++--
 1 file changed, 8 insertions(+), 2 deletions(-)

diff --git a/youtube_dl/extractor/videomega.py b/youtube_dl/extractor/videomega.py
index 9fc64d172..fc6e05fe0 100644
--- a/youtube_dl/extractor/videomega.py
+++ b/youtube_dl/extractor/videomega.py
@@ -1,12 +1,15 @@
 # coding: utf-8
 from __future__ import unicode_literals
 
+import re
+
 from .common import InfoExtractor
 from ..compat import (
     compat_urllib_parse,
     compat_urllib_request,
 )
 from ..utils import (
+    ExtractorError,
     remove_start,
 )
 
@@ -35,8 +38,11 @@ class VideoMegaIE(InfoExtractor):
         req.add_header('Referer', url)
         webpage = self._download_webpage(req, video_id)
 
-        escaped_data = self._search_regex(
-            r'unescape\("([^"]+)"\)', webpage, 'escaped data')
+        try:
+            escaped_data = re.findall(r'unescape\("([^"]+)"\)', webpage)[-1]
+        except IndexError:
+            raise ExtractorError('Unable to extract escaped data')
+
         playlist = compat_urllib_parse.unquote(escaped_data)
 
         thumbnail = self._search_regex(

From 240b9b7a5cd5f19f50ae28fb3ef991fede216ef6 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Sergey=20M=E2=80=A4?= <dstftw@gmail.com>
Date: Thu, 22 Jan 2015 23:11:22 +0600
Subject: [PATCH 27/40] [twitch] Add support for streams (Closes #893, closes
 #3693, closes #1884)

---
 youtube_dl/extractor/__init__.py |  1 +
 youtube_dl/extractor/twitch.py   | 97 +++++++++++++++++++++++++++++++-
 2 files changed, 95 insertions(+), 3 deletions(-)

diff --git a/youtube_dl/extractor/__init__.py b/youtube_dl/extractor/__init__.py
index be3fecd0d..a983fff76 100644
--- a/youtube_dl/extractor/__init__.py
+++ b/youtube_dl/extractor/__init__.py
@@ -466,6 +466,7 @@ from .twitch import (
     TwitchVodIE,
     TwitchProfileIE,
     TwitchPastBroadcastsIE,
+    TwitchStreamIE,
 )
 from .ubu import UbuIE
 from .udemy import (
diff --git a/youtube_dl/extractor/twitch.py b/youtube_dl/extractor/twitch.py
index 8866e5182..2891c4680 100644
--- a/youtube_dl/extractor/twitch.py
+++ b/youtube_dl/extractor/twitch.py
@@ -6,6 +6,7 @@ import re
 
 from .common import InfoExtractor
 from ..compat import (
+    compat_str,
     compat_urllib_parse,
     compat_urllib_request,
 )
@@ -16,9 +17,10 @@ from ..utils import (
 
 
 class TwitchBaseIE(InfoExtractor):
-    _VALID_URL_BASE = r'http://(?:www\.)?twitch\.tv'
+    _VALID_URL_BASE = r'https?://(?:www\.)?twitch\.tv'
 
     _API_BASE = 'https://api.twitch.tv'
+    _USHER_BASE = 'http://usher.twitch.tv'
     _LOGIN_URL = 'https://secure.twitch.tv/user/login'
 
