X-Git-Url: https://git.rapsys.eu/youtubedl/blobdiff_plain/33cd347759d6d999325ebf3c69b7ed5692c343b2..415fdb62500dca2e22067a05008dfbf87c75b662:/youtube_dl/extractor/common.py?ds=inline

diff --git a/youtube_dl/extractor/common.py b/youtube_dl/extractor/common.py
index da50abf..49e7540 100644
--- a/youtube_dl/extractor/common.py
+++ b/youtube_dl/extractor/common.py
@@ -1,21 +1,28 @@
 import base64
+import hashlib
+import json
 import os
 import re
 import socket
 import sys
 import netrc
+import xml.etree.ElementTree
 
 from ..utils import (
     compat_http_client,
     compat_urllib_error,
-    compat_urllib_request,
+    compat_urllib_parse_urlparse,
     compat_str,
 
     clean_html,
     compiled_regex_type,
     ExtractorError,
+    RegexNotFoundError,
+    sanitize_filename,
     unescapeHTML,
 )
+_NO_DEFAULT = object()
+
 
 class InfoExtractor(object):
     """Information Extractor class.
@@ -31,36 +38,94 @@ class InfoExtractor(object):
     The dictionaries must include the following fields:
 
     id:             Video identifier.
-    url:            Final video URL.
     title:          Video title, unescaped.
+
+    Additionally, it must contain either a formats entry or a url one:
+
+    formats:        A list of dictionaries for each format available, 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").
+                                Technically optional, but strongly recommended.
+                    * 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
+                    * resolution Textual description of width and height
+                    * tbr        Average bitrate of audio and video in KBit/s
+                    * abr        Average audio bitrate in KBit/s
+                    * acodec     Name of the audio codec in use
+                    * asr        Audio sampling rate in Hertz
+                    * vbr        Average video bitrate in KBit/s
+                    * vcodec     Name of the video codec in use
+                    * container  Name of the container format
+                    * filesize   The number of bytes, if known in advance
+                    * player_url SWF Player URL (used for rtmpdump).
+                    * protocol   The protocol that will be used for the actual
+                                 download, lower-case.
+                                 "http", "https", "rtsp", "rtmp", "m3u8" or so.
+                    * preference Order number of this format. If this field is
+                                 present and not None, the formats get sorted
+                                 by this field, regardless of all other values.
+                                 -1 for default (order by other properties),
+                                 -2 or smaller for less than default.
+                    * quality    Order number of the video quality of this
+                                 format, irrespective of the file format.
+                                 -1 for default (order by other properties),
+                                 -2 or smaller for less than default.
+    url:            Final video URL.
     ext:            Video filename extension.
+    format:         The video format, defaults to ext (used for --get-format)
+    player_url:     SWF Player URL (used for rtmpdump).
 
