youtube_dl/extractor/common.py

   1 import base64
   2 import os
   3 import re
   4 import socket
   5 import sys
   6 import netrc
   7 import xml.etree.ElementTree
   8
   9 from ..utils import (
  10     compat_http_client,
  11     compat_urllib_error,
  12     compat_str,
  13
  14     clean_html,
  15     compiled_regex_type,
  16     ExtractorError,
  17     RegexNotFoundError,
  18     sanitize_filename,
  19     unescapeHTML,
  20 )
  21
  22
  23 class InfoExtractor(object):
  24     """Information Extractor class.
  25
  26     Information extractors are the classes that, given a URL, extract
  27     information about the video (or videos) the URL refers to. This
  28     information includes the real video URL, the video title, author and
  29     others. The information is stored in a dictionary which is then
  30     passed to the FileDownloader. The FileDownloader processes this
  31     information possibly downloading the video to the file system, among
  32     other possible outcomes.
  33
  34     The dictionaries must include the following fields:
  35
  36     id:             Video identifier.
  37     url:            Final video URL.
  38     title:          Video title, unescaped.
  39     ext:            Video filename extension.
  40
  41     Instead of url and ext, formats can also specified.
  42
  43     The following fields are optional:
  44
  45     format:         The video format, defaults to ext (used for --get-format)
  46     thumbnails:     A list of dictionaries (with the entries "resolution" and
  47                     "url") for the varying thumbnails
  48     thumbnail:      Full URL to a video thumbnail image.
  49     description:    One-line video description.
  50     uploader:       Full name of the video uploader.
  51     upload_date:    Video upload date (YYYYMMDD).
  52     uploader_id:    Nickname or id of the video uploader.
  53     location:       Physical location of the video.
  54     player_url:     SWF Player URL (used for rtmpdump).
  55     subtitles:      The subtitle file contents as a dictionary in the format
  56                     {language: subtitles}.
  57     view_count:     How many users have watched the video on the platform.
  58     like_count:     Number of positive ratings of the video
  59     dislike_count:  Number of negative ratings of the video
  60     comment_count:  Number of comments on the video
  61     urlhandle:      [internal] The urlHandle to be used to download the file,
  62                     like returned by urllib.request.urlopen
  63     age_limit:      Age restriction for the video, as an integer (years)
  64     formats:        A list of dictionaries for each format available, it must
  65                     be ordered from worst to best quality. Potential fields:
  66                     * url       Mandatory. The URL of the video file
  67                     * ext       Will be calculated from url if missing
  68                     * format    A human-readable description of the format
  69                                 ("mp4 container with h264/opus").
  70                                 Calculated from the format_id, width, height.
  71                                 and format_note fields if missing.
  72                     * format_id A short description of the format
  73                                 ("mp4_h264_opus" or "19")
  74                     * format_note Additional info about the format
  75                                 ("3D" or "DASH video")
  76                     * width     Width of the video, if known
  77                     * height    Height of the video, if known
  78                     * abr       Average audio bitrate in KBit/s
  79                     * acodec    Name of the audio codec in use
  80                     * vbr       Average video bitrate in KBit/s
  81                     * vcodec    Name of the video codec in use
  82                     * filesize  The number of bytes, if known in advance
  83     webpage_url:    The url to the video webpage, if given to youtube-dl it
  84                     should allow to get the same result again. (It will be set
  85                     by YoutubeDL if it's missing)
  86
  87     Unless mentioned otherwise, the fields should be Unicode strings.
  88
  89     Subclasses of this one should re-define the _real_initialize() and
  90     _real_extract() methods and define a _VALID_URL regexp.
  91     Probably, they should also be added to the list of extractors.
  92
  93     _real_extract() must return a *list* of information dictionaries as
  94     described above.
  95
  96     Finally, the _WORKING attribute should be set to False for broken IEs
  97     in order to warn the users and skip the tests.
  98     """
  99
 100     _ready = False
 101     _downloader = None
 102     _WORKING = True
 103
 104     def __init__(self, downloader=None):
 105         """Constructor. Receives an optional downloader."""
 106         self._ready = False
 107         self.set_downloader(downloader)
 108
 109     @classmethod
 110     def suitable(cls, url):
 111         """Receives a URL and returns True if suitable for this IE."""
 112
 113         # This does not use has/getattr intentionally - we want to know whether
 114         # we have cached the regexp for *this* class, whereas getattr would also
 115         # match the superclass
 116         if '_VALID_URL_RE' not in cls.__dict__:
 117             cls._VALID_URL_RE = re.compile(cls._VALID_URL)
 118         return cls._VALID_URL_RE.match(url) is not None
 119
 120     @classmethod
 121     def working(cls):
 122         """Getter method for _WORKING."""