     def _handle_error(self, response):
@@ -191,8 +193,8 @@ class TwitchVodIE(TwitchItemBaseIE):
             '%s/api/vods/%s/access_token' % (self._API_BASE, item_id), item_id,
             'Downloading %s access token' % self._ITEM_TYPE)
         formats = self._extract_m3u8_formats(
-            'http://usher.twitch.tv/vod/%s?nauth=%s&nauthsig=%s'
-            % (item_id, access_token['token'], access_token['sig']),
+            '%s/vod/%s?nauth=%s&nauthsig=%s'
+            % (self._USHER_BASE, item_id, access_token['token'], access_token['sig']),
             item_id, 'mp4')
         info['formats'] = formats
         return info
@@ -254,3 +256,92 @@ class TwitchPastBroadcastsIE(TwitchPlaylistBaseIE):
         },
         'playlist_mincount': 54,
     }
+
+
+class TwitchStreamIE(TwitchBaseIE):
+    IE_NAME = 'twitch:stream'
+    _VALID_URL = r'%s/(?P<id>[^/]+)/?(?:\#.*)?$' % TwitchBaseIE._VALID_URL_BASE
+
+    _TEST = {
+        'url': 'http://www.twitch.tv/shroomztv',
+        'info_dict': {
+            'id': '12772022048',
+            'display_id': 'shroomztv',
+            'ext': 'mp4',
+            'title': 're:^ShroomzTV [0-9]{4}-[0-9]{2}-[0-9]{2} [0-9]{2}:[0-9]{2}$',
+            'description': 'H1Z1 - lonewolfing with ShroomzTV | A3 Battle Royale later - @ShroomzTV',
+            'is_live': True,
+            'timestamp': 1421928037,
+            'upload_date': '20150122',
+            'uploader': 'ShroomzTV',
+            'uploader_id': 'shroomztv',
+            'view_count': int,
+        },
+        'params': {
+            # m3u8 download
+            'skip_download': True,
+        },
+    }
+
+    def _real_extract(self, url):
+        channel_id = self._match_id(url)
+
+        stream = self._download_json(
+            '%s/kraken/streams/%s' % (self._API_BASE, channel_id), channel_id,
+            'Downloading stream JSON').get('stream')
+
+        # Fallback on profile extraction if stream is offline
+        if not stream:
+            return self.url_result(
+                'http://www.twitch.tv/%s/profile' % channel_id,
+                'TwitchProfile', channel_id)
+
+        access_token = self._download_json(
+            '%s/api/channels/%s/access_token' % (self._API_BASE, channel_id), channel_id,
+            'Downloading channel access token')
+
+        query = {
+            'allow_source': 'true',
+            'p': '9386337',
+            'player': 'twitchweb',
+            'segment_preference': '4',
+            'sig': access_token['sig'],
+            'token': access_token['token'],
+        }
+
+        formats = self._extract_m3u8_formats(
+            '%s/api/channel/hls/%s.m3u8?%s'
+            % (self._USHER_BASE, channel_id, compat_urllib_parse.urlencode(query).encode('utf-8')),
+            channel_id, 'mp4')
+
+        view_count = stream.get('viewers')
+        timestamp = parse_iso8601(stream.get('created_at'))
+
+        channel = stream['channel']
+        title = self._live_title(channel.get('display_name') or channel.get('name'))
+        description = channel.get('status')
+
+        thumbnails = []
+        for thumbnail_key, thumbnail_url in stream['preview'].items():
+            m = re.search(r'(?P<width>\d+)x(?P<height>\d+)\.jpg$', thumbnail_key)
+            if not m:
+                continue
+            thumbnails.append({
+                'url': thumbnail_url,
+                'width': int(m.group('width')),
+                'height': int(m.group('height')),
+            })
+
+        return {
+            'id': compat_str(stream['_id']),
+            'display_id': channel_id,
+            'title': title,
+            'description': description,
+            'thumbnails': thumbnails,
+            'uploader': channel.get('display_name'),
+            'uploader_id': channel.get('name'),
+            'timestamp': timestamp,
+            'view_count': view_count,
+            'formats': formats,
+            'is_live': True,
+        }
\ No newline at end of file

From 7932de6352f1fcb0418e78dc049fe60b92ca6c68 Mon Sep 17 00:00:00 2001
From: Philipp Hagemeister <phihag@phihag.de>
Date: Thu, 22 Jan 2015 18:15:04 +0100
Subject: [PATCH 28/40] [hearthisat] Correct error message

---
 youtube_dl/extractor/hearthisat.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/youtube_dl/extractor/hearthisat.py b/youtube_dl/extractor/hearthisat.py
index d83072c1a..a19b31ac0 100644
--- a/youtube_dl/extractor/hearthisat.py
+++ b/youtube_dl/extractor/hearthisat.py
@@ -75,7 +75,7 @@ class HearThisAtIE(InfoExtractor):
         formats = []
         mp3_url = self._search_regex(
             r'(?s)<a class="player-link"\s+(?:[a-zA-Z0-9_:-]+="[^"]+"\s+)*?data-mp3="([^"]+)"',
-            webpage, 'title', fatal=False)
+            webpage, 'mp3 URL', fatal=False)
         if mp3_url:
             formats.append({
                 'format_id': 'mp3',

From 650ab5beebfa555383365602c914729249fad4c6 Mon Sep 17 00:00:00 2001
From: Philipp Hagemeister <phihag@phihag.de>
Date: Thu, 22 Jan 2015 18:15:58 +0100
Subject: [PATCH 29/40] [comedycentral:shows] Remove references to colbert
 report

---
 test/test_all_urls.py                 | 2 --
 youtube_dl/extractor/comedycentral.py | 4 ++--
 2 files changed, 2 insertions(+), 4 deletions(-)

diff --git a/test/test_all_urls.py b/test/test_all_urls.py
index bd4fe17bf..88fda76a4 100644
--- a/test/test_all_urls.py
+++ b/test/test_all_urls.py
@@ -115,8 +115,6 @@ class TestAllURLsMatching(unittest.TestCase):
         self.assertMatch(':ythistory', ['youtube:history'])
         self.assertMatch(':thedailyshow', ['ComedyCentralShows'])
         self.assertMatch(':tds', ['ComedyCentralShows'])
-        self.assertMatch(':colbertreport', ['ComedyCentralShows'])
-        self.assertMatch(':cr', ['ComedyCentralShows'])
 
     def test_vimeo_matching(self):
         self.assertMatch('http://vimeo.com/channels/tributes', ['vimeo:channel'])
diff --git a/youtube_dl/extractor/comedycentral.py b/youtube_dl/extractor/comedycentral.py
index 8d27af5e5..15ca361f0 100644
--- a/youtube_dl/extractor/comedycentral.py
+++ b/youtube_dl/extractor/comedycentral.py
@@ -34,12 +34,12 @@ class ComedyCentralIE(MTVServicesInfoExtractor):
 
