+from __future__ import unicode_literals
+
import base64
+import datetime
import hashlib
import json
import netrc
import time
import xml.etree.ElementTree
-from ..utils import (
+from ..compat import (
+ compat_cookiejar,
compat_http_client,
compat_urllib_error,
compat_urllib_parse_urlparse,
+ compat_urlparse,
compat_str,
-
+)
+from ..utils import (
+ age_restricted,
clean_html,
compiled_regex_type,
ExtractorError,
+ float_or_none,
int_or_none,
RegexNotFoundError,
sanitize_filename,
information about the video (or videos) the URL refers to. This
information includes the real video URL, the video title, author and
others. The information is stored in a dictionary which is then
- passed to the FileDownloader. The FileDownloader processes this
+ passed to the YoutubeDL. The YoutubeDL processes this
information possibly downloading the video to the file system, among
other possible outcomes.
- The dictionaries must include the following fields:
+ The type field determines the the type of the result.
+ By far the most common value (and the default if _type is missing) is
+ "video", which indicates a single video.
+
+ For a video, the dictionaries must include the following fields:
id: Video identifier.
title: Video title, unescaped.
* acodec Name of the audio codec in use
* asr Audio sampling rate in Hertz
* vbr Average video bitrate in KBit/s
+ * fps Frame rate
* vcodec Name of the video codec in use
* container Name of the container format
* filesize The number of bytes, if known in advance
by this field, regardless of all other values.
-1 for default (order by other properties),
-2 or smaller for less than default.
+ < -1000 to hide the format (if there is
+ another one which is strictly better)
+ * language_preference Is this in the correct requested
+ language?
+ 10 if it's what the URL is about,
+ -1 for default (don't know),
+ -10 otherwise, other values reserved for now.
* quality Order number of the video quality of this
format, irrespective of the file format.
-1 for default (order by other properties),
-2 or smaller for less than default.
+ * source_preference Order number for this video source
+ (quality takes higher priority)
+ -1 for default (order by other properties),
+ -2 or smaller for less than default.
* http_referer HTTP Referer header value to set.
* http_method HTTP method to use for the download.
* http_headers A dictionary of additional HTTP headers
The following fields are optional:
+ alt_title: A secondary title of the video.
display_id An alternative identifier for the video, not necessarily
unique, but available before title. Typically, id is
something like "4234987", title "Dancing naked mole rats",
* "resolution" (optional, string "{width}x{height"},
deprecated)
thumbnail: Full URL to a video thumbnail image.
- description: One-line video description.
+ description: Full video description.
uploader: Full name of the video uploader.
timestamp: UNIX timestamp of the moment the video became available.
upload_date: Video upload date (YYYYMMDD).
by YoutubeDL if it's missing)
categories: A list of categories that the video falls in, for example
["Sports", "Berlin"]
+ is_live: True, False, or None (=unknown). Whether this video is a
+ live stream that goes on instead of a fixed-length video.
Unless mentioned otherwise, the fields should be Unicode strings.
+ Unless mentioned otherwise, None is equivalent to absence of information.
+
+
+ _type "playlist" indicates multiple videos.
+ There must be a key "entries", which is a list, an iterable, or a PagedList
+ object, each element of which is a valid dictionary by this specification.
+
+ Additionally, playlists can have "title" and "id" attributes with the same
+ semantics as videos (see above).
+
+
+ _type "multi_video" indicates that there are multiple videos that
+ form a single show, for examples multiple acts of an opera or TV episode.
+ It must have an entries key like a playlist and contain all the keys
+ required for a video at the same time.
+
+
+ _type "url" indicates that the video must be extracted from another
+ location, possibly by a different extractor. Its only required key is:
+ "url" - the next URL to extract.
+ The key "ie_key" can be set to the class name (minus the trailing "IE",
+ e.g. "Youtube") if the extractor class is known in advance.
+ Additionally, the dictionary may have any properties of the resolved entity
+ known in advance, for example "title" if the title of the referred video is
+ known ahead of time.
+
+
+ _type "url_transparent" entities have the same specification as "url", but
+ indicate that the given additional information is more precise than the one
+ associated with the resolved URL.
+ This is useful when a site employs a video service that hosts the video and
+ its technical metadata, but that video service does not embed a useful
+ title, description etc.
