Raphaël G. Git Repositories - youtubedl/blob - youtube_dl/extractor/common.py

   1 import base64
   2 import hashlib
   3 import json
   4 import os
   5 import re
   6 import socket
   7 import sys
   8 import netrc
   9 import xml.etree.ElementTree
  10
  11 from ..utils import (
  12     compat_http_client,
  13     compat_urllib_error,
  14     compat_urllib_parse_urlparse,
  15     compat_str,
  16
  17     clean_html,
  18     compiled_regex_type,
  19     ExtractorError,
  20     RegexNotFoundError,
  21     sanitize_filename,
  22     unescapeHTML,
  23 )
  24 _NO_DEFAULT = object()
  25
  26
  27 class InfoExtractor(object):
  28     """Information Extractor class.
  29
  30     Information extractors are the classes that, given a URL, extract
  31     information about the video (or videos) the URL refers to. This
  32     information includes the real video URL, the video title, author and
  33     others. The information is stored in a dictionary which is then
  34     passed to the FileDownloader. The FileDownloader processes this
  35     information possibly downloading the video to the file system, among
  36     other possible outcomes.
  37
  38     The dictionaries must include the following fields:
  39
  40     id:             Video identifier.
  41     title:          Video title, unescaped.
  42
  43     Additionally, it must contain either a formats entry or a url one:
  44
  45     formats:        A list of dictionaries for each format available, ordered
  46                     from worst to best quality.
  47
  48                     Potential fields:
  49                     * url        Mandatory. The URL of the video file
  50                     * ext        Will be calculated from url if missing
  51                     * format     A human-readable description of the format
  52                                  ("mp4 container with h264/opus").
  53                                  Calculated from the format_id, width, height.
  54                                  and format_note fields if missing.
  55                     * format_id  A short description of the format
  56                                  ("mp4_h264_opus" or "19").
  57                                 Technically optional, but strongly recommended.
  58                     * format_note Additional info about the format
  59                                  ("3D" or "DASH video")
  60                     * width      Width of the video, if known
  61                     * height     Height of the video, if known
  62                     * resolution Textual description of width and height
  63                     * tbr        Average bitrate of audio and video in KBit/s
  64                     * abr        Average audio bitrate in KBit/s
  65                     * acodec     Name of the audio codec in use
  66                     * asr        Audio sampling rate in Hertz
  67                     * vbr        Average video bitrate in KBit/s
  68                     * vcodec     Name of the video codec in use
  69                     * container  Name of the container format
  70                     * filesize   The number of bytes, if known in advance
  71                     * player_url SWF Player URL (used for rtmpdump).
  72                     * protocol   The protocol that will be used for the actual
  73                                  download, lower-case.
  74                                  "http", "https", "rtsp", "rtmp", "m3u8" or so.
  75                     * preference Order number of this format. If this field is
  76                                  present and not None, the formats get sorted
  77                                  by this field.
  78                                  -1 for default (order by other properties),
  79                                  -2 or smaller for less than default.
  80                     * quality    Order number of the video quality of this
  81                                  format, irrespective of the file format.
  82                                  -1 for default (order by other properties),
  83                                  -2 or smaller for less than default.
  84     url:            Final video URL.
  85     ext:            Video filename extension.
  86     format:         The video format, defaults to ext (used for --get-format)
  87     player_url:     SWF Player URL (used for rtmpdump).
  88
  89     The following fields are optional:
  90
  91     thumbnails:     A list of dictionaries (with the entries "resolution" and
  92                     "url") for the varying thumbnails
  93     thumbnail:      Full URL to a video thumbnail image.
  94     description:    One-line video description.
  95     uploader:       Full name of the video uploader.
  96     upload_date:    Video upload date (YYYYMMDD).
  97     uploader_id:    Nickname or id of the video uploader.
  98     location:       Physical location of the video.
  99     subtitles:      The subtitle file contents as a dictionary in the format
 100                     {language: subtitles}.