 class ComedyCentralShowsIE(MTVServicesInfoExtractor):
     IE_DESC = 'The Daily Show / The Colbert Report'
-    # urls can be abbreviations like :thedailyshow or :colbert
+    # urls can be abbreviations like :thedailyshow
     # urls for episodes like:
     # or urls for clips like: http://www.thedailyshow.com/watch/mon-december-10-2012/any-given-gun-day
     #                     or: http://www.colbertnation.com/the-colbert-report-videos/421667/november-29-2012/moon-shattering-news
     #                     or: http://www.colbertnation.com/the-colbert-report-collections/422008/festival-of-lights/79524
-    _VALID_URL = r'''(?x)^(:(?P<shortname>tds|thedailyshow|cr|colbert|colbertnation|colbertreport)
+    _VALID_URL = r'''(?x)^(:(?P<shortname>tds|thedailyshow)
                       |https?://(:www\.)?
                           (?P<showname>thedailyshow|thecolbertreport)\.(?:cc\.)?com/
                          ((?:full-)?episodes/(?:[0-9a-z]{6}/)?(?P<episode>.*)|

From 78111136db788918452a80233e74fd4c506eb6f8 Mon Sep 17 00:00:00 2001
From: Philipp Hagemeister <phihag@phihag.de>
Date: Thu, 22 Jan 2015 18:18:21 +0100
Subject: [PATCH 30/40] [twitch] Move URL matching tests into extractor

---
 test/test_all_urls.py          | 13 -------------
 youtube_dl/extractor/twitch.py |  7 +++++--
 2 files changed, 5 insertions(+), 15 deletions(-)

diff --git a/test/test_all_urls.py b/test/test_all_urls.py
index 88fda76a4..e66264b4b 100644
--- a/test/test_all_urls.py
+++ b/test/test_all_urls.py
@@ -14,7 +14,6 @@ from test.helper import gettestcases
 from youtube_dl.extractor import (
     FacebookIE,
     gen_extractors,
-    TwitchIE,
     YoutubeIE,
 )
 
@@ -72,18 +71,6 @@ class TestAllURLsMatching(unittest.TestCase):
         self.assertMatch('http://www.youtube.com/results?search_query=making+mustard', ['youtube:search_url'])
         self.assertMatch('https://www.youtube.com/results?baz=bar&search_query=youtube-dl+test+video&filters=video&lclk=video', ['youtube:search_url'])
 
-    def test_twitch_channelid_matching(self):
-        self.assertTrue(TwitchIE.suitable('twitch.tv/vanillatv'))
-        self.assertTrue(TwitchIE.suitable('www.twitch.tv/vanillatv'))
-        self.assertTrue(TwitchIE.suitable('http://www.twitch.tv/vanillatv'))
-        self.assertTrue(TwitchIE.suitable('http://www.twitch.tv/vanillatv/'))
-
-    def test_twitch_videoid_matching(self):
-        self.assertTrue(TwitchIE.suitable('http://www.twitch.tv/vanillatv/b/328087483'))
-
-    def test_twitch_chapterid_matching(self):
-        self.assertTrue(TwitchIE.suitable('http://www.twitch.tv/tsm_theoddone/c/2349361'))
-
     def test_youtube_extract(self):
         assertExtractId = lambda url, id: self.assertEqual(YoutubeIE.extract_id(url), id)
         assertExtractId('http://www.youtube.com/watch?&v=BaW_jenozKc', 'BaW_jenozKc')
diff --git a/youtube_dl/extractor/twitch.py b/youtube_dl/extractor/twitch.py
index 8866e5182..017eff742 100644
--- a/youtube_dl/extractor/twitch.py
+++ b/youtube_dl/extractor/twitch.py
@@ -148,14 +148,17 @@ class TwitchChapterIE(TwitchItemBaseIE):
     _ITEM_TYPE = 'chapter'
     _ITEM_SHORTCUT = 'c'
 
-    _TEST = {
+    _TESTS = [{
         'url': 'http://www.twitch.tv/acracingleague/c/5285812',
         'info_dict': {
             'id': 'c5285812',
             'title': 'ACRL Off Season - Sports Cars @ Nordschleife',
         },
         'playlist_mincount': 3,
-    }
+    }, {
+        'url': 'http://www.twitch.tv/tsm_theoddone/c/2349361',
+        'only_matching': True,
+    }]
 
 
 class TwitchVodIE(TwitchItemBaseIE):

From b4a64c592b5f0f547c11c72e6c840612b336de5d Mon Sep 17 00:00:00 2001
From: Philipp Hagemeister <phihag@phihag.de>
Date: Thu, 22 Jan 2015 18:21:17 +0100
Subject: [PATCH 31/40] [README] Add an FAQ entry about destination folder

---
 README.md | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/README.md b/README.md
index 60bd7e2b6..efa48f280 100644
--- a/README.md
+++ b/README.md
@@ -488,6 +488,10 @@ To make a different directory work - either for ffmpeg, or for youtube-dl, or fo
 
 From then on, after restarting your shell, you will be able to access both youtube-dl and ffmpeg (and youtube-dl will be able to find ffmpeg) by simply typing `youtube-dl` or `ffmpeg`, no matter what directory you're in.
 