+
+
Subclasses of this one should re-define the _real_initialize() and
_real_extract() methods and define a _VALID_URL regexp.
Probably, they should also be added to the list of extractors.
cls._VALID_URL_RE = re.compile(cls._VALID_URL)
return cls._VALID_URL_RE.match(url) is not None
+ @classmethod
+ def _match_id(cls, url):
+ if '_VALID_URL_RE' not in cls.__dict__:
+ cls._VALID_URL_RE = re.compile(cls._VALID_URL)
+ m = cls._VALID_URL_RE.match(url)
+ assert m
+ return m.group('id')
+
@classmethod
def working(cls):
"""Getter method for _WORKING."""
self.report_download_webpage(video_id)
elif note is not False:
if video_id is None:
- self.to_screen(u'%s' % (note,))
+ self.to_screen('%s' % (note,))
else:
- self.to_screen(u'%s: %s' % (video_id, note))
+ self.to_screen('%s: %s' % (video_id, note))
try:
return self._downloader.urlopen(url_or_request)
except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
if errnote is False:
return False
if errnote is None:
- errnote = u'Unable to download webpage'
- errmsg = u'%s: %s' % (errnote, compat_str(err))
+ errnote = 'Unable to download webpage'
+ errmsg = '%s: %s' % (errnote, compat_str(err))
if fatal:
raise ExtractorError(errmsg, sys.exc_info()[2], cause=err)
else:
def _download_webpage_handle(self, url_or_request, video_id, note=None, errnote=None, fatal=True):
""" Returns a tuple (page content as string, URL handle) """
-
# Strip hashes from the URL (#1038)
if isinstance(url_or_request, (compat_str, str)):
url_or_request = url_or_request.partition('#')[0]
if urlh is False:
assert not fatal
return False
+ content = self._webpage_read_content(urlh, url_or_request, video_id, note, errnote, fatal)
+ return (content, urlh)
+
+ def _webpage_read_content(self, urlh, url_or_request, video_id, note=None, errnote=None, fatal=True, prefix=None):
content_type = urlh.headers.get('Content-Type', '')
webpage_bytes = urlh.read()
+ if prefix is not None:
+ webpage_bytes = prefix + webpage_bytes
m = re.match(r'[a-zA-Z0-9_.-]+/[a-zA-Z0-9_.-]+\s*;\s*charset=(.+)', content_type)
if m:
encoding = m.group(1)
url = url_or_request.get_full_url()
except AttributeError:
url = url_or_request
- self.to_screen(u'Dumping request to ' + url)
+ self.to_screen('Dumping request to ' + url)
dump = base64.b64encode(webpage_bytes).decode('ascii')
self._downloader.to_screen(dump)
if self._downloader.params.get('write_pages', False):
url = url_or_request
basen = '%s_%s' % (video_id, url)
if len(basen) > 240:
- h = u'___' + hashlib.md5(basen.encode('utf-8')).hexdigest()
+ h = '___' + hashlib.md5(basen.encode('utf-8')).hexdigest()
basen = basen[:240 - len(h)] + h
raw_filename = basen + '.dump'
filename = sanitize_filename(raw_filename, restricted=True)
- self.to_screen(u'Saving request to ' + filename)
+ self.to_screen('Saving request to ' + filename)
+ # Working around MAX_PATH limitation on Windows (see
+ # http://msdn.microsoft.com/en-us/library/windows/desktop/aa365247(v=vs.85).aspx)
+ if os.name == 'nt':
+ absfilepath = os.path.abspath(filename)
+ if len(absfilepath) > 259:
+ filename = '\\\\?\\' + absfilepath
with open(filename, 'wb') as outf:
outf.write(webpage_bytes)
except LookupError:
content = webpage_bytes.decode('utf-8', 'replace')
- if (u'<title>Access to this site is blocked</title>' in content and
- u'Websense' in content[:512]):
- msg = u'Access to this webpage has been blocked by Websense filtering software in your network.'