 101     duration:       Length of the video in seconds, as an integer.
 102     view_count:     How many users have watched the video on the platform.
 103     like_count:     Number of positive ratings of the video
 104     dislike_count:  Number of negative ratings of the video
 105     comment_count:  Number of comments on the video
 106     age_limit:      Age restriction for the video, as an integer (years)
 107     webpage_url:    The url to the video webpage, if given to youtube-dl it
 108                     should allow to get the same result again. (It will be set
 109                     by YoutubeDL if it's missing)
 110
 111     Unless mentioned otherwise, the fields should be Unicode strings.
 112
 113     Subclasses of this one should re-define the _real_initialize() and
 114     _real_extract() methods and define a _VALID_URL regexp.
 115     Probably, they should also be added to the list of extractors.
 116
 117     _real_extract() must return a *list* of information dictionaries as
 118     described above.
 119
 120     Finally, the _WORKING attribute should be set to False for broken IEs
 121     in order to warn the users and skip the tests.
 122     """
 123
 124     _ready = False
 125     _downloader = None
 126     _WORKING = True
 127
 128     def __init__(self, downloader=None):
 129         """Constructor. Receives an optional downloader."""
 130         self._ready = False
 131         self.set_downloader(downloader)
 132
 133     @classmethod
 134     def suitable(cls, url):
 135         """Receives a URL and returns True if suitable for this IE."""
 136
 137         # This does not use has/getattr intentionally - we want to know whether
 138         # we have cached the regexp for *this* class, whereas getattr would also
 139         # match the superclass
 140         if '_VALID_URL_RE' not in cls.__dict__:
 141             cls._VALID_URL_RE = re.compile(cls._VALID_URL)
 142         return cls._VALID_URL_RE.match(url) is not None
 143
 144     @classmethod
 145     def working(cls):
 146         """Getter method for _WORKING."""
 147         return cls._WORKING
 148
 149     def initialize(self):
 150         """Initializes an instance (authentication, etc)."""
 151         if not self._ready:
 152             self._real_initialize()
 153             self._ready = True
 154
 155     def extract(self, url):
 156         """Extracts URL information and returns it in list of dicts."""
 157         self.initialize()
 158         return self._real_extract(url)
 159
 160     def set_downloader(self, downloader):
 161         """Sets the downloader for this IE."""
 162         self._downloader = downloader
 163
 164     def _real_initialize(self):
 165         """Real initialization process. Redefine in subclasses."""
 166         pass
 167
 168     def _real_extract(self, url):
 169         """Real extraction process. Redefine in subclasses."""
 170         pass
 171
 172     @classmethod
 173     def ie_key(cls):
 174         """A string for getting the InfoExtractor with get_info_extractor"""
 175         return cls.__name__[:-2]
 176
 177     @property
 178     def IE_NAME(self):
 179         return type(self).__name__[:-2]
 180
 181     def _request_webpage(self, url_or_request, video_id, note=None, errnote=None, fatal=True):
 182         """ Returns the response handle """
 183         if note is None:
 184             self.report_download_webpage(video_id)
 185         elif note is not False:
 186             if video_id is None:
 187                 self.to_screen(u'%s' % (note,))
 188             else:
 189                 self.to_screen(u'%s: %s' % (video_id, note))
 190         try:
 191             return self._downloader.urlopen(url_or_request)
 192         except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
 193             if errnote is False:
 194                 return False
 195             if errnote is None:
 196                 errnote = u'Unable to download webpage'
 197             errmsg = u'%s: %s' % (errnote, compat_str(err))
 198             if fatal:
 199                 raise ExtractorError(errmsg, sys.exc_info()[2], cause=err)
 200             else:
 201                 self._downloader.report_warning(errmsg)
 202                 return False
 203
 204     def _download_webpage_handle(self, url_or_request, video_id, note=None, errnote=None, fatal=True):
