X-Git-Url: https://git.rapsys.eu/youtubedl/blobdiff_plain/b238854ce845f3796daac74edab2e8a373e8ba1a..411e5125fad4faf2f85ea292fbb894a05f929336:/youtube_dl/YoutubeDL.py diff --git a/youtube_dl/YoutubeDL.py b/youtube_dl/YoutubeDL.py old mode 100644 new mode 100755 index b68b110..dc0ba98 --- a/youtube_dl/YoutubeDL.py +++ b/youtube_dl/YoutubeDL.py @@ -1,11 +1,14 @@ #!/usr/bin/env python # -*- coding: utf-8 -*- -from __future__ import absolute_import +from __future__ import absolute_import, unicode_literals +import collections +import datetime import errno import io import json +import locale import os import platform import re @@ -22,21 +25,24 @@ if os.name == 'nt': from .utils import ( compat_cookiejar, compat_http_client, - compat_print, compat_str, compat_urllib_error, compat_urllib_request, ContentTooShortError, date_from_str, DateRange, + DEFAULT_OUTTMPL, determine_ext, DownloadError, encodeFilename, ExtractorError, format_bytes, + formatSeconds, + get_term_width, locked_file, make_HTTPS_handler, MaxDownloadsReached, + PagedList, PostProcessingError, platform_name, preferredencoding, @@ -45,12 +51,15 @@ from .utils import ( subtitles_filename, takewhile_inclusive, UnavailableVideoError, + url_basename, write_json_file, write_string, YoutubeDLHandler, + prepend_extension, ) from .extractor import get_info_extractor, gen_extractors -from .FileDownloader import FileDownloader +from .downloader import get_suitable_downloader +from .postprocessor import FFmpegMergerPP from .version import __version__ @@ -87,12 +96,14 @@ class YoutubeDL(object): usenetrc: Use netrc for authentication instead. verbose: Print additional info to stdout. quiet: Do not print messages to stdout. + no_warnings: Do not print out anything for warnings. forceurl: Force printing final URL. forcetitle: Force printing title. forceid: Force printing ID. forcethumbnail: Force printing thumbnail URL. forcedescription: Force printing description. forcefilename: Force printing final filename. + forceduration: Force printing duration. forcejson: Force printing info_dict as JSON. simulate: Do not download the video files. format: Video format code. @@ -126,18 +137,40 @@ class YoutubeDL(object): noplaylist: Download single video instead of a playlist if in doubt. age_limit: An integer representing the user's age in years. Unsuitable videos for the given age are skipped. - download_archive: File name of a file where all downloads are recorded. + min_views: An integer representing the minimum view count the video + must have in order to not be skipped. + Videos without view count information are always + downloaded. None for no limit. + max_views: An integer representing the maximum view count. + Videos that are more popular than that are not + downloaded. + Videos without view count information are always + downloaded. None for no limit. + download_archive: 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 + prefer_insecure: Use HTTP instead of HTTPS to retrieve information. + At the moment, this is only supported by YouTube. proxy: URL of the proxy server to use socket_timeout: Time to wait for unresponsive hosts, in seconds + bidi_workaround: Work around buggy terminals without bidirectional text + support, using fridibi + debug_printtraffic:Print out sent and received HTTP traffic + include_ads: Download ads as well + default_search: Prepend this string if an input url is not valid. + 'auto' for elaborate guessing + encoding: Use this encoding instead of the system-specified. The following parameters are not used by YoutubeDL itself, they are used by the FileDownloader: nopart, updatetime, buffersize, ratelimit, min_filesize, max_filesize, test, noresizebuffer, retries, continuedl, noprogress, consoletitle + + The following options are used by the post processors: + prefer_ffmpeg: If True, use ffmpeg instead of avconv if both are available, + otherwise prefer avconv. """ params = None @@ -149,6 +182,8 @@ class YoutubeDL(object): def __init__(self, params=None): """Create a FileDownloader object with the given options.""" + if params is None: + params = {} self._ies = [] self._ies_instances = {} self._pps = [] @@ -156,22 +191,48 @@ class YoutubeDL(object): self._download_retcode = 0 self._num_downloads = 0 self._screen_file = [sys.stdout, sys.stderr][params.get('logtostderr', False)] - self.params = {} if params is None else params + self._err_file = sys.stderr + self.params = params + + if params.get('bidi_workaround', False): + try: + import pty + master, slave = pty.openpty() + width = get_term_width() + if width is None: + width_args = [] + else: + width_args = ['-w', str(width)] + sp_kwargs = dict( + stdin=subprocess.PIPE, + stdout=slave, + stderr=self._err_file) + try: + self._output_process = subprocess.Popen( + ['bidiv'] + width_args, **sp_kwargs + ) + except OSError: + self._output_process = subprocess.Popen( + ['fribidi', '-c', 'UTF-8'] + width_args, **sp_kwargs) + self._output_channel = os.fdopen(master, 'rb') + except OSError as ose: + if ose.errno == 2: + self.report_warning('Could not find fribidi executable, ignoring --bidi-workaround . Make sure that fribidi is an executable file in one of the directories in your $PATH.') + else: + raise if (sys.version_info >= (3,) and sys.platform != 'win32' and sys.getfilesystemencoding() in ['ascii', 'ANSI_X3.4-1968'] and not params['restrictfilenames']): # On Python 3, the Unicode filesystem API will throw errors (#1474) self.report_warning( - u'Assuming --restrict-filenames since file system encoding ' - u'cannot encode all charactes. ' - u'Set the LC_ALL environment variable to fix this.') + 'Assuming --restrict-filenames since file system encoding ' + 'cannot encode all charactes. ' + 'Set the LC_ALL environment variable to fix this.') self.params['restrictfilenames'] = True - self.fd = FileDownloader(self, self.params) - if '%(stitle)s' in self.params.get('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.report_warning('%(stitle)s is deprecated. Use the %(title)s and the --restrict-filenames flag(which also secures %(uploader)s et al) instead.') self._setup_opener() @@ -205,25 +266,50 @@ class YoutubeDL(object): self._pps.append(pp) pp.set_downloader(self) + def add_progress_hook(self, ph): + """Add the progress hook (currently only for the file downloader)""" + self._progress_hooks.append(ph) + + def _bidi_workaround(self, message): + if not hasattr(self, '_output_channel'): + return message + + assert hasattr(self, '_output_process') + assert type(message) == type('') + line_count = message.count('\n') + 1 + self._output_process.stdin.write((message + '\n').encode('utf-8')) + self._output_process.stdin.flush() + res = ''.join(self._output_channel.readline().decode('utf-8') + for _ in range(line_count)) + return res[:-len('\n')] + def to_screen(self, message, skip_eol=False): + """Print message to stdout if not in quiet mode.""" + return self.to_stdout(message, skip_eol, check_quiet=True) + + def _write_string(self, s, out=None): + write_string(s, out=out, encoding=self.params.get('encoding')) + + def to_stdout(self, message, skip_eol=False, check_quiet=False): """Print message to stdout if not in quiet mode.""" if self.params.get('logger'): self.params['logger'].debug(message) - elif not self.params.get('quiet', False): - terminator = [u'\n', u''][skip_eol] + elif not check_quiet or not self.params.get('quiet', False): + message = self._bidi_workaround(message) + terminator = ['\n', ''][skip_eol] output = message + terminator - write_string(output, self._screen_file) + + self._write_string(output, self._screen_file) def to_stderr(self, message): """Print message to stderr.""" - assert type(message) == type(u'') + assert type(message) == type('') 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) + message = self._bidi_workaround(message) + output = message + '\n' + self._write_string(output, self._err_file) def to_console_title(self, message): if not self.params.get('consoletitle', False): @@ -233,21 +319,21 @@ class YoutubeDL(object): # 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) + self._write_string('\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) + self._write_string('\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) + self._write_string('\033[23;0t', self._screen_file) def __enter__(self): self.save_console_title() @@ -255,7 +341,7 @@ class YoutubeDL(object): def __exit__(self, *args): self.restore_console_title() - + if self.params.get('cookiefile') is not None: self.cookiejar.save() @@ -273,13 +359,13 @@ class YoutubeDL(object): if self.params.get('verbose'): if tb is None: if sys.exc_info()[0]: # if .trouble has been called from an except block - tb = u'' + tb = '' if hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]: - tb += u''.join(traceback.format_exception(*sys.exc_info()[1].exc_info)) + tb += ''.join(traceback.format_exception(*sys.exc_info()[1].exc_info)) tb += compat_str(traceback.format_exc()) else: tb_data = traceback.format_list(traceback.extract_stack()) - tb = u''.join(tb_data) + tb = ''.join(tb_data) self.to_stderr(tb) if not self.params.get('ignoreerrors', False): if sys.exc_info()[0] and hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]: @@ -294,51 +380,36 @@ class YoutubeDL(object): Print the message to stderr, it will be prefixed with 'WARNING:' 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' + if self.params.get('logger') is not None: + self.params['logger'].warning(message) else: - _msg_header = u'WARNING:' - warning_message = u'%s %s' % (_msg_header, message) - self.to_stderr(warning_message) + if self.params.get('no_warnings'): + return + if self._err_file.isatty() and os.name != 'nt': + _msg_header = '\033[0;33mWARNING:\033[0m' + else: + _msg_header = 'WARNING:' + warning_message = '%s %s' % (_msg_header, message) + self.to_stderr(warning_message) def report_error(self, message, tb=None): ''' Do the same as trouble, but prefixes the message with 'ERROR:', colored in red if stderr is a tty file. ''' - if sys.stderr.isatty() and os.name != 'nt': - _msg_header = u'\033[0;31mERROR:\033[0m' + if self._err_file.isatty() and os.name != 'nt': + _msg_header = '\033[0;31mERROR:\033[0m' else: - _msg_header = u'ERROR:' - error_message = u'%s %s' % (_msg_header, message) + _msg_header = 'ERROR:' + error_message = '%s %s' % (_msg_header, message) self.trouble(error_message, tb) - def report_writedescription(self, descfn): - """ Report that the description file is being written """ - self.to_screen(u'[info] Writing video description to: ' + descfn) - - def report_writesubtitles(self, sub_filename): - """ Report that the subtitles file is being written """ - self.to_screen(u'[info] Writing video subtitles to: ' + sub_filename) - - def report_writeinfojson(self, infofn): - """ Report that the metadata file has been written """ - self.to_screen(u'[info] Video description