diff options
Diffstat (limited to 'youtube_dl/YoutubeDL.py')
| -rw-r--r-- | youtube_dl/YoutubeDL.py | 462 | 
1 files changed, 348 insertions, 114 deletions
diff --git a/youtube_dl/YoutubeDL.py b/youtube_dl/YoutubeDL.py index 296c0f992..30ba94666 100644 --- a/youtube_dl/YoutubeDL.py +++ b/youtube_dl/YoutubeDL.py @@ -5,17 +5,53 @@ from __future__ import absolute_import  import errno  import io +import json  import os +import platform  import re  import shutil +import subprocess  import socket  import sys  import time  import traceback -from .utils import * +if os.name == 'nt': +    import ctypes + +from .utils import ( +    compat_cookiejar, +    compat_http_client, +    compat_print, +    compat_str, +    compat_urllib_error, +    compat_urllib_request, +    ContentTooShortError, +    date_from_str, +    DateRange, +    determine_ext, +    DownloadError, +    encodeFilename, +    ExtractorError, +    format_bytes, +    locked_file, +    make_HTTPS_handler, +    MaxDownloadsReached, +    PostProcessingError, +    platform_name, +    preferredencoding, +    SameFileError, +    sanitize_filename, +    subtitles_filename, +    takewhile_inclusive, +    UnavailableVideoError, +    write_json_file, +    write_string, +    YoutubeDLHandler, +)  from .extractor import get_info_extractor, gen_extractors  from .FileDownloader import FileDownloader +from .version import __version__  class YoutubeDL(object): @@ -57,6 +93,7 @@ class YoutubeDL(object):      forcethumbnail:    Force printing thumbnail URL.      forcedescription:  Force printing description.      forcefilename:     Force printing final filename. +    forcejson:         Force printing info_dict as JSON.      simulate:          Do not download the video files.      format:            Video format code.      format_limit:      Highest quality format to try. @@ -68,6 +105,7 @@ class YoutubeDL(object):      playlistend:       Playlist item to end at.      matchtitle:        Download only matching titles.      rejecttitle:       Reject downloads for matching titles. +    logger:            Log messages to a logging.Logger instance.      logtostderr:       Log messages to stderr instead of stdout.      writedescription:  Write the video description to a .description file      writeinfojson:     Write the video description to a .info.json file @@ -91,7 +129,10 @@ class YoutubeDL(object):      downloadarchive:   File name of a file where all downloads are recorded.                         Videos already present in the file are not downloaded                         again. -     +    cookiefile:        File name where cookies should be read from and dumped to. +    nocheckcertificate:Do not verify SSL certificates +    proxy:             URL of the proxy server to use +      The following parameters are not used by YoutubeDL itself, they are used by      the FileDownloader:      nopart, updatetime, buffersize, ratelimit, min_filesize, max_filesize, test, @@ -131,6 +172,8 @@ class YoutubeDL(object):          if '%(stitle)s' in self.params['outtmpl']:              self.report_warning(u'%(stitle)s is deprecated. Use the %(title)s and the --restrict-filenames flag(which also secures %(uploader)s et al) instead.') +        self._setup_opener() +      def add_info_extractor(self, ie):          """Add an InfoExtractor object to the end of the list."""          self._ies.append(ie) @@ -163,7 +206,9 @@ class YoutubeDL(object):      def to_screen(self, message, skip_eol=False):          """Print message to stdout if not in quiet mode.""" -        if not self.params.get('quiet', False): +        if self.params.get('logger'): +            self.params['logger'].debug(message) +        elif not self.params.get('quiet', False):              terminator = [u'\n', u''][skip_eol]              output = message + terminator              write_string(output, self._screen_file) @@ -171,10 +216,47 @@ class YoutubeDL(object):      def to_stderr(self, message):          """Print message to stderr."""          assert type(message) == type(u'') -        output = message + u'\n' -        if 'b' in getattr(self._screen_file, 'mode', '') or sys.version_info[0] < 3: # Python 2 lies about the mode of sys.stdout/sys.stderr -            output = output.encode(preferredencoding()) -        sys.stderr.write(output) +        if self.params.get('logger'): +            self.params['logger'].error(message) +        else: +            output = message + u'\n' +            if 'b' in getattr(self._screen_file, 'mode', '') or sys.version_info[0] < 3: # Python 2 lies about the mode of sys.stdout/sys.stderr +                output = output.encode(preferredencoding()) +            sys.stderr.write(output) + +    def to_console_title(self, message): +        