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.

464 lines
18 KiB

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