metadata as JSON to: ' + infofn) - - def report_writeannotations(self, annofn): - """ Report that the annotations file has been written. """ - self.to_screen(u'[info] Writing video annotations to: ' + annofn) - def report_file_already_downloaded(self, file_name): """Report file has already been fully downloaded.""" try: - self.to_screen(u'[download] %s has already been downloaded' % file_name) + self.to_screen('[download] %s has already been downloaded' % file_name) except UnicodeEncodeError: - self.to_screen(u'[download] The file has already been downloaded') - - def increment_downloads(self): - """Increment the ordinal that assigns a number to each file.""" - self._num_downloads += 1 + self.to_screen('[download] The file has already been downloaded') def prepare_filename(self, info_dict): """Generate the output filename.""" @@ -349,54 +420,69 @@ class YoutubeDL(object): autonumber_size = self.params.get('autonumber_size') if autonumber_size is None: autonumber_size = 5 - autonumber_templ = u'%0' + str(autonumber_size) + u'd' + autonumber_templ = '%0' + str(autonumber_size) + 'd' template_dict['autonumber'] = autonumber_templ % self._num_downloads if template_dict.get('playlist_index') is not None: - template_dict['playlist_index'] = u'%05d' % template_dict['playlist_index'] + template_dict['playlist_index'] = '%05d' % template_dict['playlist_index'] + if template_dict.get('resolution') is None: + if template_dict.get('width') and template_dict.get('height'): + template_dict['resolution'] = '%dx%d' % (template_dict['width'], template_dict['height']) + elif template_dict.get('height'): + template_dict['resolution'] = '%sp' % template_dict['height'] + elif template_dict.get('width'): + template_dict['resolution'] = '?x%d' % template_dict['width'] sanitize = lambda k, v: sanitize_filename( - u'NA' if v is None else compat_str(v), + compat_str(v), restricted=self.params.get('restrictfilenames'), - is_id=(k == u'id')) + is_id=(k == 'id')) template_dict = dict((k, sanitize(k, v)) - for k, v in template_dict.items()) + for k, v in template_dict.items() + if v is not None) + template_dict = collections.defaultdict(lambda: 'NA', template_dict) - tmpl = os.path.expanduser(self.params['outtmpl']) + outtmpl = self.params.get('outtmpl', DEFAULT_OUTTMPL) + tmpl = os.path.expanduser(outtmpl) filename = tmpl % template_dict return filename - except KeyError as err: - self.report_error(u'Erroneous output template') - return None except ValueError as err: - self.report_error(u'Error in output template: ' + str(err) + u' (encoding: ' + repr(preferredencoding()) + ')') + self.report_error('Error in output template: ' + str(err) + ' (encoding: ' + repr(preferredencoding()) + ')') return None def _match_entry(self, info_dict): """ Returns None iff the file should be downloaded """ + video_title = info_dict.get('title', info_dict.get('id', 'video')) 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 + '"' + return '"' + 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 + '"' + return '"' + title + '" title matched reject pattern "' + rejecttitle + '"' date = info_dict.get('upload_date', None) if date is not None: dateRange = self.params.get('daterange', DateRange()) if date not in dateRange: - return u'[download] %s upload date is not in range %s' % (date_from_str(date).isoformat(), dateRange) + return '%s upload date is not in range %s' % (date_from_str(date).isoformat(), dateRange) + view_count = info_dict.get('view_count', None) + if view_count is not None: + min_views = self.params.get('min_views') + if min_views is not None and view_count < min_views: + return 'Skipping %s, because it has not reached minimum view count (%d/%d)' % (video_title, view_count, min_views) + max_views = self.params.get('max_views') + if max_views is not None and view_count > max_views: + return 'Skipping %s, because it has exceeded the maximum view count (%d/%d)' % (video_title, view_count, max_views) age_limit = self.params.get('age_limit') if age_limit is not None: if age_limit < info_dict.get('age_limit', 0): - return u'Skipping "' + title + '" because it is age restricted' + return 'Skipping "' + title + '" because it is age restricted' if self.in_download_archive(info_dict): - return (u'%s has already been recorded in archive' - % info_dict.get('title', info_dict.get('id', u'video'))) + return '%s has already been recorded in archive' % video_title return None @staticmethod @@ -405,7 +491,8 @@ class YoutubeDL(object): for key, value in extra_info.items(): info_dict.setdefault(key, value) - def extract_info(self, url, download=True, ie_key=None, extra_info={}): + def extract_info(self, url, download=True, ie_key=None, extra_info={}, + process=True): ''' Returns a list with a dictionary for each video we find. If 'download', also downloads the videos. @@ -422,8 +509,8 @@ class YoutubeDL(object): continue if not ie.working(): - self.report_warning(u'The program functionality for this site has been marked as broken, ' - u'and will probably not work.') + self.report_warning('The program functionality for this site has been marked as broken, ' + 'and will probably not work.') try: ie_result = ie.extract(url) @@ -435,16 +522,16 @@ class YoutubeDL(object): '_type': 'compat_list', 'entries': ie_result, } - 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) + self.add_default_extra_info(ie_result, ie, url) + if process: + return self.process_ie_result(ie_result, download, extra_info) + else: + return ie_result except ExtractorError as de: # An error we somewhat expected self.report_error(compat_str(de), de.format_traceback()) break + except MaxDownloadsReached: + raise except Exception as e: if self.params.get('ignoreerrors', False): self.report_error(compat_str(e), tb=compat_str(traceback.format_exc())) @@ -452,7 +539,15 @@ class YoutubeDL(object): else: raise else: - self.report_error(u'no suitable InfoExtractor: %s' % url) + self.report_error('no suitable InfoExtractor for URL %s' % url) + + def add_default_extra_info(self, ie_result, ie, url): + self.add_extra_info(ie_result, { + 'extractor': ie.IE_NAME, + 'webpage_url': url, + 'webpage_url_basename': url_basename(url), + 'extractor_key': ie.ie_key(), + }) def process_ie_result(self, ie_result, download=True, extra_info={}): """ @@ -474,41 +569,75 @@ class YoutubeDL(object): download, ie_key=ie_result.get('ie_key'), extra_info=extra_info) + elif result_type == 'url_transparent': + # Use the information from the embedding page + info = self.extract_info( + ie_result['url'], ie_key=ie_result.get('ie_key'), + extra_info=extra_info, download=False, process=False) + + def make_result(embedded_info): + new_result = ie_result.copy() + for f in ('_type', 'url', 'ext', 'player_url', 'formats', + 'entries', 'ie_key', 'duration', + 'subtitles', 'annotations', 'format', + 'thumbnail', 'thumbnails'): + if f in new_result: + del new_result[f] + if f in embedded_info: + new_result[f] = embedded_info[f] + return new_result + new_result = make_result(info) + + assert new_result.get('_type') != 'url_transparent' + if new_result.get('_type') == 'compat_list': + new_result['entries'] = [ + make_result(e) for e in new_result['entries']] + + return self.process_ie_result( + new_result, download=download, 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('[download] Downloading playlist: %s' % playlist) playlist_results = [] - n_all_entries = len(ie_result['entries']) playliststart = self.params.get('playliststart', 1) - 1 - playlistend = self.params.get('playlistend', -1) - + playlistend = self.params.get('playlistend', None) + # For backwards compatibility, interpret -1 as whole list if playlistend == -1: - entries = ie_result['entries'][playliststart:] - else: - entries = ie_result['entries'][playliststart:playlistend] + playlistend = None - n_entries = len(entries) - - self.to_screen(u"[%s] playlist '%s': Collected %d video ids (downloading %d of them)" % - (ie_result['extractor'], playlist, n_all_entries, n_entries)) + if isinstance(ie_result['entries'], list): + n_all_entries = len(ie_result['entries']) + entries = ie_result['entries'][playliststart:playlistend] + n_entries = len(entries) + self.to_screen( + "[%s] playlist %s: Collected %d video ids (downloading %d of them)" % + (ie_result['extractor'], playlist, n_all_entries, n_entries)) + else: + assert isinstance(ie_result['entries'], PagedList) + entries = ie_result['entries'].getslice( + playliststart, playlistend) + n_entries = len(entries) + self.to_screen( + "[%s] playlist %s: Downloading %d videos" % + (ie_result['extractor'], playlist, n_entries)) for i, entry in enumerate(entries, 1): - self.to_screen(u'[download] Downloading video #%s of %s' % (i, n_entries)) + self.to_screen('[download] Downloading video #%s of %s' % (i, n_entries)) extra = { 'playlist': playlist, 'playlist_index': i + playliststart, 'extractor': ie_result['extractor'], 'webpage_url': ie_result['webpage_url'], + 'webpage_url_basename': url_basename(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) + self.to_screen('[download] ' + reason) continue entry_result = self.process_ie_result(entry, @@ -523,6 +652,7 @@ class YoutubeDL(object): { 'extractor': ie_result['extractor'], 'webpage_url': ie_result['webpage_url'], + 'webpage_url_basename': url_basename(ie_result['webpage_url']), 'extractor_key': ie_result['extractor_key'], }) return r @@ -539,8 +669,32 @@ class YoutubeDL(object): return available_formats[-1] elif format_spec == 'worst': return available_formats[0] + elif format_spec == 'bestaudio': + audio_formats = [ + f for f in available_formats + if f.get('vcodec') == 'none'] + if audio_formats: + return audio_formats[-1] + elif format_spec == 'worstaudio': + audio_formats = [ + f for f in available_formats + if f.get('vcodec') == 'none'] + if audio_formats: + return audio_formats[0] + elif format_spec == 'bestvideo': + video_formats = [ + f for f in available_formats + if f.get('acodec') == 'none'] + if video_formats: + return video_formats[-1] + elif format_spec == 'worstvideo': + video_formats = [ + f for f in available_formats + if f.get('acodec') == 'none'] + if video_formats: + return video_formats[0] else: - extensions = [u'mp4', u'flv', u'webm', u'3gp'] + extensions = ['mp4', 'flv', 'webm', '3gp'] if format_spec in extensions: filter_f = lambda f: f['ext'] == format_spec else: @@ -553,13 +707,37 @@ class YoutubeDL(object): def process_video_result(self, info_dict, download=True): assert info_dict.get('_type', 'video') == 'video' + if 'id' not in info_dict: + raise ExtractorError('Missing "id" field in extractor result') + if 'title' not in info_dict: + raise ExtractorError('Missing "title" field in extractor result') + if 'playlist' not in info_dict: # It isn't part of a playlist info_dict['playlist'] = None info_dict['playlist_index'] = None + thumbnails = info_dict.get('thumbnails') + if thumbnails: + thumbnails.sort(key=lambda t: ( + t.get('width'), t.get('height'), t.get('url'))) + for t in thumbnails: + if 'width' in t and 'height' in t: + t['resolution'] = '%dx%d' % (t['width'], t['height']) + + if thumbnails and 'thumbnail' not in info_dict: + info_dict['thumbnail'] = thumbnails[-1]['url'] + + if 'display_id' not in info_dict and 'id' in info_dict: + info_dict['display_id'] = info_dict['id'] + + if info_dict.get('upload_date') is None and info_dict.get('timestamp') is not None: + upload_date = datetime.datetime.utcfromtimestamp( + info_dict['timestamp']) + info_dict['upload_date'] = upload_date.strftime('%Y%m%d') + # This extractors handle format selection themselves - if info_dict['extractor'] in [u'youtube', u'Youku']: + if info_dict['extractor'] in ['Youku']: if download: self.process_info(info_dict) return info_dict @@ -571,40 +749,45 @@ class YoutubeDL(object): else: formats = info_dict['formats'] + if not formats: + raise ExtractorError('No video formats found!') + # We check that all the formats have the format and format_id fields - for (i, format) in enumerate(formats): + for i, format in enumerate(formats): + if 'url' not in format: + raise ExtractorError('Missing "url" key in result (index %d)' % i) + 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( + format['format'] = '{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 '', + note=' ({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) - return + format['ext'] = determine_ext(format['url']).lower() format_limit = self.params.get('format_limit', None) if format_limit: formats = list(takewhile_inclusive( lambda f: f['format_id'] != format_limit, formats )) - if self.params.get('prefer_free_formats'): - def _free_formats_key(f): - try: - ext_ord = [u'flv', u'mp4', u'webm'].index(f['ext']) - except ValueError: - ext_ord = -1 - # We only compare the extension if they have the same height and width - return (f.get('height'), f.get('width'), ext_ord) - formats = sorted(formats, key=_free_formats_key) - - req_format = self.params.get('format', 'best') + + # TODO Central sorting goes here + + if formats[0] is not info_dict: + # only set the 'formats' fields if the original info_dict list them + # otherwise we end up with a circular reference, the first (and unique) + # element in the 'formats' field in info_dict is info_dict itself, + # wich can't be exported to json + info_dict['formats'] = formats + if self.params.get('listformats', None): + self.list_formats(info_dict) + return + + req_format = self.params.get('format') if req_format is None: req_format = 'best' formats_to_download = [] @@ -612,21 +795,35 @@ class YoutubeDL(object): if req_format in ('-1', 'all'): formats_to_download = formats else: - # We can accept formats requestd in the format: 34/5/best, we pick + # We can accept formats requested 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: - selected_format = self.select_format(rf, formats) + if re.match(r'.+?\+.+?', rf) is not None: + # Two formats have been requested like '137+139' + format_1, format_2 = rf.split('+') + formats_info = (self.select_format(format_1, formats), + self.select_format(format_2, formats)) + if all(formats_info): + selected_format = { + 'requested_formats': formats_info, + 'format': rf, + 'ext': formats_info[0]['ext'], + } + else: + selected_format = None + else: + 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('requested format not available', expected=True) if download: if len(formats_to_download) > 1: - self.to_screen(u'[info] %s: downloading video in %s formats' % (info_dict['id'], len(formats_to_download))) + self.to_screen('[info] %s: downloading video in %s formats' % (info_dict['id'], len(formats_to_download))) for format in formats_to_download: new_info = dict(info_dict) new_info.update(format) @@ -639,12 +836,15 @@ class YoutubeDL(object): """Process a single resolved IE result.""" assert info_dict.get('_type', 'video') == 'video' - #We increment the download the download count here to match the previous behaviour. - self.increment_downloads() + + max_downloads = self.params.get('max_downloads') + if max_downloads is not None: + if self._num_downloads >= int(max_downloads): + raise MaxDownloadsReached() info_dict['fulltitle'] = info_dict['title'] if len(info_dict['title']) > 200: - info_dict['title'] = info_dict['title'][:197] + u'...' + info_dict['title'] = info_dict['title'][:197] + '...' # Keep for backwards compatibility info_dict['stitle'] = info_dict['title'] @@ -654,34 +854,34 @@ class YoutubeDL(object): reason = self._match_entry(info_dict) if reason is not None: - self.to_screen(u'[download] ' + reason) + self.to_screen('[download] ' + reason) return - max_downloads = self.params.get('max_downloads') - if max_downloads is not None: - if self._num_downloads > int(max_downloads): - raise MaxDownloadsReached() + self._num_downloads += 1 filename = self.prepare_filename(info_dict) # Forced printings if self.params.get('forcetitle', False): - compat_print(info_dict['fulltitle']) + self.to_stdout(info_dict['fulltitle']) if self.params.get('forceid', False): - compat_print(info_dict['id']) + self.to_stdout(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'')) + self.to_stdout(info_dict['url'] + info_dict.get('play_path', '')) if self.params.get('forcethumbnail', False) and info_dict.get('thumbnail') is not None: - compat_print(info_dict['thumbnail']) + self.to_stdout(info_dict['thumbnail']) if