 205         """ Returns a tuple (page content as string, URL handle) """
 206
 207         # Strip hashes from the URL (#1038)
 208         if isinstance(url_or_request, (compat_str, str)):
 209             url_or_request = url_or_request.partition('#')[0]
 210
 211         urlh = self._request_webpage(url_or_request, video_id, note, errnote, fatal)
 212         if urlh is False:
 213             assert not fatal
 214             return False
 215         content_type = urlh.headers.get('Content-Type', '')
 216         webpage_bytes = urlh.read()
 217         m = re.match(r'[a-zA-Z0-9_.-]+/[a-zA-Z0-9_.-]+\s*;\s*charset=(.+)', content_type)
 218         if m:
 219             encoding = m.group(1)
 220         else:
 221             m = re.search(br'<meta[^>]+charset=[\'"]?([^\'")]+)[ /\'">]',
 222                           webpage_bytes[:1024])
 223             if m:
 224                 encoding = m.group(1).decode('ascii')
 225             elif webpage_bytes.startswith(b'\xff\xfe'):
 226                 encoding = 'utf-16'
 227             else:
 228                 encoding = 'utf-8'
 229         if self._downloader.params.get('dump_intermediate_pages', False):
 230             try:
 231                 url = url_or_request.get_full_url()
 232             except AttributeError:
 233                 url = url_or_request
 234             self.to_screen(u'Dumping request to ' + url)
 235             dump = base64.b64encode(webpage_bytes).decode('ascii')
 236             self._downloader.to_screen(dump)
 237         if self._downloader.params.get('write_pages', False):
 238             try:
 239                 url = url_or_request.get_full_url()
 240             except AttributeError:
 241                 url = url_or_request
 242             if len(url) > 200:
 243                 h = u'___' + hashlib.md5(url.encode('utf-8')).hexdigest()
 244                 url = url[:200 - len(h)] + h
 245             raw_filename = ('%s_%s.dump' % (video_id, url))
 246             filename = sanitize_filename(raw_filename, restricted=True)
 247             self.to_screen(u'Saving request to ' + filename)
 248             with open(filename, 'wb') as outf:
 249                 outf.write(webpage_bytes)
 250
 251         content = webpage_bytes.decode(encoding, 'replace')
 252         return (content, urlh)
 253
 254     def _download_webpage(self, url_or_request, video_id, note=None, errnote=None, fatal=True):
 255         """ Returns the data of the page as a string """
 256         res = self._download_webpage_handle(url_or_request, video_id, note, errnote, fatal)
 257         if res is False:
 258             return res
 259         else:
 260             content, _ = res
 261             return content
 262
 263     def _download_xml(self, url_or_request, video_id,
 264                       note=u'Downloading XML', errnote=u'Unable to download XML',
 265                       transform_source=None):
 266         """Return the xml as an xml.etree.ElementTree.Element"""
 267         xml_string = self._download_webpage(url_or_request, video_id, note, errnote)
 268         if transform_source:
 269             xml_string = transform_source(xml_string)
 270         return xml.etree.ElementTree.fromstring(xml_string.encode('utf-8'))
 271
 272     def _download_json(self, url_or_request, video_id,
 273                        note=u'Downloading JSON metadata',
 274                        errnote=u'Unable to download JSON metadata',
 275                        transform_source=None):
 276         json_string = self._download_webpage(url_or_request, video_id, note, errnote)
 277         if transform_source:
 278             json_string = transform_source(json_string)
 279         try:
 280             return json.loads(json_string)
 281         except ValueError as ve:
 282             raise ExtractorError('Failed to download JSON', cause=ve)
 283
 284     def report_warning(self, msg, video_id=None):
 285         idstr = u'' if video_id is None else u'%s: ' % video_id
 286         self._downloader.report_warning(
 287             u'[%s] %s%s' % (self.IE_NAME, idstr, msg))
 288
 289     def to_screen(self, msg):
 290         """Print msg to screen, prefixing it with '[ie_name]'"""
 291         self._downloader.to_screen(u'[%s] %s' % (self.IE_NAME, msg))
 292
 293     def report_extraction(self, id_or_name):
 294         """Report information extraction."""