if not self.params.get('consoletitle', False): +            return +        if os.name == 'nt' and ctypes.windll.kernel32.GetConsoleWindow(): +            # c_wchar_p() might not be necessary if `message` is +            # already of type unicode() +            ctypes.windll.kernel32.SetConsoleTitleW(ctypes.c_wchar_p(message)) +        elif 'TERM' in os.environ: +            write_string(u'\033]0;%s\007' % message, self._screen_file) + +    def save_console_title(self): +        if not self.params.get('consoletitle', False): +            return +        if 'TERM' in os.environ: +            # Save the title on stack +            write_string(u'\033[22;0t', self._screen_file) + +    def restore_console_title(self): +        if not self.params.get('consoletitle', False): +            return +        if 'TERM' in os.environ: +            # Restore the title from stack +            write_string(u'\033[23;0t', self._screen_file) + +    def __enter__(self): +        self.save_console_title() +        return self + +    def __exit__(self, *args): +        self.restore_console_title() +     +        if self.params.get('cookiefile') is not None: +            self.cookiejar.save()      def fixed_template(self):          """Checks if the output template is fixed.""" @@ -216,10 +298,10 @@ class YoutubeDL(object):          If stderr is a tty file the 'WARNING:' will be colored          '''          if sys.stderr.isatty() and os.name != 'nt': -            _msg_header=u'\033[0;33mWARNING:\033[0m' +            _msg_header = u'\033[0;33mWARNING:\033[0m'          else: -            _msg_header=u'WARNING:' -        warning_message=u'%s %s' % (_msg_header,message) +            _msg_header = u'WARNING:' +        warning_message = u'%s %s' % (_msg_header, message)          self.to_stderr(warning_message)      def report_error(self, message, tb=None): @@ -234,19 +316,6 @@ class YoutubeDL(object):          error_message = u'%s %s' % (_msg_header, message)          self.trouble(error_message, tb) -    def slow_down(self, start_time, byte_counter): -        """Sleep if the download speed is over the rate limit.""" -        rate_limit = self.params.get('ratelimit', None) -        if rate_limit is None or byte_counter == 0: -            return -        now = time.time() -        elapsed = now - start_time -        if elapsed <= 0.0: -            return -        speed = float(byte_counter) / elapsed -        if speed > rate_limit: -            time.sleep((byte_counter - rate_limit * (now - start_time)) / rate_limit) -      def report_writedescription(self, descfn):          """ Report that the description file is being written """          self.to_screen(u'[info] Writing video description to: ' + descfn) @@ -267,7 +336,7 @@ class YoutubeDL(object):          """Report file has already been fully downloaded."""          try:              self.to_screen(u'[download] %s has already been downloaded' % file_name) -        except (UnicodeEncodeError) as err: +        except UnicodeEncodeError:              self.to_screen(u'[download] The file has already been downloaded')      def increment_downloads(self): @@ -285,16 +354,18 @@ class YoutubeDL(object):                  autonumber_size = 5              autonumber_templ = u'%0' + str(autonumber_size) + u'd'              template_dict['autonumber'] = autonumber_templ % self._num_downloads -            if template_dict['playlist_index'] is not None: +            if template_dict.get('playlist_index') is not None:                  template_dict['playlist_index'] = u'%05d' % template_dict['playlist_index'] -            sanitize = lambda k,v: sanitize_filename( +            sanitize = lambda k, v: sanitize_filename(                  u'NA' if v is None else compat_str(v),                  restricted=self.params.get('restrictfilenames'), -                is_id=(k==u'id')) -            template_dict = dict((k, sanitize(k, v)) for k,v in template_dict.items()) +                is_id=(k == u'id')) +            template_dict = dict((k, sanitize(k, v)) +                                 for k, v in template_dict.items()) -            filename = self.params['outtmpl'] % template_dict +            tmpl = os.path.expanduser(self.params['outtmpl']) +            filename = tmpl % template_dict              return filename          except KeyError as err:              self.report_error(u'Erroneous output template') @@ -306,15 +377,17 @@ class YoutubeDL(object):      def _match_entry(self, info_dict):          """ Returns None iff the file should be downloaded """ -        title = info_dict['title'] -        matchtitle = self.params.get('matchtitle', False) -        if matchtitle: -            if not re.search(matchtitle, title, re.IGNORECASE): -                