self.params.get('forcedescription', False) and info_dict.get('description') is not None: - compat_print(info_dict['description']) + self.to_stdout(info_dict['description']) if self.params.get('forcefilename', False) and filename is not None: - compat_print(filename) + self.to_stdout(filename) + if self.params.get('forceduration', False) and info_dict.get('duration') is not None: + self.to_stdout(formatSeconds(info_dict['duration'])) if self.params.get('forceformat', False): - compat_print(info_dict['format']) + self.to_stdout(info_dict['format']) if self.params.get('forcejson', False): - compat_print(json.dumps(info_dict)) + info_dict['_filename'] = filename + self.to_stdout(json.dumps(info_dict)) # Do nothing else if in simulate mode if self.params.get('simulate', False): @@ -692,35 +892,41 @@ class YoutubeDL(object): try: dn = os.path.dirname(encodeFilename(filename)) - if dn != '' and not os.path.exists(dn): + if dn and not os.path.exists(dn): os.makedirs(dn) except (OSError, IOError) as err: - self.report_error(u'unable to create directory ' + compat_str(err)) + self.report_error('unable to create directory ' + compat_str(err)) return if self.params.get('writedescription', False): - try: - descfn = filename + u'.description' - self.report_writedescription(descfn) - with io.open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile: - descfile.write(info_dict['description']) - except (KeyError, TypeError): - self.report_warning(u'There\'s no description to write.') - except (OSError, IOError): - self.report_error(u'Cannot write description file ' + descfn) - return + descfn = filename + '.description' + if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(descfn)): + self.to_screen('[info] Video description is already present') + else: + try: + self.to_screen('[info] Writing video description to: ' + descfn) + with io.open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile: + descfile.write(info_dict['description']) + except (KeyError, TypeError): + self.report_warning('There\'s no description to write.') + except (OSError, IOError): + self.report_error('Cannot write description file ' + descfn) + return 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']) - 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 + annofn = filename + '.annotations.xml' + if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(annofn)): + self.to_screen('[info] Video annotations are already present') + else: + try: + self.to_screen('[info] Writing video annotations to: ' + annofn) + with io.open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile: + annofile.write(info_dict['annotations']) + except (KeyError, TypeError): + self.report_warning('There are no annotations to write.') + except (OSError, IOError): + self.report_error('Cannot write annotations file: ' + annofn) + return subtitles_are_requested = any([self.params.get('writesubtitles', False), self.params.get('writeautomaticsub')]) @@ -736,88 +942,144 @@ class YoutubeDL(object): continue try: sub_filename = subtitles_filename(filename, sub_lang, sub_format) - self.report_writesubtitles(sub_filename) - with io.open(encodeFilename(sub_filename), 'w', encoding='utf-8') as subfile: - subfile.write(sub) + if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(sub_filename)): + self.to_screen('[info] Video subtitle %s.%s is already_present' % (sub_lang, sub_format)) + else: + self.to_screen('[info] Writing video subtitles to: ' + sub_filename) + with io.open(encodeFilename(sub_filename), 'w', encoding='utf-8') as subfile: + subfile.write(sub) except (OSError, IOError): - self.report_error(u'Cannot write subtitles file ' + descfn) + self.report_error('Cannot write subtitles file ' + sub_filename) return if self.params.get('writeinfojson', False): - 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']) - write_json_file(json_info_dict, encodeFilename(infofn)) - except (OSError, IOError): - self.report_error(u'Cannot write metadata to JSON file ' + infofn) - return + infofn = os.path.splitext(filename)[0] + '.info.json' + if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(infofn)): + self.to_screen('[info] Video description metadata is already present') + else: + self.to_screen('[info] Writing video description metadata as JSON to: ' + infofn) + try: + write_json_file(info_dict, encodeFilename(infofn)) + except (OSError, IOError): + self.report_error('Cannot write metadata to JSON file ' + infofn) + return if self.params.get('writethumbnail', False): if info_dict.get('thumbnail') is not None: - thumb_format = determine_ext(info_dict['thumbnail'], u'jpg') - thumb_filename = filename.rpartition('.')