 295         self.to_screen(u'%s: Extracting information' % id_or_name)
 296
 297     def report_download_webpage(self, video_id):
 298         """Report webpage download."""
 299         self.to_screen(u'%s: Downloading webpage' % video_id)
 300
 301     def report_age_confirmation(self):
 302         """Report attempt to confirm age."""
 303         self.to_screen(u'Confirming age')
 304
 305     def report_login(self):
 306         """Report attempt to log in."""
 307         self.to_screen(u'Logging in')
 308
 309     #Methods for following #608
 310     @staticmethod
 311     def url_result(url, ie=None, video_id=None):
 312         """Returns a url that points to a page that should be processed"""
 313         #TODO: ie should be the class used for getting the info
 314         video_info = {'_type': 'url',
 315                       'url': url,
 316                       'ie_key': ie}
 317         if video_id is not None:
 318             video_info['id'] = video_id
 319         return video_info
 320     @staticmethod
 321     def playlist_result(entries, playlist_id=None, playlist_title=None):
 322         """Returns a playlist"""
 323         video_info = {'_type': 'playlist',
 324                       'entries': entries}
 325         if playlist_id:
 326             video_info['id'] = playlist_id
 327         if playlist_title:
 328             video_info['title'] = playlist_title
 329         return video_info
 330
 331     def _search_regex(self, pattern, string, name, default=_NO_DEFAULT, fatal=True, flags=0):
 332         """
 333         Perform a regex search on the given string, using a single or a list of
 334         patterns returning the first matching group.
 335         In case of failure return a default value or raise a WARNING or a
 336         RegexNotFoundError, depending on fatal, specifying the field name.
 337         """
 338         if isinstance(pattern, (str, compat_str, compiled_regex_type)):
 339             mobj = re.search(pattern, string, flags)
 340         else:
 341             for p in pattern:
 342                 mobj = re.search(p, string, flags)
 343                 if mobj: break
 344
 345         if os.name != 'nt' and sys.stderr.isatty():
 346             _name = u'\033[0;34m%s\033[0m' % name
 347         else:
 348             _name = name
 349
 350         if mobj:
 351             # return the first matching group
 352             return next(g for g in mobj.groups() if g is not None)
 353         elif default is not _NO_DEFAULT:
 354             return default
 355         elif fatal:
 356             raise RegexNotFoundError(u'Unable to extract %s' % _name)
 357         else:
 358             self._downloader.report_warning(u'unable to extract %s; '
 359                 u'please report this issue on http://yt-dl.org/bug' % _name)
 360             return None
 361
 362     def _html_search_regex(self, pattern, string, name, default=_NO_DEFAULT, fatal=True, flags=0):
 363         """