     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
+    display_id      An alternative identifier for the video, not necessarily
+                    unique, but available before title. Typically, id is
+                    something like "4234987", title "Dancing naked mole rats",
+                    and display_id "dancing-naked-mole-rats"
+    thumbnails:     A list of dictionaries, with the following entries:
+                        * "url"
+                        * "width" (optional, int)
+                        * "height" (optional, int)
+                        * "resolution" (optional, string "{width}x{height"},
+                                        deprecated)
     thumbnail:      Full URL to a video thumbnail image.
     description:    One-line video description.
     uploader:       Full name of the video uploader.
+    timestamp:      UNIX timestamp of the moment the video became available.
     upload_date:    Video upload date (YYYYMMDD).
+                    If not explicitly set, calculated from timestamp.
     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.
+    subtitles:      The subtitle file contents as a dictionary in the format
+                    {language: subtitles}.
+    duration:       Length of the video in seconds, as an integer.
     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
-
-    The fields should all be Unicode strings.
+    like_count:     Number of positive ratings of the video
+    dislike_count:  Number of negative ratings of the video
+    comment_count:  Number of comments on the video
+    age_limit:      Age restriction for the video, as an integer (years)
+    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)
+    categories:     A list of categories that the video falls in, for example
+                    ["Sports", "Berlin"]
+
+    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.
     """
@@ -77,7 +142,13 @@ class InfoExtractor(object):
     @classmethod
     def suitable(cls, url):
         """Receives a URL and returns True if suitable for this IE."""
-        return re.match(cls._VALID_URL, url) is not None
+
+        # 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):
@@ -107,38 +178,63 @@ class InfoExtractor(object):
         """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):
+    def _request_webpage(self, url_or_request, video_id, note=None, errnote=None, fatal=True):
         """ 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))
+            if video_id is None:
+                self.to_screen(u'%s' % (note,))
+            else:
+                self.to_screen(u'%s: %s' % (video_id, note))
         try:
-            return compat_urllib_request.urlopen(url_or_request)
+            return self._downloader.urlopen(url_or_request)
         except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err:
+            if errnote is False:
+                return False
             if errnote is None:
                 errnote = u'Unable to download webpage'
-            raise ExtractorError(u'%s: %s' % (errnote, compat_str(err)), sys.exc_info()[2])
-
-    def _download_webpage_handle(self, url_or_request, video_id, note=None, errnote=None):
+            errmsg = u'%s: %s' % (errnote, compat_str(err))
+            if fatal:
+                raise ExtractorError(errmsg, sys.exc_info()[2], cause=err)
+            else:
+                self._downloader.report_warning(errmsg)
+                return False
+
+    def _download_webpage_handle(self, url_or_request, video_id, note=None, errnote=None, fatal=True):
         """ 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)
+        urlh = self._request_webpage(url_or_request, video_id, note, errnote, fatal)
+        if urlh is False:
+            assert not fatal
+            return False
         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:
-            encoding = 'utf-8'
-        webpage_bytes = urlh.read()
+            m = re.search(br'<meta[^>]+charset=[\'"]?([^\'")]+)[ /\'">]',
+                          webpage_bytes[:1024])
+            if m:
+                encoding = m.group(1).decode('ascii')
+            elif webpage_bytes.startswith(b'\xff\xfe'):
+                encoding = 'utf-16'
+            else:
+                encoding = 'utf-8'
         if self._downloader.params.get('dump_intermediate_pages', False):
             try:
                 url = url_or_request.get_full_url()
@@ -147,12 +243,75 @@ class InfoExtractor(object):
             self.to_screen(u'Dumping request to ' + url)
             dump = base64.b64encode(webpage_bytes).decode('ascii')
             self._downloader.to_screen(dump)
-        content = webpage_bytes.decode(encoding, 'replace')
+        if self._downloader.params.get('write_pages', False):
+            try:
+                url = url_or_request.get_full_url()
+            except AttributeError:
+                url = url_or_request
+            basen = '%s_%s' % (video_id, url)
+            if len(basen) > 240:
+                h = u'___' + hashlib.md5(basen.encode('utf-8')).hexdigest()
+                basen = basen[:240 - len(h)] + h
+            raw_filename = basen + '.dump'
+            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)
+
+        try:
+            content = webpage_bytes.decode(encoding, 'replace')
+        except LookupError:
+            content = webpage_bytes.decode('utf-8', 'replace')
+
+        if (u'<title>Access to this site is blocked</title>' in content and
+                u'Websense' in content[:512]):
+            msg = u'Access to this webpage has been blocked by Websense filtering software in your network.'
+            blocked_iframe = self._html_search_regex(
+                r'<iframe src="([^"]+)"', content,
+                u'Websense information URL', default=None)
+            if blocked_iframe:
+                msg += u' Visit %s for more details' % blocked_iframe
+            raise ExtractorError(msg, expected=True)
+
         return (content, urlh)
 
-    def _download_webpage(self, url_or_request, video_id, note=None, errnote=None):
+    def _download_webpage(self, url_or_request, video_id, note=None, errnote=None, fatal=True):
         """ Returns the data of the page as a string """
-        return self._download_webpage_handle(url_or_request, video_id, note, errnote)[0]
+        res = self._download_webpage_handle(url_or_request, video_id, note, errnote, fatal)
+        if res is False:
+            return res
+        else:
+            content, _ = res
+            return content
+
+    def _download_xml(self, url_or_request, video_id,
+                      note=u'Downloading XML', errnote=u'Unable to download XML',
+                      transform_source=None, fatal=True):
+        """Return the xml as an xml.etree.ElementTree.Element"""
+        xml_string = self._download_webpage(
+            url_or_request, video_id, note, errnote, fatal=fatal)
+        if xml_string is False:
+            return xml_string
+        if transform_source:
+            xml_string = transform_source(xml_string)
+        return xml.etree.ElementTree.fromstring(xml_string.encode('utf-8'))
+
+    def _download_json(self, url_or_request, video_id,
+                       note=u'Downloading JSON metadata',
+                       errnote=u'Unable to download JSON metadata',
+                       transform_source=None):
+        json_string = self._download_webpage(url_or_request, video_id, note, errnote)
+        if transform_source:
+            json_string = transform_source(json_string)
+        try:
+            return json.loads(json_string)
+        except ValueError as ve:
+            raise ExtractorError('Failed to download JSON', cause=ve)
+
+    def report_warning(self, msg, video_id=None):
+        idstr = u'' if video_id is None else u'%s: ' % video_id
+        self._downloader.report_warning(
+            u'[%s] %s%s' % (self.IE_NAME, idstr, msg))
 