[0] + u'.' + thumb_format - self.to_screen(u'[%s] %s: Downloading thumbnail ...' % - (info_dict['extractor'], info_dict['id'])) - try: - uf = compat_urllib_request.urlopen(info_dict['thumbnail']) - with open(thumb_filename, 'wb') as thumbf: - shutil.copyfileobj(uf, thumbf) - self.to_screen(u'[%s] %s: Writing thumbnail to: %s' % - (info_dict['extractor'], info_dict['id'], thumb_filename)) - except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err: - self.report_warning(u'Unable to download thumbnail "%s": %s' % - (info_dict['thumbnail'], compat_str(err))) + thumb_format = determine_ext(info_dict['thumbnail'], 'jpg') + thumb_filename = os.path.splitext(filename)[0] + '.' + thumb_format + if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(thumb_filename)): + self.to_screen('[%s] %s: Thumbnail is already present' % + (info_dict['extractor'], info_dict['id'])) + else: + self.to_screen('[%s] %s: Downloading thumbnail ...' % + (info_dict['extractor'], info_dict['id'])) + try: + uf = self.urlopen(info_dict['thumbnail']) + with open(thumb_filename, 'wb') as thumbf: + shutil.copyfileobj(uf, thumbf) + self.to_screen('[%s] %s: Writing thumbnail to: %s' % + (info_dict['extractor'], info_dict['id'], thumb_filename)) + except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err: + self.report_warning('Unable to download thumbnail "%s": %s' % + (info_dict['thumbnail'], compat_str(err))) if not self.params.get('skip_download', False): if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(filename)): success = True else: try: - success = self.fd._do_download(filename, info_dict) + def dl(name, info): + fd = get_suitable_downloader(info)(self, self.params) + for ph in self._progress_hooks: + fd.add_progress_hook(ph) + return fd.download(name, info) + if info_dict.get('requested_formats') is not None: + downloaded = [] + success = True + merger = FFmpegMergerPP(self) + if not merger._get_executable(): + postprocessors = [] + self.report_warning('You have requested multiple ' + 'formats but ffmpeg or avconv are not installed.' + ' The formats won\'t be merged') + else: + postprocessors = [merger] + for f in info_dict['requested_formats']: + new_info = dict(info_dict) + new_info.update(f) + fname = self.prepare_filename(new_info) + fname = prepend_extension(fname, 'f%s' % f['format_id']) + downloaded.append(fname) + partial_success = dl(fname, new_info) + success = success and partial_success + info_dict['__postprocessors'] = postprocessors + info_dict['__files_to_merge'] = downloaded + else: + # Just a single file + success = dl(filename, info_dict) except (compat_urllib_error.URLError, compat_http_client.HTTPException, socket.error) as err: - self.report_error(u'unable to download video data: %s' % str(err)) + self.report_error('unable to download video data: %s' % str(err)) return except (OSError, IOError) as err: raise UnavailableVideoError(err) except (ContentTooShortError, ) as err: - self.report_error(u'content too short (expected %s bytes and served %s)' % (err.expected, err.downloaded)) + self.report_error('content too short (expected %s bytes and served %s)' % (err.expected, err.downloaded)) return if success: try: self.post_process(filename, info_dict) except (PostProcessingError) as err: - self.report_error(u'postprocessing: %s' % str(err)) + self.report_error('postprocessing: %s' % str(err)) return self.record_download_archive(info_dict) def download(self, url_list): """Download a given list of URLs.""" + outtmpl = self.params.get('outtmpl', DEFAULT_OUTTMPL) if (len(url_list) > 1 and - '%' not in self.params['outtmpl'] + '%' not in outtmpl and self.params.get('max_downloads') != 1): - raise SameFileError(self.params['outtmpl']) + raise SameFileError(outtmpl) for url in url_list: try: #It also downloads the videos self.extract_info(url) except UnavailableVideoError: - self.report_error(u'unable to download video') + self.report_error('unable to download video') except MaxDownloadsReached: - self.to_screen(u'[info] Maximum number of downloaded files reached.') + self.to_screen('[info] Maximum number of downloaded files reached.') raise return self._download_retcode + def download_with_info_file(self, info_filename): + with io.open(info_filename, 'r', encoding='utf-8') as f: + info = json.load(f) + try: + self.process_ie_result(info, download=True) + except DownloadError: + webpage_url = info.get('webpage_url') + if webpage_url is not None: + self.report_warning('The info failed to download, trying with "%s"' % webpage_url) + return self.download([webpage_url]) + else: + raise + return self._download_retcode + def post_process(self, filename, ie_info): """Run all the postprocessors on the given file.""" info = dict(ie_info) info['filepath'] = filename keep_video = None - for pp in self._pps: + pps_chain = [] + if ie_info.get('__postprocessors') is not None: + pps_chain.extend(ie_info['__postprocessors']) + pps_chain.extend(self._pps) + for pp in pps_chain: try: keep_video_wish, new_info = pp.run(info) if keep_video_wish is not None: @@ -830,10 +1092,10 @@ class YoutubeDL(object): self.report_error(e.msg) if keep_video is False and not self.params.get('keepvideo', False): try: - self.to_screen(u'Deleting original file %s (pass -k to keep)' % filename) + self.to_screen('Deleting original file %s (pass -k to keep)' % filename) os.remove(encodeFilename(filename)) except (IOError, OSError): - self.report_warning(u'Unable to remove downloaded video file') + self.report_warning('Unable to remove downloaded video file') def _make_archive_id(self, info_dict): # Future-proof against any change in case @@ -844,7 +1106,7 @@ class YoutubeDL(object): extractor = info_dict.get('ie_key') # key in a playlist if extractor is None: return None # Incomplete video information - return extractor.lower() + u' ' + info_dict['id'] + return extractor.lower() + ' ' + info_dict['id'] def in_download_archive(self, info_dict): fn = self.params.get('download_archive') @@ -872,81 +1134,110 @@ class YoutubeDL(object): vid_id = self._make_archive_id(info_dict) assert vid_id with locked_file(fn, 'a', encoding='utf-8') as archive_file: - archive_file.write(vid_id + u'\n') + archive_file.write(vid_id + '\n') @staticmethod def