mirror of
https://codeberg.org/polarisfm/youtube-dl
synced 2024-11-05 01:44:33 +01:00
9103bbc5cd
The url for the video page, it must allow to reproduce the result. It's automatically set by YoutubeDL if it's missing.
392 lines
15 KiB
Python
392 lines
15 KiB
Python
import base64
|
|
import os
|
|
import re
|
|
import socket
|
|
import sys
|
|
import netrc
|
|
|
|
from ..utils import (
|
|
compat_http_client,
|
|
compat_urllib_error,
|
|
compat_urllib_request,
|
|
compat_str,
|
|
|
|
clean_html,
|
|
compiled_regex_type,
|
|
ExtractorError,
|
|
RegexNotFoundError,
|
|
sanitize_filename,
|
|
unescapeHTML,
|
|
)
|
|
|
|
class InfoExtractor(object):
|
|
"""Information Extractor class.
|
|
|
|
Information extractors are the classes that, given a URL, extract
|
|
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
|
|
information possibly downloading the video to the file system, among
|
|
other possible outcomes.
|
|
|
|
The dictionaries must include the following fields:
|
|
|
|
id: Video identifier.
|
|
url: Final video URL.
|
|
title: Video title, unescaped.
|
|
ext: Video filename extension.
|
|
|
|
Instead of url and ext, formats can also specified.
|
|
|
|
The following fields are optional:
|
|
|
|
format: The video format, defaults to ext (used for --get-format)
|
|
thumbnails: A list of dictionaries (with the entries "resolution" and
|
|
"url") for the varying thumbnails
|
|
thumbnail: Full URL to a video thumbnail image.
|
|
description: One-line video description.
|
|
uploader: Full name of the video uploader.
|
|
upload_date: Video upload date (YYYYMMDD).
|
|
uploader_id: Nickname or id of the video uploader.
|
|
location: Physical location of the video.
|
|
player_url: SWF Player URL (used for rtmpdump).
|
|
subtitles: The subtitle file contents as a dictionary in the format
|
|
{language: subtitles}.
|
|
view_count: How many users have watched the video on the platform.
|
|
urlhandle: [internal] The urlHandle to be used to download the file,
|
|
like returned by urllib.request.urlopen
|
|
age_limit: Age restriction for the video, as an integer (years)
|
|
formats: A list of dictionaries for each format available, it must
|
|
be ordered from worst to best quality. Potential fields:
|
|
* url Mandatory. The URL of the video file
|
|
* ext Will be calculated from url if missing
|
|
* format A human-readable description of the format
|
|
("mp4 container with h264/opus").
|
|
Calculated from the format_id, width, height.
|
|
and format_note fields if missing.
|
|
* format_id A short description of the format
|
|
("mp4_h264_opus" or "19")
|
|
* format_note Additional info about the format
|
|
("3D" or "DASH video")
|
|
* width Width of the video, if known
|
|
* height Height of the video, if known
|
|
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
|
|
by YoutubeDL if it's missing)
|
|
|
|
Unless mentioned otherwise, the fields should be Unicode strings.
|
|
|
|
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.
|
|
|
|
_real_extract() must return a *list* of information dictionaries as
|
|
described above.
|
|
|
|
Finally, the _WORKING attribute should be set to False for broken IEs
|
|
in order to warn the users and skip the tests.
|
|
"""
|
|
|
|
_ready = False
|
|
_downloader = None
|
|
_WORKING = True
|
|
|
|
def __init__(self, downloader=None):
|
|
"""Constructor. Receives an optional downloader."""
|
|
self._ready = False
|
|
self.set_downloader(downloader)
|
|
|
|
@classmethod
|
|
def suitable(cls, url):
|
|
"""Receives a URL and returns True if suitable for this IE."""
|
|
|
|
# This does not use has/getattr intentionally - we want to know whether
|
|
# we have cached the regexp for *this* class, whereas getattr would also
|
|
# match the superclass
|
|
if '_VALID_URL_RE' not in cls.__dict__:
|
|
cls._VALID_URL_RE = re.compile(cls._VALID_URL)
|
|
return cls._VALID_URL_RE.match(url) is not None
|
|
|
|
@classmethod
|
|
def working(cls):
|
|
"""Getter method for _WORKING."""
|
|
return cls._WORKING
|
|
|
|
def initialize(self):
|
|
"""Initializes an instance (authentication, etc)."""
|
|
if not self._ready:
|
|
self._real_initialize()
|
|
self._ready = True
|
|
|
|
def extract(self, url):
|
|
"""Extracts URL information and returns it in list of dicts."""
|
|
self.initialize()
|
|
return self._real_extract(url)
|
|
|
|
def set_downloader(self, downloader):
|
|
"""Sets the downloader for this IE."""
|
|
self._downloader = downloader
|
|
|
|
def _real_initialize(self):
|
|
"""Real initialization process. Redefine in subclasses."""
