X-Git-Url: https://git.rapsys.eu/youtubedl/blobdiff_plain/46113edab215c2211a604c06245c16d5d4e57dcf..620b803afb94df95e3761acf1e764a9eb3a4b00c:/youtube_dl/extractor/common.py
diff --git a/youtube_dl/extractor/common.py b/youtube_dl/extractor/common.py
index 7e41132..df1a441 100644
--- a/youtube_dl/extractor/common.py
+++ b/youtube_dl/extractor/common.py
@@ -12,17 +12,22 @@ import sys
import time
import xml.etree.ElementTree
-from ..utils import (
+from ..compat import (
+ compat_cookiejar,
+ compat_HTTPError,
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,
+ HEADRequest,
int_or_none,
RegexNotFoundError,
sanitize_filename,
@@ -38,11 +43,15 @@ class InfoExtractor(object):
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.
@@ -80,12 +89,20 @@ class InfoExtractor(object):
* player_url SWF Player URL (used for rtmpdump).
* protocol The protocol that will be used for the actual
download, lower-case.
- "http", "https", "rtsp", "rtmp", "m3u8" or so.
+ "http", "https", "rtsp", "rtmp", "rtmpe",
+ "m3u8", or "m3u8_native".
* preference Order number of this format. If this field is
present and not None, the formats get sorted
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),
@@ -94,12 +111,17 @@ class InfoExtractor(object):
(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
to add to the request.
* http_post_data Additional data to send with a POST
request.
+ * stretched_ratio If given and not 1, indicates that the
+ video's pixels are not square.
+ width : height ratio as float.
+ * no_resume The server does not support resuming the
+ (HTTP or RTMP) download. Boolean.
+
url: Final video URL.
ext: Video filename extension.
format: The video format, defaults to ext (used for --get-format)
@@ -107,19 +129,23 @@ class InfoExtractor(object):
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",
and display_id "dancing-naked-mole-rats"
thumbnails: A list of dictionaries, with the following entries:
+ * "id" (optional, string) - Thumbnail format ID
* "url"
+ * "preference" (optional, int) - quality of the image
* "width" (optional, int)
* "height" (optional, int)
* "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.
+ creator: The main artist who created the video.
timestamp: UNIX timestamp of the moment the video became available.
upload_date: Video upload date (YYYYMMDD).
If not explicitly set, calculated from timestamp.
@@ -132,6 +158,17 @@ class InfoExtractor(object):
like_count: Number of positive ratings of the video
dislike_count: Number of negative ratings of the video
comment_count: Number of comments on the video
+ comments: A list of comments, each with one or more of the following
+ properties (all but one of text or html optional):
+ * "author" - human-readable name of the comment author
+ * "author_id" - user ID of the comment author
+ * "id" - Comment ID
+ * "html" - Comment as HTML
+ * "text" - Plain text of the comment
+ * "timestamp" - UNIX timestamp of comment
+ * "parent" - ID of the comment this one is replying to.
+ Set to "root" to indicate that this is a
+ comment to the original video.
age_limit: Age restriction for the video, as an integer (years)
webpage_url: The url to the video webpage, if given to youtube-dl it
should allow to get the same result again. (It will be set
@@ -145,6 +182,39 @@ class InfoExtractor(object):
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.
@@ -254,9 +324,11 @@ class InfoExtractor(object):
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):
+ 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)
@@ -315,9 +387,19 @@ class InfoExtractor(object):
return content
- def _download_webpage(self, url_or_request, video_id, note=None, errnote=None, fatal=True):
+ def _download_webpage(self, url_or_request, video_id, note=None, errnote=None, fatal=True, tries=1, timeout=5):
""" Returns the data of the page as a string """
- res = self._download_webpage_handle(url_or_request, video_id, note, errnote, fatal)
+ success = False
+ try_count = 0
+ while success is False:
+ try:
+ res = self._download_webpage_handle(url_or_request, video_id, note, errnote, fatal)
+ success = True
+ except compat_http_client.IncompleteRead as e:
+ try_count += 1
+ if try_count >= tries:
+ raise e
+ self._sleep(timeout, video_id)
if res is False:
return res
else:
@@ -345,6 +427,10 @@ class InfoExtractor(object):
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:
@@ -381,19 +467,20 @@ class InfoExtractor(object):
"""Report attempt to log 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}
@@ -401,9 +488,11 @@ class InfoExtractor(object):
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.