     def to_screen(self, msg):
         """Print msg to screen, prefixing it with '[ie_name]'"""
@@ -175,14 +334,18 @@ class InfoExtractor(object):
         self.to_screen(u'Logging in')
 
     #Methods for following #608
-    def url_result(self, url, ie=None):
+    @staticmethod
+    def url_result(url, ie=None, video_id=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}
+        if video_id is not None:
+            video_info['id'] = video_id
         return video_info
-    def playlist_result(self, entries, playlist_id=None, playlist_title=None):
+    @staticmethod
+    def playlist_result(entries, playlist_id=None, playlist_title=None):
         """Returns a playlist"""
         video_info = {'_type': 'playlist',
                       'entries': entries}
@@ -192,12 +355,12 @@ class InfoExtractor(object):
             video_info['title'] = playlist_title
         return video_info
 
-    def _search_regex(self, pattern, string, name, default=None, fatal=True, flags=0):
+    def _search_regex(self, pattern, string, name, default=_NO_DEFAULT, 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
-        ExtractorError, depending on fatal, specifying the field name.
+        RegexNotFoundError, depending on fatal, specifying the field name.
         """
         if isinstance(pattern, (str, compat_str, compiled_regex_type)):
             mobj = re.search(pattern, string, flags)
@@ -206,7 +369,7 @@ class InfoExtractor(object):
                 mobj = re.search(p, string, flags)
                 if mobj: break
 
-        if sys.stderr.isatty() and os.name != 'nt':
+        if os.name != 'nt' and sys.stderr.isatty():
             _name = u'\033[0;34m%s\033[0m' % name
         else:
             _name = name
@@ -214,16 +377,16 @@ class InfoExtractor(object):
         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:
+        elif default is not _NO_DEFAULT:
             return default
         elif fatal:
-            raise ExtractorError(u'Unable to extract %s' % _name)
+            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):
+    def _html_search_regex(self, pattern, string, name, default=_NO_DEFAULT, fatal=True, flags=0):
         """
         Like _search_regex, but strips HTML tags and unescapes entities.
         """
@@ -265,13 +428,21 @@ class InfoExtractor(object):
 
     # Helper functions for extracting OpenGraph info
     @staticmethod
-    def _og_regex(prop):
-        return r'<meta.+?property=[\'"]og:%s[\'"].+?content=(?:"(.+?)"|\'(.+?)\')' % re.escape(prop)
+    def _og_regexes(prop):
+        content_re = r'content=(?:"([^>]+?)"|\'([^>]+?)\')'
+        property_re = r'(?:name|property)=[\'"]og:%s[\'"]' % re.escape(prop)
+        template = r'<meta[^>]+?%s[^>]+?%s'
+        return [
+            template % (property_re, content_re),
+            template % (content_re, property_re),
+        ]
 
     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)
+        escaped = self._search_regex(self._og_regexes(prop), html, name, flags=re.DOTALL, **kargs)
+        if escaped is None:
+            return None
         return unescapeHTML(escaped)
 