return u'[download] "' + title + '" title did not match pattern "' + matchtitle + '"' -        rejecttitle = self.params.get('rejecttitle', False) -        if rejecttitle: -            if re.search(rejecttitle, title, re.IGNORECASE): -                return u'"' + title + '" title matched reject pattern "' + rejecttitle + '"' +        if 'title' in info_dict: +            # This can happen when we're just evaluating the playlist +            title = info_dict['title'] +            matchtitle = self.params.get('matchtitle', False) +            if matchtitle: +                if not re.search(matchtitle, title, re.IGNORECASE): +                    return u'[download] "' + title + '" title did not match pattern "' + matchtitle + '"' +            rejecttitle = self.params.get('rejecttitle', False) +            if rejecttitle: +                if re.search(rejecttitle, title, re.IGNORECASE): +                    return u'"' + title + '" title matched reject pattern "' + rejecttitle + '"'          date = info_dict.get('upload_date', None)          if date is not None:              dateRange = self.params.get('daterange', DateRange()) @@ -325,17 +398,23 @@ class YoutubeDL(object):              if age_limit < info_dict.get('age_limit', 0):                  return u'Skipping "' + title + '" because it is age restricted'          if self.in_download_archive(info_dict): -            return (u'%(title)s has already been recorded in archive' -                    % info_dict) +            return (u'%s has already been recorded in archive' +                    % info_dict.get('title', info_dict.get('id', u'video')))          return None -         + +    @staticmethod +    def add_extra_info(info_dict, extra_info): +        '''Set the keys from extra_info in info dict if they are missing''' +        for key, value in extra_info.items(): +            info_dict.setdefault(key, value) +      def extract_info(self, url, download=True, ie_key=None, extra_info={}):          '''          Returns a list with a dictionary for each video we find.          If 'download', also downloads the videos.          extra_info is a dict containing the extra values to add to each result           ''' -         +          if ie_key:              ies = [self.get_info_extractor(ie_key)]          else: @@ -355,17 +434,17 @@ class YoutubeDL(object):                      break                  if isinstance(ie_result, list):                      # Backwards compatibility: old IE result format -                    for result in ie_result: -                        result.update(extra_info)                      ie_result = {                          '_type': 'compat_list',                          'entries': ie_result,                      } -                else: -                    ie_result.update(extra_info) -                if 'extractor' not in ie_result: -                    ie_result['extractor'] = ie.IE_NAME -                return self.process_ie_result(ie_result, download=download) +                self.add_extra_info(ie_result, +                    { +                        'extractor': ie.IE_NAME, +                        'webpage_url': url, +                        'extractor_key': ie.ie_key(), +                    }) +                return self.process_ie_result(ie_result, download, extra_info)              except ExtractorError as de: # An error we somewhat expected                  self.report_error(compat_str(de), de.format_traceback())                  break @@ -377,7 +456,7 @@ class YoutubeDL(object):                      raise          else:              self.report_error(u'no suitable InfoExtractor: %s' % url) -         +      def process_ie_result(self, ie_result, download=True, extra_info={}):          """          Take the result of the ie(may be modified) and resolve all unresolved @@ -389,8 +468,8 @@ class YoutubeDL(object):          result_type = ie_result.get('_type', 'video') # If not given we suppose it's a video, support the default old system          if result_type == 'video': -            ie_result.update(extra_info) -            return self.process_video_result(ie_result) +            self.add_extra_info(ie_result, extra_info) +            return self.process_video_result(ie_result, download=download)          elif result_type == 'url':              # We have to add extra_info to the results because it may be              # contained in a playlist @@ -399,9 +478,10 @@ class YoutubeDL(object):                                       ie_key=ie_result.get('ie_key'),                                       extra_info=extra_info)          elif result_type == 'playlist': +              # We process each entry in the playlist              