 123         return cls._WORKING
 124
 125     def initialize(self):
 126         """Initializes an instance (authentication, etc)."""
 127         if not self._ready:
 128             self._real_initialize()
 129             self._ready = True
 130
 131     def extract(self, url):
 132         """Extracts URL information and returns it in list of dicts."""
 133         self.initialize()
 134         return self._real_extract(url)
 135
 136     def set_downloader(self, downloader):
 137         """Sets the downloader for this IE."""
 138         self._downloader = downloader
 139
 140     def _real_initialize(self):
 141         """Real initialization process. Redefine in subclasses."""
 142         pass
 143
 144     def _real_extract(self, url):
 145         """Real extraction process. Redefine in subclasses."""
 146         pass
 147
 148     @classmethod
 149     def ie_key(cls):
 150         """A string for getting the InfoExtractor with get_info_extractor"""
 151         return cls.__name__[:-2]
 152
 153     @property
 154     def IE_NAME(self):
 155         return type(self).__name__[:-2]
 156
 157     def _request_webpage(self, url_or_request, video_id, note=None, errnote=None, fatal=True):
 158         """ Returns the response handle """
 159         if note is None:
 160             self.report_download_webpage(video_id)
 161         elif note is not False:
 162             if video_id is None:
 163                 self.to_screen(u'%s' % (note,))
 164             else:
 165                 self.to_screen(u'%s: %s' % (video_id, note))
 166         try:
 167             return self._downloader.urlopen(url_or_request)
 168         except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
 169             if errnote is None:
 170                 errnote = u'Unable to download webpage'
 171             errmsg = u'%s: %s' % (errnote, compat_str(err))
 172             if fatal:
 173                 raise ExtractorError(errmsg, sys.exc_info()[2], cause=err)
 174             else:
 175                 self._downloader.report_warning(errmsg)
 176                 return False
 177
 178     def _download_webpage_handle(self, url_or_request, video_id, note=None, errnote=None, fatal=True):
 179         """ Returns a tuple (page content as string, URL handle) """
 180
 181         # Strip hashes from the URL (#1038)
 182         if isinstance(url_or_request, (compat_str, str)):
 183             url_or_request = url_or_request.partition('#')[0]
 184
 185         urlh = self._request_webpage(url_or_request, video_id, note, errnote, fatal)
 186         if urlh is False:
 187             assert not fatal
 188             return False
 189         content_type = urlh.headers.get('Content-Type', '')
 190         webpage_bytes = urlh.read()
 191         m = re.match(r'[a-zA-Z0-9_.-]+/[a-zA-Z0-9_.-]+\s*;\s*charset=(.+)', content_type)
 192         if m:
 193             encoding = m.group(1)
 194         else:
 195             m = re.search(br'<meta[^>]+charset=[\'"]?([^\'")]+)[ /\'">]',
 196                           webpage_bytes[:1024])
 197             if m:
 198                 encoding = m.group(1).decode('ascii')
 199             else:
 200                 encoding = 'utf-8'
 201         if self._downloader.params.get('dump_intermediate_pages', False):
 202             try:
 203                 url = url_or_request.get_full_url()
 204             except AttributeError:
 205                 url = url_or_request
 206             self.to_screen(u'Dumping request to ' + url)
 207             dump = base64.b64encode(webpage_bytes).decode('ascii')
 208             self._downloader.to_screen(dump)
 209         if self._downloader.params.get('write_pages', False):
 210             try:
 211                 url = url_or_request.get_full_url()
 212             except AttributeError:
 213                 url = url_or_request
 214             raw_filename = ('%s_%s.dump' % (video_id, url))
 215             filename = sanitize_filename(raw_filename, restricted=True)
 216             self.to_screen(u'Saving request to ' + filename)
 217             with open(filename, 'wb') as outf:
 218                 outf.write(webpage_bytes)
 219
 220         content = webpage_bytes.decode(encoding, 'replace')
 221         return (content, urlh)
 222
 223     def _download_webpage(self, url_or_request, video_id, note=None, errnote=None, fatal=True):
 224         """ Returns the data of the page as a string """
 225         res = self._download_webpage_handle(url_or_request, video_id, note, errnote, fatal)
 226         if res is False:
 227             return res
 228         else:
 229             content, _ = res
 230             return content
 231
 232     def _download_xml(self, url_or_request, video_id,
 233                       note=u'Downloading XML', errnote=u'Unable to download XML'):
 234         """Return the xml as an xml.etree.ElementTree.Element"""
 235         xml_string = self._download_webpage(url_or_request, video_id, note, errnote)
 236         return xml.etree.ElementTree.fromstring(xml_string.encode('utf-8'))
 237
 238     def to_screen(self, msg):
 239         """Print msg to screen, prefixing it with '[ie_name]'"""
 240         self._downloader.to_screen(u'[%s] %s' % (self.IE_NAME, msg))
 241
 242     def report_extraction(self, id_or_name):
 243         """Report information extraction."""