format_resolution(format, default='unknown'): if format.get('vcodec') == 'none': return 'audio only' - if format.get('_resolution') is not None: - return format['_resolution'] + 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']) + res = '%sx%s' % (format['width'], format['height']) else: - res = u'%sp' % format['height'] + res = '%sp' % format['height'] + elif format.get('width') is not None: + res = '?x%d' % format['width'] else: res = default return res - def list_formats(self, info_dict): - 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 and - fdict.get('vcodec') != 'none'): - res += u'%-5s' % fdict['vcodec'] - elif fdict.get('vbr') is not None: - res += u'video' + def _format_note(self, fdict): + res = '' + if fdict.get('ext') in ['f4f', 'f4m']: + res += '(unsupported) ' + if fdict.get('format_note') is not None: + res += fdict['format_note'] + ' ' + if fdict.get('tbr') is not None: + res += '%4dk ' % fdict['tbr'] + if fdict.get('container') is not None: + if res: + res += ', ' + res += '%s container' % fdict['container'] + if (fdict.get('vcodec') is not None and + fdict.get('vcodec') != 'none'): + if res: + res += ', ' + res += fdict['vcodec'] 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 + res += '@' + elif fdict.get('vbr') is not None and fdict.get('abr') is not None: + res += 'video@' + if fdict.get('vbr') is not None: + res += '%4dk' % fdict['vbr'] + if fdict.get('acodec') is not None: + if res: + res += ', ' + if fdict['acodec'] == 'none': + res += 'video only' + else: + res += '%-5s' % fdict['acodec'] + elif fdict.get('abr') is not None: + if res: + res += ', ' + res += 'audio' + if fdict.get('abr') is not None: + res += '@%3dk' % fdict['abr'] + if fdict.get('asr') is not None: + res += ' (%5dHz)' % fdict['asr'] + if fdict.get('filesize') is not None: + if res: + res += ', ' + res += format_bytes(fdict['filesize']) + return res + def list_formats(self, info_dict): def line(format, idlen=20): - return ((u'%-' + compat_str(idlen + 1) + u's%-10s%-12s%s') % ( + return (('%-' + compat_str(idlen + 1) + 's%-10s%-12s%s') % ( format['format_id'], format['ext'], self.format_resolution(format), - format_note(format), + self._format_note(format), )) formats = info_dict.get('formats', [info_dict]) - idlen = max(len(u'format code'), + idlen = max(len('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)' + formats_s[0] += (' ' if self._format_note(formats[0]) else '') + '(worst)' + formats_s[-1] += (' ' if self._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))) + 'format_id': 'format code', 'ext': 'extension', + 'resolution': 'resolution', 'format_note': 'note'}, idlen=idlen) + self.to_screen('[info] Available formats for %s:\n%s\n%s' % + (info_dict['id'], header_line, '\n'.join(formats_s))) def urlopen(self, req): """ Start an HTTP download """ - return self._opener.open(req) + return self._opener.open(req, timeout=self._socket_timeout) def print_debug_header(self): if not self.params.get('verbose'): return - write_string(u'[debug] youtube-dl version ' + __version__ + u'\n') + + write_string( + '[debug] Encodings: locale %s, fs %s, out %s, pref %s\n' % ( + locale.getpreferredencoding(), + sys.getfilesystemencoding(), + sys.stdout.encoding, + self.get_encoding()), + encoding=None + ) + + self._write_string('[debug] youtube-dl version ' + __version__ + '\n') try: sp = subprocess.Popen( ['git', 'rev-parse', '--short', 'HEAD'], @@ -955,24 +1246,24 @@ class YoutubeDL(object): out, err = sp.communicate() out = out.decode().strip() if re.match('[0-9a-f]+', out): - write_string(u'[debug] Git HEAD: ' + out + u'\n') + self._write_string('[debug] Git HEAD: ' + out + '\n') except: try: sys.exc_clear() except: pass - write_string(u'[debug] Python version %s - %s' % - (platform.python_version(), platform_name()) + u'\n') + self._write_string('[debug] Python version %s - %s' % + (platform.python_version(), platform_name()) + '\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') + self._write_string('[debug] Proxy map: ' + compat_str(proxy_map) + '\n') def _setup_opener(self): timeout_val = self.params.get('socket_timeout') - timeout = 600 if timeout_val is None else float(timeout_val) + self._socket_timeout = 600 if timeout_val is None else float(timeout_val) opts_cookiefile = self.params.get('cookiefile') opts_proxy = self.params.get('proxy') @@ -998,16 +1289,31 @@ class YoutubeDL(object): if 'http' in proxies and 'https' not in proxies: proxies['https'] = proxies['http'] proxy_handler = compat_urllib_request.ProxyHandler(proxies) + + debuglevel = 1 if self.params.get('debug_printtraffic') else 0 https_handler = make_HTTPS_handler( - self.params.get('nocheckcertificate', False)) + self.params.get('nocheckcertificate', False), debuglevel=debuglevel) + ydlh = YoutubeDLHandler(debuglevel=debuglevel) opener = compat_urllib_request.build_opener( - https_handler, proxy_handler, cookie_processor, YoutubeDLHandler()) + https_handler, proxy_handler, cookie_processor, ydlh) # 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) + def encode(self, s): + if isinstance(s, bytes): + return s # Already encoded + + try: + return s.encode(self.get_encoding()) + except UnicodeEncodeError as err: + err.reason = err.reason + '. Check your system encoding configuration or use the --encoding option.' + raise + + def get_encoding(self): + encoding = self.params.get('encoding') + if encoding is None: + encoding = preferredencoding() + return encoding