playlist = ie_result.get('title', None) or ie_result.get('id', None) -            self.to_screen(u'[download] Downloading playlist: %s'  % playlist) +            self.to_screen(u'[download] Downloading playlist: %s' % playlist)              playlist_results = [] @@ -419,17 +499,21 @@ class YoutubeDL(object):              self.to_screen(u"[%s] playlist '%s': Collected %d video ids (downloading %d of them)" %                  (ie_result['extractor'], playlist, n_all_entries, n_entries)) -            for i,entry in enumerate(entries,1): -                self.to_screen(u'[download] Downloading video #%s of %s' %(i, n_entries)) +            for i, entry in enumerate(entries, 1): +                self.to_screen(u'[download] Downloading video #%s of %s' % (i, n_entries))                  extra = { -                         'playlist': playlist,  -                         'playlist_index': i + playliststart, -                         } -                if not 'extractor' in entry: -                    # We set the extractor, if it's an url it will be set then to -                    # the new extractor, but if it's already a video we must make -                    # sure it's present: see issue #877 -                    entry['extractor'] = ie_result['extractor'] +                    'playlist': playlist, +                    'playlist_index': i + playliststart, +                    'extractor': ie_result['extractor'], +                    'webpage_url': ie_result['webpage_url'], +                    'extractor_key': ie_result['extractor_key'], +                } + +                reason = self._match_entry(entry) +                if reason is not None: +                    self.to_screen(u'[download] ' + reason) +                    continue +                  entry_result = self.process_ie_result(entry,                                                        download=download,                                                        extra_info=extra) @@ -438,16 +522,37 @@ class YoutubeDL(object):              return ie_result          elif result_type == 'compat_list':              def _fixup(r): -                r.setdefault('extractor', ie_result['extractor']) +                self.add_extra_info(r, +                    { +                        'extractor': ie_result['extractor'], +                        'webpage_url': ie_result['webpage_url'], +                        'extractor_key': ie_result['extractor_key'], +                    })                  return r              ie_result['entries'] = [ -                self.process_ie_result(_fixup(r), download=download) +                self.process_ie_result(_fixup(r), download, extra_info)                  for r in ie_result['entries']              ]              return ie_result          else:              raise Exception('Invalid result type: %s' % result_type) +    def select_format(self, format_spec, available_formats): +        if format_spec == 'best' or format_spec is None: +            return available_formats[-1] +        elif format_spec == 'worst': +            return available_formats[0] +        else: +            extensions = [u'mp4', u'flv', u'webm', u'3gp'] +            if format_spec in extensions: +                filter_f = lambda f: f['ext'] == format_spec +            else: +                filter_f = lambda f: f['format_id'] == format_spec +            matches = list(filter(filter_f, available_formats)) +            if matches: +                return matches[-1] +        return None +      def process_video_result(self, info_dict, download=True):          assert info_dict.get('_type', 'video') == 'video' @@ -457,8 +562,9 @@ class YoutubeDL(object):              info_dict['playlist_index'] = None          # This extractors handle format selection themselves -        if info_dict['extractor'] in [u'youtube', u'Youku', u'YouPorn', u'mixcloud']: -            self.process_info(info_dict) +        if info_dict['extractor'] in [u'youtube', u'Youku']: +            if download: +                self.process_info(info_dict)              return info_dict          # We now pick which formats have to be downloaded @@ -470,17 +576,17 @@ class YoutubeDL(object):          # We check that all the formats have the format and format_id fields          for (i, format) in enumerate(formats): -            if format.get('format') is None: -                if format.get('height') is not None: -                    if format.get('width') is not None: -                        format_desc = u'%sx%s' % (format['width'], format['height']) -                    else: -                        format_desc = u'%sp' % format['height'] -                else: -                    format_desc = '???' -                