You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

441 lines
17 KiB

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