 244         self.to_screen(u'%s: Extracting information' % id_or_name)
 245
 246     def report_download_webpage(self, video_id):
 247         """Report webpage download."""
 248         self.to_screen(u'%s: Downloading webpage' % video_id)
 249
 250     def report_age_confirmation(self):
 251         """Report attempt to confirm age."""
 252         self.to_screen(u'Confirming age')
 253
 254     def report_login(self):
 255         """Report attempt to log in."""
 256         self.to_screen(u'Logging in')
 257
 258     #Methods for following #608
 259     def url_result(self, url, ie=None, video_id=None):
 260         """Returns a url that points to a page that should be processed"""
 261         #TODO: ie should be the class used for getting the info
 262         video_info = {'_type': 'url',
 263                       'url': url,
 264                       'ie_key': ie}
 265         if video_id is not None:
 266             video_info['id'] = video_id
 267         return video_info
 268     def playlist_result(self, entries, playlist_id=None, playlist_title=None):
 269         """Returns a playlist"""
 270         video_info = {'_type': 'playlist',
 271                       'entries': entries}
 272         if playlist_id:
 273             video_info['id'] = playlist_id
 274         if playlist_title:
 275             video_info['title'] = playlist_title
 276         return video_info
 277
 278     def _search_regex(self, pattern, string, name, default=None, fatal=True, flags=0):
 279         """
 280         Perform a regex search on the given string, using a single or a list of
 281         patterns returning the first matching group.
 282         In case of failure return a default value or raise a WARNING or a
 283         RegexNotFoundError, depending on fatal, specifying the field name.
 284         """
 285         if isinstance(pattern, (str, compat_str, compiled_regex_type)):
 286             mobj = re.search(pattern, string, flags)
 287         else:
 288             for p in pattern:
 289                 mobj = re.search(p, string, flags)
 290                 if mobj: break
 291
 292         if sys.stderr.isatty() and os.name != 'nt':
 293             _name = u'\033[0;34m%s\033[0m' % name
 294         else:
 295             _name = name
 296
 297         if mobj:
 298             # return the first matching group
 299             return next(g for g in mobj.groups() if g is not None)
 300         elif default is not None:
 301             return default
 302         elif fatal:
 303             raise RegexNotFoundError(u'Unable to extract %s' % _name)
 304         else:
 305             self._downloader.report_warning(u'unable to extract %s; '
 306                 u'please report this issue on http://yt-dl.org/bug' % _name)
 307             return None
 308
 309     def _html_search_regex(self, pattern, string, name, default=None, fatal=True, flags=0):
 310         """
 311         Like _search_regex, but strips HTML tags and unescapes entities.
 312         """
 313         res = self._search_regex(pattern, string, name, default, fatal, flags)
 314         if res:
 315             return clean_html(res).strip()
 316         else:
 317             return res
 318
 319     def _get_login_info(self):
 320         """
 321         Get the the login info as (username, password)
 322         It will look in the netrc file using the _NETRC_MACHINE value
 323         If there's no info available, return (None, None)
 324         """
 325         if self._downloader is None:
 326             return (None, None)
 327
 328         username = None
 329         password = None
 330         downloader_params = self._downloader.params
 331
 332         # Attempt to use provided username and password or .netrc data
 333         if downloader_params.get('username', None) is not None:
 334             username = downloader_params['username']
 335             password = downloader_params['password']
 336         elif downloader_params.get('usenetrc', False):
 337             try:
 338                 info = netrc.netrc().authenticators(self._NETRC_MACHINE)
 339                 if info is not None:
 340                     username = info[0]
 341                     password = info[2]
 342                 else:
 343                     raise netrc.NetrcParseError('No authenticators for %s' % self._NETRC_MACHINE)
 344             except (IOError, netrc.NetrcParseError) as err:
 345                 self._downloader.report_warning(u'parsing .netrc: %s' % compat_str(err))
 346
 347         return (username, password)
 348
 349     # Helper functions for extracting OpenGraph info
 350     @staticmethod
 351     def _og_regexes(prop):