+ if ('<title>Access to this site is blocked</title>' in content and
+ 'Websense' in content[:512]):
+ msg = 'Access to this webpage has been blocked by Websense filtering software in your network.'
blocked_iframe = self._html_search_regex(
r'<iframe src="([^"]+)"', content,
- u'Websense information URL', default=None)
+ 'Websense information URL', default=None)
if blocked_iframe:
- msg += u' Visit %s for more details' % blocked_iframe
+ msg += ' Visit %s for more details' % blocked_iframe
raise ExtractorError(msg, expected=True)
- return (content, urlh)
+ return content
def _download_webpage(self, url_or_request, video_id, note=None, errnote=None, fatal=True):
""" Returns the data of the page as a string """
return content
def _download_xml(self, url_or_request, video_id,
- note=u'Downloading XML', errnote=u'Unable to download XML',
+ note='Downloading XML', errnote='Unable to download XML',
transform_source=None, fatal=True):
"""Return the xml as an xml.etree.ElementTree.Element"""
xml_string = self._download_webpage(
return xml.etree.ElementTree.fromstring(xml_string.encode('utf-8'))
def _download_json(self, url_or_request, video_id,
- note=u'Downloading JSON metadata',
- errnote=u'Unable to download JSON metadata',
+ note='Downloading JSON metadata',
+ errnote='Unable to download JSON metadata',
transform_source=None,
fatal=True):
json_string = self._download_webpage(
url_or_request, video_id, note, errnote, fatal=fatal)
if (not fatal) and json_string is False:
return None
+ return self._parse_json(
+ json_string, video_id, transform_source=transform_source, fatal=fatal)
+
+ def _parse_json(self, json_string, video_id, transform_source=None, fatal=True):
if transform_source:
json_string = transform_source(json_string)
try:
return json.loads(json_string)
except ValueError as ve:
- raise ExtractorError('Failed to download JSON', cause=ve)
+ errmsg = '%s: Failed to parse JSON ' % video_id
+ if fatal:
+ raise ExtractorError(errmsg, cause=ve)
+ else:
+ self.report_warning(errmsg + str(ve))
def report_warning(self, msg, video_id=None):
- idstr = u'' if video_id is None else u'%s: ' % video_id
+ idstr = '' if video_id is None else '%s: ' % video_id
self._downloader.report_warning(
- u'[%s] %s%s' % (self.IE_NAME, idstr, msg))
+ '[%s] %s%s' % (self.IE_NAME, idstr, msg))
def to_screen(self, msg):
"""Print msg to screen, prefixing it with '[ie_name]'"""
- self._downloader.to_screen(u'[%s] %s' % (self.IE_NAME, msg))
+ self._downloader.to_screen('[%s] %s' % (self.IE_NAME, msg))
def report_extraction(self, id_or_name):
"""Report information extraction."""
- self.to_screen(u'%s: Extracting information' % id_or_name)
+ self.to_screen('%s: Extracting information' % id_or_name)
def report_download_webpage(self, video_id):
"""Report webpage download."""
- self.to_screen(u'%s: Downloading webpage' % video_id)
+ self.to_screen('%s: Downloading webpage' % video_id)
def report_age_confirmation(self):
"""Report attempt to confirm age."""
- self.to_screen(u'Confirming age')
+ self.to_screen('Confirming age')
def report_login(self):
"""Report attempt to log in."""
- self.to_screen(u'Logging in')
+ self.to_screen('Logging in')
- #Methods for following #608
+ # Methods for following #608
@staticmethod
def url_result(url, ie=None, video_id=None):
"""Returns a url that points to a page that should be processed"""
- #TODO: ie should be the class used for getting the info
+ # TODO: ie should be the class used for getting the info
video_info = {'_type': 'url',
'url': url,
'ie_key': ie}
if video_id is not None:
video_info['id'] = video_id
return video_info
+
@staticmethod
- def playlist_result(entries, playlist_id=None, playlist_title=None):
+ def playlist_result(entries, playlist_id=None, playlist_title=None, playlist_description=None):
"""Returns a playlist"""
video_info = {'_type': 'playlist',
'entries': entries}
video_info['id'] = playlist_id
if playlist_title:
video_info['title'] = playlist_title
+ if playlist_description:
+ video_info['description'] = playlist_description
return video_info
- def _search_regex(self, pattern, string, name, default=_NO_DEFAULT, fatal=True, flags=0):
+ def _search_regex(self, pattern, string, name, default=_NO_DEFAULT, fatal=True, flags=0, group=None):
"""
Perform a regex search on the given string, using a single or a list of
patterns returning the first matching group.