 364         Like _search_regex, but strips HTML tags and unescapes entities.
 365         """
 366         res = self._search_regex(pattern, string, name, default, fatal, flags)
 367         if res:
 368             return clean_html(res).strip()
 369         else:
 370             return res
 371
 372     def _get_login_info(self):
 373         """
 374         Get the the login info as (username, password)
 375         It will look in the netrc file using the _NETRC_MACHINE value
 376         If there's no info available, return (None, None)
 377         """
 378         if self._downloader is None:
 379             return (None, None)
 380
 381         username = None
 382         password = None
 383         downloader_params = self._downloader.params
 384
 385         # Attempt to use provided username and password or .netrc data
 386         if downloader_params.get('username', None) is not None:
 387             username = downloader_params['username']
 388             password = downloader_params['password']
 389         elif downloader_params.get('usenetrc', False):
 390             try:
 391                 info = netrc.netrc().authenticators(self._NETRC_MACHINE)
 392                 if info is not None:
 393                     username = info[0]
 394                     password = info[2]
 395                 else:
 396                     raise netrc.NetrcParseError('No authenticators for %s' % self._NETRC_MACHINE)
 397             except (IOError, netrc.NetrcParseError) as err:
 398                 self._downloader.report_warning(u'parsing .netrc: %s' % compat_str(err))
 399
 400         return (username, password)
 401
 402     # Helper functions for extracting OpenGraph info
 403     @staticmethod
 404     def _og_regexes(prop):
 405         content_re = r'content=(?:"([^>]+?)"|\'([^>]+?)\')'
 406         property_re = r'(?:name|property)=[\'"]og:%s[\'"]' % re.escape(prop)
 407         template = r'<meta[^>]+?%s[^>]+?%s'
 408         return [
 409             template % (property_re, content_re),
 410             template % (content_re, property_re),
 411         ]
 412
 413     def _og_search_property(self, prop, html, name=None, **kargs):
 414         if name is None:
 415             name = 'OpenGraph %s' % prop
 416         escaped = self._search_regex(self._og_regexes(prop), html, name, flags=re.DOTALL, **kargs)
 417         if escaped is None:
 418             return None
 419         return unescapeHTML(escaped)
 420
 421     def _og_search_thumbnail(self, html, **kargs):
 422         return self._og_search_property('image', html, u'thumbnail url', fatal=False, **kargs)
 423
 424     def _og_search_description(self, html, **kargs):
 425         return self._og_search_property('description', html, fatal=False, **kargs)
 426
 427     def _og_search_title(self, html, **kargs):
 428         return self._og_search_property('title', html, **kargs)
 429
 430     def _og_search_video_url(self, html, name='video url', secure=True, **kargs):
 431         regexes = self._og_regexes('video')
 432         if secure: regexes = self._og_regexes('video:secure_url') + regexes
 433         return self._html_search_regex(regexes, html, name, **kargs)
 434
 435     def _html_search_meta(self, name, html, display_name=None):
 436         if display_name is None:
 437             display_name = name
 438         return self._html_search_regex(
 439             r'''(?ix)<meta
 440                     (?=[^>]+(?:itemprop|name|property)=["\']%s["\'])
 441                     [^>]+content=["\']([^"\']+)["\']''' % re.escape(name),
 442             html, display_name, fatal=False)
 443
 444     def _dc_search_uploader(self, html):
 445         return self._html_search_meta('dc.creator', html, 'uploader')
 446
 447     def _rta_search(self, html):
 448         # See http://www.rtalabel.org/index.php?content=howtofaq#single
 449         if re.search(r'(?ix)<meta\s+name="rating"\s+'
 450                      r'     content="RTA-5042-1996-1400-1577-RTA"',
 451                      html):
 452             return 18
 453         return 0
 454
 455     def _media_rating_search(self, html):
 456         # See http://www.tjg-designs.com/WP/metadata-code-examples-adding-metadata-to-your-web-pages/
 457         rating = self._html_search_meta('rating', html)
 458
 459         if not rating:
 460             return None
 461
 462         RATING_TABLE = {
 463             'safe for kids': 0,
 464             'general': 8,
 465             '14 years': 14,
 466             'mature': 17,
 467             'restricted': 19,
 468         }
 469         return RATING_TABLE.get(rating.lower(), None)