 352         content_re = r'content=(?:"([^>]+?)"|\'(.+?)\')'
 353         property_re = r'property=[\'"]og:%s[\'"]' % re.escape(prop)
 354         template = r'<meta[^>]+?%s[^>]+?%s'
 355         return [
 356             template % (property_re, content_re),
 357             template % (content_re, property_re),
 358         ]
 359
 360     def _og_search_property(self, prop, html, name=None, **kargs):
 361         if name is None:
 362             name = 'OpenGraph %s' % prop
 363         escaped = self._search_regex(self._og_regexes(prop), html, name, flags=re.DOTALL, **kargs)
 364         if escaped is None:
 365             return None
 366         return unescapeHTML(escaped)
 367
 368     def _og_search_thumbnail(self, html, **kargs):
 369         return self._og_search_property('image', html, u'thumbnail url', fatal=False, **kargs)
 370
 371     def _og_search_description(self, html, **kargs):
 372         return self._og_search_property('description', html, fatal=False, **kargs)
 373
 374     def _og_search_title(self, html, **kargs):
 375         return self._og_search_property('title', html, **kargs)
 376
 377     def _og_search_video_url(self, html, name='video url', secure=True, **kargs):
 378         regexes = self._og_regexes('video')
 379         if secure: regexes = self._og_regexes('video:secure_url') + regexes
 380         return self._html_search_regex(regexes, html, name, **kargs)
 381
 382     def _html_search_meta(self, name, html, display_name=None):
 383         if display_name is None:
 384             display_name = name
 385         return self._html_search_regex(
 386             r'''(?ix)<meta
 387                     (?=[^>]+(?:itemprop|name|property)=["\']%s["\'])
 388                     [^>]+content=["\']([^"\']+)["\']''' % re.escape(name),
 389             html, display_name, fatal=False)
 390
 391     def _dc_search_uploader(self, html):
 392         return self._html_search_meta('dc.creator', html, 'uploader')
 393
 394     def _rta_search(self, html):
 395         # See http://www.rtalabel.org/index.php?content=howtofaq#single
 396         if re.search(r'(?ix)<meta\s+name="rating"\s+'
 397                      r'     content="RTA-5042-1996-1400-1577-RTA"',
 398                      html):
 399             return 18
 400         return 0
 401
 402     def _media_rating_search(self, html):
 403         # See http://www.tjg-designs.com/WP/metadata-code-examples-adding-metadata-to-your-web-pages/
 404         rating = self._html_search_meta('rating', html)
 405
 406         if not rating:
 407             return None
 408
 409         RATING_TABLE = {
 410             'safe for kids': 0,
 411             'general': 8,
 412             '14 years': 14,
 413             'mature': 17,
 414             'restricted': 19,
 415         }
 416         return RATING_TABLE.get(rating.lower(), None)
 417
 418
 419
 420 class SearchInfoExtractor(InfoExtractor):
 421     """
 422     Base class for paged search queries extractors.
 423     They accept urls in the format _SEARCH_KEY(|all|[0-9]):{query}
 424     Instances should define _SEARCH_KEY and _MAX_RESULTS.
 425     """
 426
 427     @classmethod
 428     def _make_valid_url(cls):
 429         return r'%s(?P<prefix>|[1-9][0-9]*|all):(?P<query>[\s\S]+)' % cls._SEARCH_KEY
 430
 431     @classmethod
 432     def suitable(cls, url):
 433         return re.match(cls._make_valid_url(), url) is not None
 434
 435     def _real_extract(self, query):
 436         mobj = re.match(self._make_valid_url(), query)
 437         if mobj is None:
 438             raise ExtractorError(u'Invalid search query "%s"' % query)
 439
 440         prefix = mobj.group('prefix')
 441         query = mobj.group('query')
 442         if prefix == '':
 443             return self._get_n_results(query, 1)
 444         elif prefix == 'all':
 445             return self._get_n_results(query, self._MAX_RESULTS)
 446         else:
 447             n = int(prefix)
 448             if n <= 0:
 449                 raise ExtractorError(u'invalid download number %s for query "%s"' % (n, query))
 450             elif n > self._MAX_RESULTS:
 451                 self._downloader.report_warning(u'%s returns max %i results (you requested %i)' % (self._SEARCH_KEY, self._MAX_RESULTS, n))
 452                 n = self._MAX_RESULTS
 453             return self._get_n_results(query, n)
 454
 455     def _get_n_results(self, query, n):
 456         """Get a specified number of results for a query"""
 457         raise NotImplementedError("This method must be implemented by subclasses")
 458
 459     @property
 460     def SEARCH_KEY(self):
 461         return self._SEARCH_KEY