break
if os.name != 'nt' and sys.stderr.isatty():
- _name = u'\033[0;34m%s\033[0m' % name
+ _name = '\033[0;34m%s\033[0m' % name
else:
_name = name
if mobj:
- # return the first matching group
- return next(g for g in mobj.groups() if g is not None)
+ if group is None:
+ # return the first matching group
+ return next(g for g in mobj.groups() if g is not None)
+ else:
+ return mobj.group(group)
elif default is not _NO_DEFAULT:
return default
elif fatal:
- raise RegexNotFoundError(u'Unable to extract %s' % _name)
+ raise RegexNotFoundError('Unable to extract %s' % _name)
else:
- self._downloader.report_warning(u'unable to extract %s; '
- u'please report this issue on http://yt-dl.org/bug' % _name)
+ self._downloader.report_warning('unable to extract %s; '
+ 'please report this issue on http://yt-dl.org/bug' % _name)
return None
- def _html_search_regex(self, pattern, string, name, default=_NO_DEFAULT, fatal=True, flags=0):
+ def _html_search_regex(self, pattern, string, name, default=_NO_DEFAULT, fatal=True, flags=0, group=None):
"""
Like _search_regex, but strips HTML tags and unescapes entities.
"""
- res = self._search_regex(pattern, string, name, default, fatal, flags)
+ res = self._search_regex(pattern, string, name, default, fatal, flags, group)
if res:
return clean_html(res).strip()
else:
else:
raise netrc.NetrcParseError('No authenticators for %s' % self._NETRC_MACHINE)
except (IOError, netrc.NetrcParseError) as err:
- self._downloader.report_warning(u'parsing .netrc: %s' % compat_str(err))
-
+ self._downloader.report_warning('parsing .netrc: %s' % compat_str(err))
+
return (username, password)
def _get_tfa_info(self):
return unescapeHTML(escaped)
def _og_search_thumbnail(self, html, **kargs):
- return self._og_search_property('image', html, u'thumbnail url', fatal=False, **kargs)
+ return self._og_search_property('image', html, 'thumbnail url', fatal=False, **kargs)
def _og_search_description(self, html, **kargs):
return self._og_search_property('description', html, fatal=False, **kargs)
if display_name is None:
display_name = name
return self._html_search_regex(
- r'''(?ix)<meta
- (?=[^>]+(?:itemprop|name|property)=["\']?%s["\']?)
- [^>]+content=["\']([^"\']+)["\']''' % re.escape(name),
- html, display_name, fatal=fatal, **kwargs)
+ r'''(?isx)<meta
+ (?=[^>]+(?:itemprop|name|property)=(["\']?)%s\1)
+ [^>]+?content=(["\'])(?P<content>.*?)\2''' % re.escape(name),
+ html, display_name, fatal=fatal, group='content', **kwargs)
def _dc_search_uploader(self, html):
return self._html_search_meta('dc.creator', html, 'uploader')
def _twitter_search_player(self, html):
return self._html_search_meta('twitter:player', html,
- 'twitter card player')
+ 'twitter card player')
def _sort_formats(self, formats):
if not formats:
- raise ExtractorError(u'No video formats found')
+ raise ExtractorError('No video formats found')
def _formats_key(f):
# TODO remove the following workaround
if f.get('vcodec') == 'none': # audio only
if self._downloader.params.get('prefer_free_formats'):
- ORDER = [u'aac', u'mp3', u'm4a', u'webm', u'ogg', u'opus']
+ ORDER = ['aac', 'mp3', 'm4a', 'webm', 'ogg', 'opus']
else:
- ORDER = [u'webm', u'opus', u'ogg', u'mp3', u'aac', u'm4a']
+ ORDER = ['webm', 'opus', 'ogg', 'mp3', 'aac', 'm4a']
ext_preference = 0
try:
audio_ext_preference = ORDER.index(f['ext'])
audio_ext_preference = -1
else:
if self._downloader.params.get('prefer_free_formats'):
- ORDER = [u'flv', u'mp4', u'webm']
+ ORDER = ['flv', 'mp4', 'webm']
else:
- ORDER = [u'webm', u'flv', u'mp4']