 470
 471     def _twitter_search_player(self, html):
 472         return self._html_search_meta('twitter:player', html,
 473             'twitter card player')
 474
 475     def _sort_formats(self, formats):
 476         if not formats:
 477             raise ExtractorError(u'No video formats found')
 478
 479         def _formats_key(f):
 480             # TODO remove the following workaround
 481             from ..utils import determine_ext
 482             if not f.get('ext') and 'url' in f:
 483                 f['ext'] = determine_ext(f['url'])
 484
 485             preference = f.get('preference')
 486             if preference is None:
 487                 proto = f.get('protocol')
 488                 if proto is None:
 489                     proto = compat_urllib_parse_urlparse(f.get('url', '')).scheme
 490
 491                 preference = 0 if proto in ['http', 'https'] else -0.1
 492                 if f.get('ext') in ['f4f', 'f4m']:  # Not yet supported
 493                     preference -= 0.5
 494
 495             if f.get('vcodec') == 'none':  # audio only
 496                 if self._downloader.params.get('prefer_free_formats'):
 497                     ORDER = [u'aac', u'mp3', u'm4a', u'webm', u'ogg', u'opus']
 498                 else:
 499                     ORDER = [u'webm', u'opus', u'ogg', u'mp3', u'aac', u'm4a']
 500                 ext_preference = 0
 501                 try:
 502                     audio_ext_preference = ORDER.index(f['ext'])
 503                 except ValueError:
 504                     audio_ext_preference = -1
 505             else:
 506                 if self._downloader.params.get('prefer_free_formats'):
 507                     ORDER = [u'flv', u'mp4', u'webm']
 508                 else:
 509                     ORDER = [u'webm', u'flv', u'mp4']
 510                 try:
 511                     ext_preference = ORDER.index(f['ext'])
 512                 except ValueError:
 513                     ext_preference = -1
 514                 audio_ext_preference = 0
 515
 516             return (
 517                 preference,
 518                 f.get('quality') if f.get('quality') is not None else -1,
 519                 f.get('height') if f.get('height') is not None else -1,
 520                 f.get('width') if f.get('width') is not None else -1,
 521                 ext_preference,
 522                 f.get('tbr') if f.get('tbr') is not None else -1,
 523                 f.get('vbr') if f.get('vbr') is not None else -1,
 524                 f.get('abr') if f.get('abr') is not None else -1,
 525                 audio_ext_preference,
 526                 f.get('filesize') if f.get('filesize') is not None else -1,
 527                 f.get('format_id'),
 528             )
 529         formats.sort(key=_formats_key)
 530
 531
 532 class SearchInfoExtractor(InfoExtractor):
 533     """
 534     Base class for paged search queries extractors.
 535     They accept urls in the format _SEARCH_KEY(|all|[0-9]):{query}
 536     Instances should define _SEARCH_KEY and _MAX_RESULTS.
 537     """
 538
 539     @classmethod
 540     def _make_valid_url(cls):
 541         return r'%s(?P<prefix>|[1-9][0-9]*|all):(?P<query>[\s\S]+)' % cls._SEARCH_KEY
 542
 543     @classmethod
 544     def suitable(cls, url):
 545         return re.match(cls._make_valid_url(), url) is not None
 546
 547     def _real_extract(self, query):
 548         mobj = re.match(self._make_valid_url(), query)
 549         if mobj is None:
 550             raise ExtractorError(u'Invalid search query "%s"' % query)
 551
 552         prefix = mobj.group('prefix')
 553         query = mobj.group('query')
 554         if prefix == '':
 555             return self._get_n_results(query, 1)
 556         elif prefix == 'all':
 557             return self._get_n_results(query, self._MAX_RESULTS)
 558         else:
 559             n = int(prefix)
 560             if n <= 0:
 561                 raise ExtractorError(u'invalid download number %s for query "%s"' % (n, query))
 562             elif n > self._MAX_RESULTS:
 563                 self._downloader.report_warning(u'%s returns max %i results (you requested %i)' % (self._SEARCH_KEY, self._MAX_RESULTS, n))
 564                 n = self._MAX_RESULTS
 565             return self._get_n_results(query, n)
 566
 567     def _get_n_results(self, query, n):
 568         """Get a specified number of results for a query"""
 569         raise NotImplementedError("This method must be implemented by subclasses")
 570
 571     @property
 572     def SEARCH_KEY(self):
 573         return self._SEARCH_KEY