|
|
pass
|
|
|
|
def _real_extract(self, url):
|
|
"""Real extraction process. Redefine in subclasses."""
|
|
pass
|
|
|
|
@classmethod
|
|
def ie_key(cls):
|
|
"""A string for getting the InfoExtractor with get_info_extractor"""
|
|
return cls.__name__[:-2]
|
|
|
|
@property
|
|
def IE_NAME(self):
|
|
return type(self).__name__[:-2]
|
|
|
|
def _request_webpage(self, url_or_request, video_id, note=None, errnote=None):
|
|
""" Returns the response handle """
|
|
if note is None:
|
|
self.report_download_webpage(video_id)
|
|
elif note is not False:
|
|
self.to_screen(u'%s: %s' % (video_id, note))
|
|
try:
|
|
return compat_urllib_request.urlopen(url_or_request)
|
|
except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
|
|
if errnote is None:
|
|
errnote = u'Unable to download webpage'
|
|
raise ExtractorError(u'%s: %s' % (errnote, compat_str(err)), sys.exc_info()[2], cause=err)
|
|
|
|
def _download_webpage_handle(self, url_or_request, video_id, note=None, errnote=None):
|
|
""" 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]
|
|
|
|
urlh = self._request_webpage(url_or_request, video_id, note, errnote)
|
|
content_type = urlh.headers.get('Content-Type', '')
|
|
webpage_bytes = urlh.read()
|
|
m = re.match(r'[a-zA-Z0-9_.-]+/[a-zA-Z0-9_.-]+\s*;\s*charset=(.+)', content_type)
|
|
if m:
|
|
encoding = m.group(1)
|
|
else:
|
|
m = re.search(br'<meta[^>]+charset=[\'"]?([^\'")]+)[ /\'">]',
|
|
webpage_bytes[:1024])
|
|
if m:
|
|
encoding = m.group(1).decode('ascii')
|
|
else:
|
|
encoding = 'utf-8'
|
|
if self._downloader.params.get('dump_intermediate_pages', False):
|
|
try:
|
|
url = url_or_request.get_full_url()
|
|
except AttributeError:
|
|
url = url_or_request
|
|
self.to_screen(u'Dumping request to ' + url)
|
|
dump = base64.b64encode(webpage_bytes).decode('ascii')
|
|
self._downloader.to_screen(dump)
|
|
if self._downloader.params.get('write_pages', False):
|
|
try:
|
|
url = url_or_request.get_full_url()
|
|
except AttributeError:
|
|
url = url_or_request
|
|
raw_filename = ('%s_%s.dump' % (video_id, url))
|
|
filename = sanitize_filename(raw_filename, restricted=True)
|
|
self.to_screen(u'Saving request to ' + filename)
|
|
with open(filename, 'wb') as outf:
|
|
outf.write(webpage_bytes)
|
|
|
|
content = webpage_bytes.decode(encoding, 'replace')
|
|
return (content, urlh)
|
|
|
|
def _download_webpage(self, url_or_request, video_id, note=None, errnote=None):
|
|
""" Returns the data of the page as a string """
|
|
return self._download_webpage_handle(url_or_request, video_id, note, errnote)[0]
|
|
|
|
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))
|
|
|
|
def report_extraction(self, id_or_name):
|
|
"""Report information extraction."""
|
|
self.to_screen(u'%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)
|
|
|
|
def report_age_confirmation(self):
|
|
"""Report attempt to confirm age."""
|
|
self.to_screen(u'Confirming age')
|
|
|
|
def report_login(self):
|
|
"""Report attempt to log in."""
|
|
self.to_screen(u'Logging in')
|
|
|
|
#Methods for following #608
|
|
def url_result(self, url, ie=None):
|
|
"""Returns a url that points to a page that should be processed"""
|
|
#TODO: ie should be the class used for getting the info
|
|
video_info = {'_type': 'url',
|
|
'url': url,
|
|
'ie_key': ie}
|
|
return video_info
|
|
def playlist_result(self, entries, playlist_id=None, playlist_title=None):
|
|
"""Returns a playlist"""
|
|
video_info = {'_type': 'playlist',
|
|
'entries': entries}
|
|
if playlist_id:
|
|
video_info['id'] = playlist_id
|
|
if playlist_title:
|
|
video_info['title'] = playlist_title
|
|
return video_info
|
|
|
|
def _search_regex(self, pattern, string, name, default=None, fatal=True, flags=0):
|
|
"""
|
|
Perform a regex search on the given string, using a single or a list of
|
|
patterns returning the first matching group.
|
|
In case of failure return a default value or raise a WARNING or a
|
|
RegexNotFoundError, depending on fatal, specifying the field name.