+### How do I put downloads into a specific folder?
+
+Use the `-o` to specify an [output template](#output-template), for example `-o "/home/user/videos/%(title)s-%(id)s.%(ext)s"`. If you want this for all of your downloads, put the option into your [configuration file](#configuration).
+
 ### How can I detect whether a given URL is supported by youtube-dl?
 
 For one, have a look at the [list of supported sites](docs/supportedsites). Note that it can sometimes happen that the site changes its URL scheme (say, from http://example.com/v/1234567 to http://example.com/v/1234567 ) and youtube-dl reports an URL of a service in that list as unsupported. In that case, simply report a bug.

From f353cbdb2f91564d4637462875c1a12bfbcd2ada Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Sergey=20M=E2=80=A4?= <dstftw@gmail.com>
Date: Thu, 22 Jan 2015 23:34:40 +0600
Subject: [PATCH 32/40] [twitch:stream] Randomize query

---
 youtube_dl/extractor/twitch.py | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/youtube_dl/extractor/twitch.py b/youtube_dl/extractor/twitch.py
index 14314916a..87e3c453d 100644
--- a/youtube_dl/extractor/twitch.py
+++ b/youtube_dl/extractor/twitch.py
@@ -3,6 +3,7 @@ from __future__ import unicode_literals
 
 import itertools
 import re
+import random
 
 from .common import InfoExtractor
 from ..compat import (
@@ -305,7 +306,7 @@ class TwitchStreamIE(TwitchBaseIE):
 
         query = {
             'allow_source': 'true',
-            'p': '9386337',
+            'p': random.randint(1000000, 10000000),
             'player': 'twitchweb',
             'segment_preference': '4',
             'sig': access_token['sig'],

From dc1b027cd4672c6b4772c12daf2a6a2a93469270 Mon Sep 17 00:00:00 2001
From: Philipp Hagemeister <phihag@phihag.de>
Date: Thu, 22 Jan 2015 23:06:03 +0100
Subject: [PATCH 33/40] [twitch] PEP8

---
 youtube_dl/extractor/twitch.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/youtube_dl/extractor/twitch.py b/youtube_dl/extractor/twitch.py
index 14314916a..6831bb596 100644
--- a/youtube_dl/extractor/twitch.py
+++ b/youtube_dl/extractor/twitch.py
@@ -347,4 +347,4 @@ class TwitchStreamIE(TwitchBaseIE):
             'view_count': view_count,
             'formats': formats,
             'is_live': True,
-        }
\ No newline at end of file
+        }

From 50789175edbe3aaad77f45e6fe883ba09580cc6f Mon Sep 17 00:00:00 2001
From: Philipp Hagemeister <phihag@phihag.de>
Date: Thu, 22 Jan 2015 23:48:58 +0100
Subject: [PATCH 34/40] [pornhub] Detect private videos and emit an error
 message (Closes #4764)

---
 youtube_dl/extractor/pornhub.py | 10 ++++++++++
 1 file changed, 10 insertions(+)

diff --git a/youtube_dl/extractor/pornhub.py b/youtube_dl/extractor/pornhub.py
index 634142d0d..fb2032832 100644
--- a/youtube_dl/extractor/pornhub.py
+++ b/youtube_dl/extractor/pornhub.py
@@ -10,6 +10,7 @@ from ..compat import (
     compat_urllib_request,
 )
 from ..utils import (
+    ExtractorError,
     str_to_int,
 )
 from ..aes import (
@@ -44,6 +45,15 @@ class PornHubIE(InfoExtractor):
         req.add_header('Cookie', 'age_verified=1')
         webpage = self._download_webpage(req, video_id)
 
+        error_msg = self._html_search_regex(
+            r'(?s)<div class="userMessageSection[^"]*".*?>(.*?)</div>',
+            webpage, 'error message', default=None)
+        if error_msg:
+            error_msg = re.sub(r'\s+', ' ', error_msg)
+            raise ExtractorError(
+                'PornHub said: %s' % error_msg,
+                expected=True, video_id=video_id)
+
         video_title = self._html_search_regex(r'<h1 [^>]+>([^<]+)', webpage, 'title')
         video_uploader = self._html_search_regex(
             r'(?s)From:&nbsp;.+?<(?:a href="/users/|a href="/channels/|<span class="username)[^>]+>(.+?)<',

From 083c9df93b08a24e967b68fbdd2f4a71ae74c8c8 Mon Sep 17 00:00:00 2001
From: Philipp Hagemeister <phihag@phihag.de>
Date: Fri, 23 Jan 2015 00:04:05 +0100
Subject: [PATCH 35/40] [YoutubeDL] Allow filtering by properties (Fixes #4584)