format['format'] = format_desc              if format.get('format_id') is None:                  format['format_id'] = compat_str(i) +            if format.get('format') is None: +                format['format'] = u'{id} - {res}{note}'.format( +                    id=format['format_id'], +                    res=self.format_resolution(format), +                    note=u' ({0})'.format(format['format_note']) if format.get('format_note') is not None else '', +                ) +            # Automatically determine file extension if missing +            if 'ext' not in format: +                format['ext'] = determine_ext(format['url'])          if self.params.get('listformats', None):              self.list_formats(info_dict) @@ -502,25 +608,24 @@ class YoutubeDL(object):              formats = sorted(formats, key=_free_formats_key)          req_format = self.params.get('format', 'best') +        if req_format is None: +            req_format = 'best'          formats_to_download = [] -        if req_format == 'best' or req_format is None: -            formats_to_download = [formats[-1]] -        elif req_format == 'worst': -            formats_to_download = [formats[0]]          # The -1 is for supporting YoutubeIE -        elif req_format in ('-1', 'all'): +        if req_format in ('-1', 'all'):              formats_to_download = formats          else: -            # We can accept formats requestd in the format: 34/10/5, we pick +            # We can accept formats requestd in the format: 34/5/best, we pick              # the first that is available, starting from left              req_formats = req_format.split('/')              for rf in req_formats: -                matches = filter(lambda f:f['format_id'] == rf ,formats) -                if matches: -                    formats_to_download = [matches[0]] +                selected_format = self.select_format(rf, formats) +                if selected_format is not None: +                    formats_to_download = [selected_format]                      break          if not formats_to_download: -            raise ExtractorError(u'requested format not available') +            raise ExtractorError(u'requested format not available', +                                 expected=True)          if download:              if len(formats_to_download) > 1: @@ -564,20 +669,22 @@ class YoutubeDL(object):          # Forced printings          if self.params.get('forcetitle', False): -            compat_print(info_dict['title']) +            compat_print(info_dict['fulltitle'])          if self.params.get('forceid', False):              compat_print(info_dict['id'])          if self.params.get('forceurl', False):              # For RTMP URLs, also include the playpath              compat_print(info_dict['url'] + info_dict.get('play_path', u'')) -        if self.params.get('forcethumbnail', False) and 'thumbnail' in info_dict: +        if self.params.get('forcethumbnail', False) and info_dict.get('thumbnail') is not None:              compat_print(info_dict['thumbnail']) -        if self.params.get('forcedescription', False) and 'description' in info_dict: +        if self.params.get('forcedescription', False) and info_dict.get('description') is not None:              compat_print(info_dict['description'])          if self.params.get('forcefilename', False) and filename is not None:              compat_print(filename)          if self.params.get('forceformat', False):              compat_print(info_dict['format']) +        if self.params.get('forcejson', False): +            compat_print(json.dumps(info_dict))          # Do nothing else if in simulate mode          if self.params.get('simulate', False): @@ -608,24 +715,24 @@ class YoutubeDL(object):          if self.params.get('writeannotations', False):              try: -               annofn = filename + u'.annotations.xml' -               self.report_writeannotations(annofn) -               with io.open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile: -                   annofile.write(info_dict['annotations']) +                annofn = filename + u'.annotations.xml' +                self.report_writeannotations(annofn) +                with io.open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile: +                    annofile.write(info_dict['annotations'])              except (KeyError, TypeError):                  self.report_warning(u'There are no annotations to write.')              