|
|
"""
|
|
if isinstance(pattern, (str, compat_str, compiled_regex_type)):
|
|
mobj = re.search(pattern, string, flags)
|
|
else:
|
|
for p in pattern:
|
|
mobj = re.search(p, string, flags)
|
|
if mobj: break
|
|
|
|
if sys.stderr.isatty() and os.name != 'nt':
|
|
_name = u'\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)
|
|
elif default is not None:
|
|
return default
|
|
elif fatal:
|
|
raise RegexNotFoundError(u'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)
|
|
return None
|
|
|
|
def _html_search_regex(self, pattern, string, name, default=None, fatal=True, flags=0):
|
|
"""
|
|
Like _search_regex, but strips HTML tags and unescapes entities.
|
|
"""
|
|
res = self._search_regex(pattern, string, name, default, fatal, flags)
|
|
if res:
|
|
return clean_html(res).strip()
|
|
else:
|
|
return res
|
|
|
|
def _get_login_info(self):
|
|
"""
|
|
Get the the login info as (username, password)
|
|
It will look in the netrc file using the _NETRC_MACHINE value
|
|
If there's no info available, return (None, None)
|
|
"""
|
|
if self._downloader is None:
|
|
return (None, None)
|
|
|
|
username = None
|
|
password = None
|
|
downloader_params = self._downloader.params
|
|
|
|
# Attempt to use provided username and password or .netrc data
|
|
if downloader_params.get('username', None) is not None:
|
|
username = downloader_params['username']
|
|
password = downloader_params['password']
|
|
elif downloader_params.get('usenetrc', False):
|
|
try:
|
|
info = netrc.netrc().authenticators(self._NETRC_MACHINE)
|
|
if info is not None:
|
|
username = info[0]
|
|
password = info[2]
|
|
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))
|
|
|
|
return (username, password)
|
|
|
|
# Helper functions for extracting OpenGraph info
|
|
@staticmethod
|
|
def _og_regex(prop):
|
|
return r'<meta.+?property=[\'"]og:%s[\'"].+?content=(?:"(.+?)"|\'(.+?)\')' % re.escape(prop)
|
|
|
|
def _og_search_property(self, prop, html, name=None, **kargs):
|
|
if name is None:
|
|
name = 'OpenGraph %s' % prop
|
|
escaped = self._search_regex(self._og_regex(prop), html, name, flags=re.DOTALL, **kargs)
|
|
return unescapeHTML(escaped)
|
|
|
|
def _og_search_thumbnail(self, html, **kargs):
|
|
return self._og_search_property('image', html, u'thumbnail url', fatal=False, **kargs)
|
|
|
|
def _og_search_description(self, html, **kargs):
|
|
return self._og_search_property('description', html, fatal=False, **kargs)
|
|
|
|
def _og_search_title(self, html, **kargs):
|
|
return self._og_search_property('title', html, **kargs)
|
|
|
|
def _og_search_video_url(self, html, name='video url', secure=True, **kargs):
|
|
regexes = [self._og_regex('video')]
|
|
if secure: regexes.insert(0, self._og_regex('video:secure_url'))
|
|
return self._html_search_regex(regexes, html, name, **kargs)
|
|
|
|
def _rta_search(self, html):
|
|
# See http://www.rtalabel.org/index.php?content=howtofaq#single
|
|
if re.search(r'(?ix)<meta\s+name="rating"\s+'
|
|
r' content="RTA-5042-1996-1400-1577-RTA"',
|
|
html):
|
|
return 18
|
|
return 0
|
|
|
|
|
|
class SearchInfoExtractor(InfoExtractor):
|
|
"""
|
|
Base class for paged search queries extractors.
|
|
They accept urls in the format _SEARCH_KEY(|all|[0-9]):{query}
|
|
Instances should define _SEARCH_KEY and _MAX_RESULTS.
|
|
"""
|
|
|
|
@classmethod
|
|
def _make_valid_url(cls):
|
|
return r'%s(?P<prefix>|[1-9][0-9]*|all):(?P<query>[\s\S]+)' % cls._SEARCH_KEY
|
|
|
|
@classmethod
|
|
def suitable(cls, url):
|
|
return re.match(cls._make_valid_url(), url) is not None
|
|
|
|
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)
|
|
|
|
prefix = mobj.group('prefix')
|
|
query = mobj.group('query')
|
|
if prefix == '':
|
|
return self._get_n_results(query, 1)
|
|
elif prefix == 'all':
|
|
return self._get_n_results(query, self._MAX_RESULTS)
|
|
else:
|
|
n = int(prefix)
|
|
if n <= 0:
|
|
raise ExtractorError(u'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))
|
|
n = self._MAX_RESULTS
|
|
return self._get_n_results(query, n)
|
|
|
|
def _get_n_results(self, query, n):
|
|
"""Get a specified number of results for a query"""
|
|
raise NotImplementedError("This method must be implemented by subclasses")
|
|
|
|
@property
|
|
def SEARCH_KEY(self):
|
|
return self._SEARCH_KEY
|