---
 test/test_YoutubeDL.py  | 55 +++++++++++++++++++++++++++++++++++++++++
 youtube_dl/YoutubeDL.py | 54 ++++++++++++++++++++++++++++++++++++++++
 youtube_dl/options.py   | 11 +++++++++
 3 files changed, 120 insertions(+)

diff --git a/test/test_YoutubeDL.py b/test/test_YoutubeDL.py
index 85d87f2c3..678b9f7d1 100644
--- a/test/test_YoutubeDL.py
+++ b/test/test_YoutubeDL.py
@@ -281,6 +281,61 @@ class TestFormatSelection(unittest.TestCase):
             downloaded = ydl.downloaded_info_dicts[0]
             self.assertEqual(downloaded['format_id'], f1id)
 
+    def test_format_filtering(self):
+        formats = [
+            {'format_id': 'A', 'filesize': 500, 'width': 1000},
+            {'format_id': 'B', 'filesize': 1000, 'width': 500},
+            {'format_id': 'C', 'filesize': 1000, 'width': 400},
+            {'format_id': 'D', 'filesize': 2000, 'width': 600},
+            {'format_id': 'E', 'filesize': 3000},
+            {'format_id': 'F'},
+            {'format_id': 'G', 'filesize': 1000000},
+        ]
+        for f in formats:
+            f['url'] = 'http://_/'
+            f['ext'] = 'unknown'
+        info_dict = _make_result(formats)
+
+        ydl = YDL({'format': 'best[filesize<3000]'})
+        ydl.process_ie_result(info_dict)
+        downloaded = ydl.downloaded_info_dicts[0]
+        self.assertEqual(downloaded['format_id'], 'D')
+
+        ydl = YDL({'format': 'best[filesize<=3000]'})
+        ydl.process_ie_result(info_dict)
+        downloaded = ydl.downloaded_info_dicts[0]
+        self.assertEqual(downloaded['format_id'], 'E')
+
+        ydl = YDL({'format': 'best[filesize <= ? 3000]'})
+        ydl.process_ie_result(info_dict)
+        downloaded = ydl.downloaded_info_dicts[0]
+        self.assertEqual(downloaded['format_id'], 'F')
+
+        ydl = YDL({'format': 'best [filesize = 1000] [width>450]'})
+        ydl.process_ie_result(info_dict)
+        downloaded = ydl.downloaded_info_dicts[0]
+        self.assertEqual(downloaded['format_id'], 'B')
+
+        ydl = YDL({'format': 'best [filesize = 1000] [width!=450]'})
+        ydl.process_ie_result(info_dict)
+        downloaded = ydl.downloaded_info_dicts[0]
+        self.assertEqual(downloaded['format_id'], 'C')
+
+        ydl = YDL({'format': '[filesize>?1]'})
+        ydl.process_ie_result(info_dict)
+        downloaded = ydl.downloaded_info_dicts[0]
+        self.assertEqual(downloaded['format_id'], 'G')
+
+        ydl = YDL({'format': '[filesize<1M]'})
+        ydl.process_ie_result(info_dict)
+        downloaded = ydl.downloaded_info_dicts[0]
+        self.assertEqual(downloaded['format_id'], 'E')
+
+        ydl = YDL({'format': '[filesize<1MiB]'})
+        ydl.process_ie_result(info_dict)
+        downloaded = ydl.downloaded_info_dicts[0]
+        self.assertEqual(downloaded['format_id'], 'G')
+
     def test_add_extra_info(self):
         test_dict = {
             'extractor': 'Foo',
diff --git a/youtube_dl/YoutubeDL.py b/youtube_dl/YoutubeDL.py
index 772fddd45..8ef74e414 100755
--- a/youtube_dl/YoutubeDL.py
+++ b/youtube_dl/YoutubeDL.py
@@ -10,6 +10,7 @@ import io
 import itertools
 import json
 import locale
+import operator
 import os
 import platform
 import re
@@ -49,6 +50,7 @@ from .utils import (
     make_HTTPS_handler,
     MaxDownloadsReached,
     PagedList,
+    parse_filesize,
     PostProcessingError,
     platform_name,
     preferredencoding,
@@ -768,7 +770,59 @@ class YoutubeDL(object):
         else:
             raise Exception('Invalid result type: %s' % result_type)
 