@@ -424,22 +513,25 @@ class InfoExtractor(object):
_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('Unable to extract %s' % _name)
else:
self._downloader.report_warning('unable to extract %s; '
- 'please report this issue on http://yt-dl.org/bug' % _name)
+ '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:
@@ -472,7 +564,7 @@ class InfoExtractor(object):
raise netrc.NetrcParseError('No authenticators for %s' % self._NETRC_MACHINE)
except (IOError, netrc.NetrcParseError) as err:
self._downloader.report_warning('parsing .netrc: %s' % compat_str(err))
-
+
return (username, password)
def _get_tfa_info(self):
@@ -532,10 +624,10 @@ class InfoExtractor(object):
if display_name is None:
display_name = name
return self._html_search_regex(
- r'''(?ix)]+(?:itemprop|name|property)=["\']?%s["\']?)
- [^>]+content=["\']([^"\']+)["\']''' % re.escape(name),
- html, display_name, fatal=fatal, **kwargs)
+ r'''(?isx)]+(?:itemprop|name|property)=(["\']?)%s\1)
+ [^>]+?content=(["\'])(?P.*?)\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')
@@ -566,7 +658,7 @@ class InfoExtractor(object):
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:
@@ -611,12 +703,13 @@ class InfoExtractor(object):
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,
- ext_preference,
f.get('tbr') if f.get('tbr') is not None else -1,
f.get('vbr') if f.get('vbr') is not None else -1,
+ ext_preference,
+ 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('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,
@@ -627,6 +720,27 @@ class InfoExtractor(object):
)
formats.sort(key=_formats_key)
+ def _check_formats(self, formats, video_id):
+ if formats:
+ formats[:] = filter(
+ lambda f: self._is_valid_url(
+ f['url'], video_id,
+ item='%s video format' % f.get('format_id') if f.get('format_id') else 'video'),
+ formats)
+
+ def _is_valid_url(self, url, video_id, item='video'):
+ try:
+ self._request_webpage(
+ HEADRequest(url), video_id,
+ 'Checking %s URL' % item)
+ return True
+ except ExtractorError as e:
+ if isinstance(e.cause, compat_HTTPError):
+ self.report_warning(
+ '%s URL is invalid, skipping' % item, video_id)
+ return False
+ raise
+
def http_scheme(self):
""" Either "http:" or "https:", depending on the user's preferences """
return (
@@ -651,33 +765,41 @@ class InfoExtractor(object):
self.to_screen(msg)
time.sleep(timeout)
- def _extract_f4m_formats(self, manifest_url, video_id):
+ def _extract_f4m_formats(self, manifest_url, video_id, preference=None, f4m_id=None):
manifest = self._download_xml(
manifest_url, video_id, 'Downloading f4m manifest',
'Unable to download f4m manifest')
formats = []
+ manifest_version = '1.0'
media_nodes = manifest.findall('{http://ns.adobe.com/f4m/1.0}media')
+ if not media_nodes:
+ manifest_version = '2.0'
+ media_nodes = manifest.findall('{http://ns.adobe.com/f4m/2.0}media')
for i, media_el in enumerate(media_nodes):
+ if manifest_version == '2.0':
+ manifest_url = ('/'.join(manifest_url.split('/')[:-1]) + '/'
+ + (media_el.attrib.get('href') or media_el.attrib.get('url')))
tbr = int_or_none(media_el.attrib.get('bitrate'))
- format_id = 'f4m-%d' % (i if tbr is None else tbr)
formats.append({
- 'format_id': format_id,
+ 'format_id': '-'.join(filter(None, [f4m_id, 'f4m-%d' % (i if tbr is None else tbr)])),
'url': manifest_url,
'ext': 'flv',
'tbr': tbr,
'width': int_or_none(media_el.attrib.get('width')),
'height': int_or_none(media_el.attrib.get('height')),
+ 'preference': preference,
})
self._sort_formats(formats)
return formats
def _extract_m3u8_formats(self, m3u8_url, video_id, ext=None,
- entry_protocol='m3u8', preference=None):
+ entry_protocol='m3u8', preference=None,
+ m3u8_id=None):
formats = [{
- 'format_id': 'm3u8-meta',
+ 'format_id': '-'.join(filter(None, [m3u8_id, 'm3u8-meta'])),
'url': m3u8_url,
'ext': ext,
'protocol': 'm3u8',
@@ -713,9 +835,8 @@ class InfoExtractor(object):
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)),
+ 'format_id': '-'.join(filter(None, [m3u8_id, 'm3u8-%d' % (tbr if tbr else len(formats))])),
'url': format_url(line.strip()),
'tbr': tbr,
'ext': ext,
@@ -740,6 +861,52 @@ class InfoExtractor(object):
self._sort_formats(formats)
return formats
+ # TODO: improve extraction
+ def _extract_smil_formats(self, smil_url, video_id, fatal=True):
+ smil = self._download_xml(
+ smil_url, video_id, 'Downloading SMIL file',
+ 'Unable to download SMIL file', fatal=fatal)
+ if smil is False:
+ assert not fatal
+ return []
+
+ 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()
@@ -768,6 +935,41 @@ class InfoExtractor(object):
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):
"""