+ ORDER = ['webm', 'flv', 'mp4']
try:
ext_preference = ORDER.index(f['ext'])
except ValueError:
return (
preference,
+ f.get('language_preference') if f.get('language_preference') is not None else -1,
f.get('quality') if f.get('quality') is not None else -1,
f.get('height') if f.get('height') is not None else -1,
f.get('width') if f.get('width') is not None else -1,
f.get('vbr') if f.get('vbr') is not None else -1,
f.get('abr') if f.get('abr') is not None else -1,
audio_ext_preference,
+ f.get('fps') if f.get('fps') is not None else -1,
f.get('filesize') if f.get('filesize') is not None else -1,
f.get('filesize_approx') if f.get('filesize_approx') is not None else -1,
+ f.get('source_preference') if f.get('source_preference') is not None else -1,
f.get('format_id'),
)
formats.sort(key=_formats_key)
def http_scheme(self):
- """ Either "https:" or "https:", depending on the user's preferences """
+ """ Either "http:" or "https:", depending on the user's preferences """
return (
'http:'
if self._downloader.params.get('prefer_insecure', False)
def _sleep(self, timeout, video_id, msg_template=None):
if msg_template is None:
- msg_template = u'%(video_id)s: Waiting for %(timeout)s seconds'
+ msg_template = '%(video_id)s: Waiting for %(timeout)s seconds'
msg = msg_template % {'video_id': video_id, 'timeout': timeout}
self.to_screen(msg)
time.sleep(timeout)
return formats
- def _extract_m3u8_formats(self, m3u8_url, video_id, ext=None):
+ def _extract_m3u8_formats(self, m3u8_url, video_id, ext=None,
+ entry_protocol='m3u8', preference=None):
+
formats = [{
'format_id': 'm3u8-meta',
'url': m3u8_url,
'format_note': 'Quality selection URL',
}]
- m3u8_doc = self._download_webpage(m3u8_url, video_id)
+ format_url = lambda u: (
+ u
+ if re.match(r'^https?://', u)
+ else compat_urlparse.urljoin(m3u8_url, u))
+
+ m3u8_doc = self._download_webpage(
+ m3u8_url, video_id,
+ note='Downloading m3u8 information',
+ errnote='Failed to download m3u8 information')
last_info = None
kv_rex = re.compile(
r'(?P<key>[a-zA-Z_-]+)=(?P<val>"[^"]+"|[^",]+)(?:,|$)')
elif line.startswith('#') or not line.strip():
continue
else:
+ if last_info is None:
+ formats.append({'url': format_url(line)})
+ continue
tbr = int_or_none(last_info.get('BANDWIDTH'), scale=1000)
f = {
'format_id': 'm3u8-%d' % (tbr if tbr else len(formats)),
- 'url': line.strip(),
+ 'url': format_url(line.strip()),
'tbr': tbr,
'ext': ext,
+ 'protocol': entry_protocol,
+ 'preference': preference,
}
codecs = last_info.get('CODECS')
if codecs:
- video, audio = codecs.split(',')
- f['vcodec'] = video.partition('.')[0]
- f['acodec'] = audio.partition('.')[0]
+ # TODO: looks like video codec is not always necessarily goes first
+ va_codecs = codecs.split(',')
+ if va_codecs[0]:
+ f['vcodec'] = va_codecs[0].partition('.')[0]
+ if len(va_codecs) > 1 and va_codecs[1]:
+ f['acodec'] = va_codecs[1].partition('.')[0]
resolution = last_info.get('RESOLUTION')
if resolution:
width_str, height_str = resolution.split('x')
self._sort_formats(formats)
return formats
+ # TODO: improve extraction
+ def _extract_smil_formats(self, smil_url, video_id):
+ smil = self._download_xml(
+ smil_url, video_id, 'Downloading SMIL file',
+ 'Unable to download SMIL file')
+
+ base = smil.find('./head/meta').get('base')
+
+ formats = []
+ rtmp_count = 0
+ for video in smil.findall('./body/switch/video'):
+ src = video.get('src')
+ if not src:
+ continue
+ bitrate = int_or_none(video.get('system-bitrate') or video.get('systemBitrate'), 1000)