+    def _apply_format_filter(self, format_spec, available_formats):
+        " Returns a tuple of the remaining format_spec and filtered formats "
+
+        OPERATORS = {
+            '<': operator.lt,
+            '<=': operator.le,
+            '>': operator.gt,
+            '>=': operator.ge,
+            '=': operator.eq,
+            '!=': operator.ne,
+        }
+        operator_rex = re.compile(r'''(?x)\s*\[
+            (?P<key>width|height|tbr|abr|vbr|filesize)
+            \s*(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
+            (?P<value>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)
+            \]$
+            ''' % '|'.join(map(re.escape, OPERATORS.keys())))
+        m = operator_rex.search(format_spec)
+        if not m:
+            raise ValueError('Invalid format specification %r' % format_spec)
+
+        try:
+            comparison_value = int(m.group('value'))
+        except ValueError:
+            comparison_value = parse_filesize(m.group('value'))
+            if comparison_value is None:
+                comparison_value = parse_filesize(m.group('value') + 'B')
+            if comparison_value is None:
+                raise ValueError(
+                    'Invalid value %r in format specification %r' % (
+                        m.group('value'), format_spec))
+        op = OPERATORS[m.group('op')]
+
+        def _filter(f):
+            actual_value = f.get(m.group('key'))
+            if actual_value is None:
+                return m.group('none_inclusive')
+            return op(actual_value, comparison_value)
+        new_formats = [f for f in available_formats if _filter(f)]
+
+        new_format_spec = format_spec[:-len(m.group(0))]
+        if not new_format_spec:
+            new_format_spec = 'best'
+
+        return (new_format_spec, new_formats)
+
     def select_format(self, format_spec, available_formats):
+        while format_spec.endswith(']'):
+            format_spec, available_formats = self._apply_format_filter(
+                format_spec, available_formats)
+        if not available_formats:
+            return None
+
         if format_spec == 'best' or format_spec is None:
             return available_formats[-1]
         elif format_spec == 'worst':
diff --git a/youtube_dl/options.py b/youtube_dl/options.py
index f25c12e52..fd7b400b2 100644
--- a/youtube_dl/options.py
+++ b/youtube_dl/options.py
@@ -289,6 +289,17 @@ def parseOpts(overrideArguments=None):
             'extensions aac, m4a, mp3, mp4, ogg, wav, webm. '
             'You can also use the special names "best",'
             ' "bestvideo", "bestaudio", "worst". '
+            ' You can filter the video results by putting a condition in'
+            ' brackets, as in -f "best[height=720]"'
+            ' (or -f "[filesize>10M]"). '
+            ' This works for filesize, height, width, tbr, abr, and vbr'
+            ' and the comparisons <, <=, >, >=, =, != .'
+            ' Formats for which the value is not known are excluded unless you'
+            ' put a question mark (?) after the operator.'
+            ' You can combine format filters, so  '
+            '-f "[height <=? 720][tbr>500]" '
+            'selects up to 720p videos (or videos where the height is not '
+            'known) with a bitrate of at least 500 KBit/s.'
             ' By default, youtube-dl will pick the best quality.'
             ' Use commas to download multiple audio formats, such as'
             ' -f  136/137/mp4/bestvideo,140/m4a/bestaudio.'

From c994e6bd63feb95e1e385af2333ae2f615314ac5 Mon Sep 17 00:00:00 2001
From: Philipp Hagemeister <phihag@phihag.de>
Date: Fri, 23 Jan 2015 00:06:11 +0100
Subject: [PATCH 36/40] release 2015.01.23

---
 README.md             | 20 ++++++++++++++++----
 youtube_dl/version.py |  2 +-
 2 files changed, 17 insertions(+), 5 deletions(-)

diff --git a/README.md b/README.md
index efa48f280..1408ebba0 100644
--- a/README.md
+++ b/README.md
@@ -267,10 +267,22 @@ which means you can modify it, redistribute it or use it however you like.
                                      by extension for the extensions aac, m4a,
                                      mp3, mp4, ogg, wav, webm. You can also use
                                      the special names "best", "bestvideo",
-                                     "bestaudio", "worst".  By default, youtube-
-                                     dl will pick the best quality. Use commas
-                                     to download multiple audio formats, such as
-                                     -f
+                                     "bestaudio", "worst".  You can filter the
+                                     video results by putting a condition in
+                                     brackets, as in -f "best[height=720]" (or
+                                     -f "[filesize>10M]").  This works for
+                                     filesize, height, width, tbr, abr, and vbr
+                                     and the comparisons <, <=, >, >=, =, != .
+                                     Formats for which the value is not known
+                                     are excluded unless you put a question mark
+                                     (?) after the operator. You can combine
+                                     format filters, so  -f "[height <=?
+                                     720][tbr>500]" selects up to 720p videos
+                                     (or videos where the height is not known)
+                                     with a bitrate of at least 500 KBit/s. By
+                                     default, youtube-dl will pick the best
+                                     quality. Use commas to download multiple
+                                     audio formats, such as -f
                                      136/137/mp4/bestvideo,140/m4a/bestaudio.
                                      You can merge the video and audio of two
                                      formats into a single file using -f <video-
diff --git a/youtube_dl/version.py b/youtube_dl/version.py
index f5df7be31..6c552064b 100644
--- a/youtube_dl/version.py
+++ b/youtube_dl/version.py
@@ -1,3 +1,3 @@
 from __future__ import unicode_literals
 
-__version__ = '2015.01.22'
+__version__ = '2015.01.23'

From c2a30b250cdea5a8c73225e6a14707a669640de0 Mon Sep 17 00:00:00 2001
From: Philipp Hagemeister <phihag@phihag.de>
Date: Fri, 23 Jan 2015 00:32:35 +0100
Subject: [PATCH 37/40] [testtube] Add new extractor (Fixes #4759)