     def _og_search_thumbnail(self, html, **kargs):
@@ -283,10 +454,124 @@ class InfoExtractor(object):
     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', **kargs):
-        return self._html_search_regex([self._og_regex('video:secure_url'),
-                                        self._og_regex('video')],
-                                       html, name, **kargs)
+    def _og_search_video_url(self, html, name='video url', secure=True, **kargs):
+        regexes = self._og_regexes('video')
+        if secure: regexes = self._og_regexes('video:secure_url') + regexes
+        return self._html_search_regex(regexes, html, name, **kargs)
+
+    def _html_search_meta(self, name, html, display_name=None, fatal=False):
+        if display_name is None:
+            display_name = name
+        return self._html_search_regex(
+            r'''(?ix)<meta
+                    (?=[^>]+(?:itemprop|name|property)=["\']%s["\'])
+                    [^>]+content=["\']([^"\']+)["\']''' % re.escape(name),
+            html, display_name, fatal=fatal)
+
+    def _dc_search_uploader(self, html):
+        return self._html_search_meta('dc.creator', html, 'uploader')
+
+    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
+
+    def _media_rating_search(self, html):
+        # See http://www.tjg-designs.com/WP/metadata-code-examples-adding-metadata-to-your-web-pages/
+        rating = self._html_search_meta('rating', html)
+
+        if not rating:
+            return None
+
+        RATING_TABLE = {
+            'safe for kids': 0,
+            'general': 8,
+            '14 years': 14,
+            'mature': 17,
+            'restricted': 19,
+        }
+        return RATING_TABLE.get(rating.lower(), None)
+
+    def _twitter_search_player(self, html):
+        return self._html_search_meta('twitter:player', html,
+            'twitter card player')
+
+    def _sort_formats(self, formats):
+        if not formats:
+            raise ExtractorError(u'No video formats found')
+
+        def _formats_key(f):
+            # TODO remove the following workaround
+            from ..utils import determine_ext
+            if not f.get('ext') and 'url' in f:
+                f['ext'] = determine_ext(f['url'])
+
+            preference = f.get('preference')
+            if preference is None:
+                proto = f.get('protocol')
+                if proto is None:
+                    proto = compat_urllib_parse_urlparse(f.get('url', '')).scheme
+
+                preference = 0 if proto in ['http', 'https'] else -0.1
+                if f.get('ext') in ['f4f', 'f4m']:  # Not yet supported
+                    preference -= 0.5
+
+            if f.get('vcodec') == 'none':  # audio only
+                if self._downloader.params.get('prefer_free_formats'):
+                    ORDER = [u'aac', u'mp3', u'm4a', u'webm', u'ogg', u'opus']
+                else:
+                    ORDER = [u'webm', u'opus', u'ogg', u'mp3', u'aac', u'm4a']
+                ext_preference = 0
+                try:
+                    audio_ext_preference = ORDER.index(f['ext'])
+                except ValueError:
+                    audio_ext_preference = -1
+            else:
+                if self._downloader.params.get('prefer_free_formats'):
+                    ORDER = [u'flv', u'mp4', u'webm']
+                else:
+                    ORDER = [u'webm', u'flv', u'mp4']
+                try:
+                    ext_preference = ORDER.index(f['ext'])
+                except ValueError:
+                    ext_preference = -1
+                audio_ext_preference = 0
+
+            return (
+                preference,
+                f.get('quality') if f.get('quality') is not None else -1,
+                f.get('height') if f.get('height') is not None else -1,
+                f.get('width') if f.get('width') is not None else -1,
+                ext_preference,
+                f.get('tbr') if f.get('tbr') is not None else -1,
+                f.get('vbr') if f.get('vbr') is not None else -1,
+                f.get('abr') if f.get('abr') is not None else -1,
+                audio_ext_preference,
+                f.get('filesize') if f.get('filesize') is not None else -1,
+                f.get('format_id'),
+            )
+        formats.sort(key=_formats_key)
+
+    def http_scheme(self):
+        """ Either "https:" or "https:", depending on the user's preferences """
+        return (
+            'http:'
+            if self._downloader.params.get('prefer_insecure', False)
+            else 'https:')
+
+    def _proto_relative_url(self, url, scheme=None):
+        if url is None:
+            return url
+        if url.startswith('//'):
+            if scheme is None:
+                scheme = self.http_scheme()
+            return scheme + url
+        else:
+            return url
+
 
 class SearchInfoExtractor(InfoExtractor):
     """
@@ -325,8 +610,9 @@ class SearchInfoExtractor(InfoExtractor):
 
     def _get_n_results(self, query, n):
         """Get a specified number of results for a query"""
-        raise NotImplementedError("This method must be implemented by sublclasses")
+        raise NotImplementedError("This method must be implemented by subclasses")
 
     @property
     def SEARCH_KEY(self):
         return self._SEARCH_KEY
+