except (OSError, IOError): -                 self.report_error(u'Cannot write annotations file: ' + annofn) -                 return +                self.report_error(u'Cannot write annotations file: ' + annofn) +                return          subtitles_are_requested = any([self.params.get('writesubtitles', False),                                         self.params.get('writeautomaticsub')]) -        if  subtitles_are_requested and 'subtitles' in info_dict and info_dict['subtitles']: +        if subtitles_are_requested and 'subtitles' in info_dict and info_dict['subtitles']:              # subtitles download errors are already managed as troubles in relevant IE              # that way it will silently go on when used with unsupporting IE              subtitles = info_dict['subtitles'] -            sub_format = self.params.get('subtitlesformat') +            sub_format = self.params.get('subtitlesformat', 'srt')              for sub_lang in subtitles.keys():                  sub = subtitles[sub_lang]                  if sub is None: @@ -640,10 +747,10 @@ class YoutubeDL(object):                      return          if self.params.get('writeinfojson', False): -            infofn = filename + u'.info.json' +            infofn = os.path.splitext(filename)[0] + u'.info.json'              self.report_writeinfojson(infofn)              try: -                json_info_dict = dict((k, v) for k,v in info_dict.items() if not k in ['urlhandle']) +                json_info_dict = dict((k, v) for k, v in info_dict.items() if not k in ['urlhandle'])                  write_json_file(json_info_dict, encodeFilename(infofn))              except (OSError, IOError):                  self.report_error(u'Cannot write metadata to JSON file ' + infofn) @@ -697,7 +804,7 @@ class YoutubeDL(object):          for url in url_list:              try:                  #It also downloads the videos -                videos = self.extract_info(url) +                self.extract_info(url)              except UnavailableVideoError:                  self.report_error(u'unable to download video')              except MaxDownloadsReached: @@ -713,7 +820,7 @@ class YoutubeDL(object):          keep_video = None          for pp in self._pps:              try: -                keep_video_wish,new_info = pp.run(info) +                keep_video_wish, new_info = pp.run(info)                  if keep_video_wish is not None:                      if keep_video_wish:                          keep_video = keep_video_wish @@ -733,7 +840,16 @@ class YoutubeDL(object):          fn = self.params.get('download_archive')          if fn is None:              return False -        vid_id = info_dict['extractor'] + u' ' + info_dict['id'] +        extractor = info_dict.get('extractor_id') +        if extractor is None: +            if 'id' in info_dict: +                extractor = info_dict.get('ie_key')  # key in a playlist +        if extractor is None: +            return False  # Incomplete video information +        # Future-proof against any change in case +        # and backwards compatibility with prior versions +        extractor = extractor.lower() +        vid_id = extractor + u' ' + info_dict['id']          try:              with locked_file(fn, 'r', encoding='utf-8') as archive_file:                  for line in archive_file: @@ -752,16 +868,134 @@ class YoutubeDL(object):          with locked_file(fn, 'a', encoding='utf-8') as archive_file:              archive_file.write(vid_id + u'\n') +    @staticmethod +    def format_resolution(format, default='unknown'): +        if format.get('_resolution') is not None: +            return format['_resolution'] +        if format.get('height') is not None: +            if format.get('width') is not None: +                res = u'%sx%s' % (format['width'], format['height']) +            else: +                res = u'%sp' % format['height'] +        else: +            res = default +        return res +      def list_formats(self, info_dict): -        formats_s = [] -        for format in info_dict.get('formats', [info_dict]): -            formats_s.append("%s\t:\t%s\t[%s]" % (format['format_id'], -                                                format['ext'], -                                                format.get('format', '???'), -                                                ) -                            ) -        if len(formats_s) != 1: -            formats_s[0]  += ' (worst)' -            formats_s[-1] += ' (best)' -        formats_s = "\n".join(formats_s) -        self.to_screen(u"[info] Available formats for %s:\nformat code\textension\n%s" % (info_dict['id'], formats_s))  +        def format_note(fdict): +            res = u'' +            if fdict.get('format_note') is not None: +                res += fdict['format_note'] + u' ' +            if fdict.get('vcodec') is not None: +                res += u'%-5s' % fdict['vcodec'] +            elif fdict.get('vbr') is