---
 youtube_dl/extractor/__init__.py |  1 +
 youtube_dl/extractor/testtube.py | 60 ++++++++++++++++++++++++++++++++
 2 files changed, 61 insertions(+)
 create mode 100644 youtube_dl/extractor/testtube.py

diff --git a/youtube_dl/extractor/__init__.py b/youtube_dl/extractor/__init__.py
index a983fff76..9ab90ac62 100644
--- a/youtube_dl/extractor/__init__.py
+++ b/youtube_dl/extractor/__init__.py
@@ -432,6 +432,7 @@ from .telemb import TeleMBIE
 from .teletask import TeleTaskIE
 from .tenplay import TenPlayIE
 from .testurl import TestURLIE
+from .testtube import TestTubeIE
 from .tf1 import TF1IE
 from .theonion import TheOnionIE
 from .theplatform import ThePlatformIE
diff --git a/youtube_dl/extractor/testtube.py b/youtube_dl/extractor/testtube.py
new file mode 100644
index 000000000..fd47e71a2
--- /dev/null
+++ b/youtube_dl/extractor/testtube.py
@@ -0,0 +1,60 @@
+from __future__ import unicode_literals
+
+from .common import InfoExtractor
+from ..utils import int_or_none
+
+
+class TestTubeIE(InfoExtractor):
+    _VALID_URL = r'https?://testtube\.com/[^/?#]+/(?P<id>[^/?#]+)'
+    _TESTS = [{
+        'url': 'https://testtube.com/dnews/5-weird-ways-plants-can-eat-animals?utm_source=FB&utm_medium=DNews&utm_campaign=DNewsSocial',
+        'info_dict': {
+            'id': '60163',
+            'display_id': '5-weird-ways-plants-can-eat-animals',
+            'duration': 275,
+            'ext': 'mp4',
+            'title': '5 Weird Ways Plants Can Eat Animals',
+            'description': 'Why have some plants evolved to eat meat?',
+            'thumbnail': 're:^https?://.*\.jpg$',
+            'uploader': 'DNews',
+            'uploader_id': 'dnews',
+        },
+    }]
+
+    def _real_extract(self, url):
+        display_id = self._match_id(url)
+
+        webpage = self._download_webpage(url, display_id)
+        video_id = self._search_regex(
+            r"player\.loadRevision3Item\('video_id',\s*([0-9]+)\);",
+            webpage, 'video ID')
+
+        all_info = self._download_json(
+            'https://testtube.com/api/getPlaylist.json?api_key=ba9c741bce1b9d8e3defcc22193f3651b8867e62&codecs=h264,vp8,theora&video_id=%s' % video_id,
+            video_id)
+        info = all_info['items'][0]
+
+        formats = []
+        for vcodec, fdatas in info['media'].items():
+            for name, fdata in fdatas.items():
+                formats.append({
+                    'format_id': '%s-%s' % (vcodec, name),
+                    'url': fdata['url'],
+                    'vcodec': vcodec,
+                    'tbr': fdata.get('bitrate'),
+                })
+        self._sort_formats(formats)
+
+        duration = int_or_none(info.get('duration'))
+
+        return {
+            'id': video_id,
+            'display_id': display_id,
+            'title': info['title'],
+            'description': info.get('summary'),
+            'thumbnail': info.get('images', {}).get('large'),
+            'uploader': info.get('show', {}).get('name'),
+            'uploader_id': info.get('show', {}).get('slug'),
+            'duration': duration,
+            'formats': formats,
+        }

From 317239b097a913824e20d436cdb8c74161564268 Mon Sep 17 00:00:00 2001
From: Philipp Hagemeister <phihag@phihag.de>
Date: Fri, 23 Jan 2015 00:33:14 +0100
Subject: [PATCH 38/40] release 2015.01.23.1

---
 youtube_dl/version.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/youtube_dl/version.py b/youtube_dl/version.py
index 6c552064b..28458fd69 100644
--- a/youtube_dl/version.py
+++ b/youtube_dl/version.py
@@ -1,3 +1,3 @@
 from __future__ import unicode_literals
 
-__version__ = '2015.01.23'
+__version__ = '2015.01.23.1'

From 61ca9a80b34b14fa0e5e19e35ee76e0086d49edf Mon Sep 17 00:00:00 2001
From: Philipp Hagemeister <phihag@phihag.de>
Date: Fri, 23 Jan 2015 01:21:30 +0100
Subject: [PATCH 39/40] [generic] Add support for BOMs (Fixes #4753)

---
 test/test_utils.py              | 17 +++++++++++++++++
 youtube_dl/extractor/generic.py |  3 ++-
 youtube_dl/utils.py             | 20 ++++++++++++++++++++
 3 files changed, 39 insertions(+), 1 deletion(-)

diff --git a/test/test_utils.py b/test/test_utils.py
index 206760d99..bdd7f268a 100644
--- a/test/test_utils.py
+++ b/test/test_utils.py
@@ -28,6 +28,7 @@ from youtube_dl.utils import (
     fix_xml_ampersands,
     InAdvancePagedList,
     intlist_to_bytes,
+    is_html,
     js_to_json,
     limit_length,
     OnDemandPagedList,
@@ -417,5 +418,21 @@ ffmpeg version 2.4.4 Copyright (c) 2000-2014 the FFmpeg ...'''), '2.4.4')
         self.assertTrue(age_restricted(18, 14))
         self.assertFalse(age_restricted(18, 18))
 