+ width = int_or_none(video.get('width'))
+ height = int_or_none(video.get('height'))
+ proto = video.get('proto')
+ if not proto:
+ if base:
+ if base.startswith('rtmp'):
+ proto = 'rtmp'
+ elif base.startswith('http'):
+ proto = 'http'
+ ext = video.get('ext')
+ if proto == 'm3u8':
+ formats.extend(self._extract_m3u8_formats(src, video_id, ext))
+ elif proto == 'rtmp':
+ rtmp_count += 1
+ streamer = video.get('streamer') or base
+ formats.append({
+ 'url': streamer,
+ 'play_path': src,
+ 'ext': 'flv',
+ 'format_id': 'rtmp-%d' % (rtmp_count if bitrate is None else bitrate),
+ 'tbr': bitrate,
+ 'width': width,
+ 'height': height,
+ })
+ self._sort_formats(formats)
+
+ return formats
+
+ def _live_title(self, name):
+ """ Generate the title for a live video """
+ now = datetime.datetime.now()
+ now_str = now.strftime("%Y-%m-%d %H:%M")
+ return name + ' ' + now_str
+
+ def _int(self, v, name, fatal=False, **kwargs):
+ res = int_or_none(v, **kwargs)
+ if 'get_attr' in kwargs:
+ print(getattr(v, kwargs['get_attr']))
+ if res is None:
+ msg = 'Failed to extract %s: Could not parse value %r' % (name, v)
+ if fatal:
+ raise ExtractorError(msg)
+ else:
+ self._downloader.report_warning(msg)
+ return res
+
+ def _float(self, v, name, fatal=False, **kwargs):
+ res = float_or_none(v, **kwargs)
+ if res is None:
+ msg = 'Failed to extract %s: Could not parse value %r' % (name, v)
+ if fatal:
+ raise ExtractorError(msg)
+ else:
+ self._downloader.report_warning(msg)
+ return res
+
+ def _set_cookie(self, domain, name, value, expire_time=None):
+ cookie = compat_cookiejar.Cookie(
+ 0, name, value, None, None, domain, None,
+ None, '/', True, False, expire_time, '', None, None, None)
+ self._downloader.cookiejar.set_cookie(cookie)
+
+ def get_testcases(self, include_onlymatching=False):
+ t = getattr(self, '_TEST', None)
+ if t:
+ assert not hasattr(self, '_TESTS'), \
+ '%s has _TEST and _TESTS' % type(self).__name__
+ tests = [t]
+ else:
+ tests = getattr(self, '_TESTS', [])
+ for t in tests:
+ if not include_onlymatching and t.get('only_matching', False):
+ continue
+ t['name'] = type(self).__name__[:-len('IE')]
+ yield t
+
+ def is_suitable(self, age_limit):
+ """ Test whether the extractor is generally suitable for the given
+ age limit (i.e. pornographic sites are not, all others usually are) """
+
+ any_restricted = False
+ for tc in self.get_testcases(include_onlymatching=False):
+ if 'playlist' in tc:
+ tc = tc['playlist'][0]
+ is_restricted = age_restricted(
+ tc.get('info_dict', {}).get('age_limit'), age_limit)
+ if not is_restricted:
+ return True
+ any_restricted = any_restricted or is_restricted
+ return not any_restricted
+
class SearchInfoExtractor(InfoExtractor):
"""
def _real_extract(self, query):
mobj = re.match(self._make_valid_url(), query)
if mobj is None:
- raise ExtractorError(u'Invalid search query "%s"' % query)
+ raise ExtractorError('Invalid search query "%s"' % query)
prefix = mobj.group('prefix')
query = mobj.group('query')
else:
n = int(prefix)
if n <= 0:
- raise ExtractorError(u'invalid download number %s for query "%s"' % (n, query))
+ raise ExtractorError('invalid download number %s for query "%s"' % (n, query))
elif n > self._MAX_RESULTS:
- self._downloader.report_warning(u'%s returns max %i results (you requested %i)' % (self._SEARCH_KEY, self._MAX_RESULTS, n))
+ self._downloader.report_warning('%s returns max %i results (you requested %i)' % (self._SEARCH_KEY, self._MAX_RESULTS, n))
n = self._MAX_RESULTS
return self._get_n_results(query, n)