not None: +                res += u'video' +            if fdict.get('vbr') is not None: +                res += u'@%4dk' % fdict['vbr'] +            if fdict.get('acodec') is not None: +                if res: +                    res += u', ' +                res += u'%-5s' % fdict['acodec'] +            elif fdict.get('abr') is not None: +                if res: +                    res += u', ' +                res += 'audio' +            if fdict.get('abr') is not None: +                res += u'@%3dk' % fdict['abr'] +            if fdict.get('filesize') is not None: +                if res: +                    res += u', ' +                res += format_bytes(fdict['filesize']) +            return res + +        def line(format, idlen=20): +            return ((u'%-' + compat_str(idlen + 1) + u's%-10s%-12s%s') % ( +                format['format_id'], +                format['ext'], +                self.format_resolution(format), +                format_note(format), +            )) + +        formats = info_dict.get('formats', [info_dict]) +        idlen = max(len(u'format code'), +                    max(len(f['format_id']) for f in formats)) +        formats_s = [line(f, idlen) for f in formats] +        if len(formats) > 1: +            formats_s[0] += (' ' if format_note(formats[0]) else '') + '(worst)' +            formats_s[-1] += (' ' if format_note(formats[-1]) else '') + '(best)' + +        header_line = line({ +            'format_id': u'format code', 'ext': u'extension', +            '_resolution': u'resolution', 'format_note': u'note'}, idlen=idlen) +        self.to_screen(u'[info] Available formats for %s:\n%s\n%s' % +                       (info_dict['id'], header_line, u"\n".join(formats_s))) + +    def urlopen(self, req): +        """ Start an HTTP download """ +        return self._opener.open(req) + +    def print_debug_header(self): +        if not self.params.get('verbose'): +            return +        write_string(u'[debug] youtube-dl version ' + __version__ + u'\n') +        try: +            sp = subprocess.Popen( +                ['git', 'rev-parse', '--short', 'HEAD'], +                stdout=subprocess.PIPE, stderr=subprocess.PIPE, +                cwd=os.path.dirname(os.path.abspath(__file__))) +            out, err = sp.communicate() +            out = out.decode().strip() +            if re.match('[0-9a-f]+', out): +                write_string(u'[debug] Git HEAD: ' + out + u'\n') +        except: +            try: +                sys.exc_clear() +            except: +                pass +        write_string(u'[debug] Python version %s - %s' % +                     (platform.python_version(), platform_name()) + u'\n') + +        proxy_map = {} +        for handler in self._opener.handlers: +            if hasattr(handler, 'proxies'): +                proxy_map.update(handler.proxies) +        write_string(u'[debug] Proxy map: ' + compat_str(proxy_map) + u'\n') + +    def _setup_opener(self, timeout=300): +        opts_cookiefile = self.params.get('cookiefile') +        opts_proxy = self.params.get('proxy') + +        if opts_cookiefile is None: +            self.cookiejar = compat_cookiejar.CookieJar() +        else: +            self.cookiejar = compat_cookiejar.MozillaCookieJar( +                opts_cookiefile) +            if os.access(opts_cookiefile, os.R_OK): +                self.cookiejar.load() + +        cookie_processor = compat_urllib_request.HTTPCookieProcessor( +            self.cookiejar) +        if opts_proxy is not None: +            if opts_proxy == '': +                proxies = {} +            else: +                proxies = {'http': opts_proxy, 'https': opts_proxy} +        else: +            proxies = compat_urllib_request.getproxies() +            # Set HTTPS proxy to HTTP one if given (https://github.com/rg3/youtube-dl/issues/805) +            if 'http' in proxies and 'https' not in proxies: +                proxies['https'] = proxies['http'] +        proxy_handler = compat_urllib_request.ProxyHandler(proxies) +        https_handler = make_HTTPS_handler( +            self.params.get('nocheckcertificate', False)) +        opener = compat_urllib_request.build_opener( +            https_handler, proxy_handler, cookie_processor, YoutubeDLHandler()) +        # Delete the default user-agent header, which would otherwise apply in +        # cases where our custom HTTP handler doesn't come into play +        # (See https://github.com/rg3/youtube-dl/issues/1309 for details) +        opener.addheaders = [] +        self._opener = opener + +        # TODO remove this global modification +        compat_urllib_request.install_opener(opener) +        socket.setdefaulttimeout(timeout)  | 