+    def test_is_html(self):
+        self.assertFalse(is_html(b'\x49\x44\x43<html'))
+        self.assertTrue(is_html(b'<!DOCTYPE foo>\xaaa'))
+        self.assertTrue(is_html(  # UTF-8 with BOM
+            b'\xef\xbb\xbf<!DOCTYPE foo>\xaaa'))
+        self.assertTrue(is_html(  # UTF-16-LE
+            b'\xff\xfe<\x00h\x00t\x00m\x00l\x00>\x00\xe4\x00'
+        ))
+        self.assertTrue(is_html(  # UTF-16-BE
+            b'\xfe\xff\x00<\x00h\x00t\x00m\x00l\x00>\x00\xe4'
+        ))
+        self.assertTrue(is_html(  # UTF-32-BE
+            b'\x00\x00\xFE\xFF\x00\x00\x00<\x00\x00\x00h\x00\x00\x00t\x00\x00\x00m\x00\x00\x00l\x00\x00\x00>\x00\x00\x00\xe4'))
+        self.assertTrue(is_html(  # UTF-32-LE
+            b'\xFF\xFE\x00\x00<\x00\x00\x00h\x00\x00\x00t\x00\x00\x00m\x00\x00\x00l\x00\x00\x00>\x00\x00\x00\xe4\x00\x00\x00'))
+
 if __name__ == '__main__':
     unittest.main()
diff --git a/youtube_dl/extractor/generic.py b/youtube_dl/extractor/generic.py
index 7a5bf9392..b893d8149 100644
--- a/youtube_dl/extractor/generic.py
+++ b/youtube_dl/extractor/generic.py
@@ -17,6 +17,7 @@ from ..utils import (
     ExtractorError,
     float_or_none,
     HEADRequest,
+    is_html,
     orderedSet,
     parse_xml,
     smuggle_url,
@@ -647,7 +648,7 @@ class GenericIE(InfoExtractor):
         # Maybe it's a direct link to a video?
         # Be careful not to download the whole thing!
         first_bytes = full_response.read(512)
-        if not re.match(r'^\s*<', first_bytes.decode('utf-8', 'replace')):
+        if not is_html(first_bytes):
             self._downloader.report_warning(
                 'URL could be a direct video link, returning it as such.')
             upload_date = unified_strdate(
diff --git a/youtube_dl/utils.py b/youtube_dl/utils.py
index 764474c33..b433b591b 100644
--- a/youtube_dl/utils.py
+++ b/youtube_dl/utils.py
@@ -1631,3 +1631,23 @@ def age_restricted(content_limit, age_limit):
     if content_limit is None:
         return False  # Content available for everyone
     return age_limit < content_limit
+
+
+def is_html(first_bytes):
+    """ Detect whether a file contains HTML by examining its first bytes. """
+
+    BOMS = [
+        (b'\xef\xbb\xbf', 'utf-8'),
+        (b'\x00\x00\xfe\xff', 'utf-32-be'),
+        (b'\xff\xfe\x00\x00', 'utf-32-le'),
+        (b'\xff\xfe', 'utf-16-le'),
+        (b'\xfe\xff', 'utf-16-be'),
+    ]
+    for bom, enc in BOMS:
+        if first_bytes.startswith(bom):
+            s = first_bytes[len(bom):].decode(enc, 'replace')
+            break
+    else:
+        s = first_bytes.decode('utf-8', 'replace')
+
+    return re.match(r'^\s*<', s)

From 08ff6ab07e9d08853ca19cf3b8745fac11abf0f6 Mon Sep 17 00:00:00 2001
From: Philipp Hagemeister <phihag@phihag.de>
Date: Fri, 23 Jan 2015 01:34:24 +0100
Subject: [PATCH 40/40] [gamestar] Modernize slightly

---
 youtube_dl/extractor/gamestar.py | 6 +-----
 1 file changed, 1 insertion(+), 5 deletions(-)

diff --git a/youtube_dl/extractor/gamestar.py b/youtube_dl/extractor/gamestar.py
index 50f8fc7e7..7591a151e 100644
--- a/youtube_dl/extractor/gamestar.py
+++ b/youtube_dl/extractor/gamestar.py
@@ -1,8 +1,6 @@
 # coding: utf-8
 from __future__ import unicode_literals
 
-import re
-
 from .common import InfoExtractor
 from ..utils import (
     int_or_none,
@@ -29,9 +27,7 @@ class GameStarIE(InfoExtractor):
     }
 
     def _real_extract(self, url):
-        mobj = re.match(self._VALID_URL, url)
-        video_id = mobj.group('id')
-
+        video_id = self._match_id(url)
         webpage = self._download_webpage(url, video_id)
 
         og_title = self._og_search_title(webpage)