X-Git-Url: http://git.bitcoin.ninja/index.cgi?a=blobdiff_plain;f=youtube_dl%2FYoutubeDL.py;h=20eed96caf6dbb824973a1ba7566d5131a5f4334;hb=6febd1c1df02870bf8b529ca758d7155cf65ab35;hp=f22a8bd0e044b9c10ecad56187a15a310a4c1d7d;hpb=1c1218fefcbabb49a0d06fc45d64a299920460ec;p=youtube-dl diff --git a/youtube_dl/YoutubeDL.py b/youtube_dl/YoutubeDL.py index f22a8bd0e..f70c356c0 100644 --- a/youtube_dl/YoutubeDL.py +++ b/youtube_dl/YoutubeDL.py @@ -1,21 +1,60 @@ #!/usr/bin/env python # -*- coding: utf-8 -*- -from __future__ import absolute_import +from __future__ import absolute_import, unicode_literals +import collections 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_str, + compat_urllib_error, + compat_urllib_request, + ContentTooShortError, + date_from_str, + DateRange, + determine_ext, + DownloadError, + encodeFilename, + ExtractorError, + format_bytes, + formatSeconds, + get_term_width, + locked_file, + make_HTTPS_handler, + MaxDownloadsReached, + PostProcessingError, + platform_name, + preferredencoding, + SameFileError, + sanitize_filename, + subtitles_filename, + takewhile_inclusive, + UnavailableVideoError, + url_basename, + write_json_file, + write_string, + YoutubeDLHandler, +) from .extractor import get_info_extractor, gen_extractors -from .FileDownloader import FileDownloader +from .downloader import get_suitable_downloader +from .version import __version__ class YoutubeDL(object): @@ -57,6 +96,8 @@ class YoutubeDL(object): 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. format_limit: Highest quality format to try. @@ -68,6 +109,7 @@ 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 @@ -88,10 +130,26 @@ 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. - downloadarchive: 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 + 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 + The following parameters are not used by YoutubeDL itself, they are used by the FileDownloader: nopart, updatetime, buffersize, ratelimit, min_filesize, max_filesize, test, @@ -105,8 +163,10 @@ class YoutubeDL(object): _num_downloads = None _screen_file = None - def __init__(self, params): + 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 = [] @@ -114,22 +174,50 @@ class YoutubeDL(object): self._download_retcode = 0 self._num_downloads = 0 self._screen_file = [sys.stdout, sys.stderr][params.get('logtostderr', False)] + 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.') - params['restrictfilenames'] = True + '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.params = params - self.fd = FileDownloader(self, self.params) + if '%(stitle)s' in self.params.get('outtmpl', ''): + self.report_warning('%(stitle)s is deprecated. Use the %(title)s and the --restrict-filenames flag(which also secures %(uploader)s et al) instead.') - 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.""" @@ -161,24 +249,81 @@ 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.""" - if not self.params.get('quiet', False): - terminator = [u'\n', u''][skip_eol] + return self.to_stdout(message, skip_eol, check_quiet=True) + + 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 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) 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) + assert type(message) == type('') + if self.params.get('logger'): + self.params['logger'].error(message) + else: + message = self._bidi_workaround(message) + output = message + '\n' + write_string(output, self._err_file) - def fixed_template(self): - """Checks if the output template is fixed.""" - return (re.search(u'(?u)%\\(.+?\\)s', self.params['outtmpl']) is None) + 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('\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('\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('\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 trouble(self, message=None, tb=None): """Determine action to take when a download problem appears. @@ -194,13 +339,13 @@ 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]: @@ -215,11 +360,11 @@ 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._err_file.isatty() and os.name != 'nt': + _msg_header = '\033[0;33mWARNING:\033[0m' else: - _msg_header=u'WARNING:' - warning_message=u'%s %s' % (_msg_header,message) + _msg_header = 'WARNING:' + warning_message = '%s %s' % (_msg_header, message) self.to_stderr(warning_message) def report_error(self, message, tb=None): @@ -227,48 +372,19 @@ class YoutubeDL(object): 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 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) - - 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) - except (UnicodeEncodeError) as err: - self.to_screen(u'[download] The file has already been downloaded') + self.to_screen('[download] %s has already been downloaded' % file_name) + except UnicodeEncodeError: + self.to_screen('[download] The file has already been downloaded') def increment_downloads(self): """Increment the ordinal that assigns a number to each file.""" @@ -283,59 +399,77 @@ 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['playlist_index'] is not None: - template_dict['playlist_index'] = u'%05d' % template_dict['playlist_index'] + if template_dict.get('playlist_index') is not None: + template_dict['playlist_index'] = '%05d' % template_dict['playlist_index'] - sanitize = lambda k,v: sanitize_filename( - u'NA' if v is None else compat_str(v), + sanitize = lambda k, v: sanitize_filename( + 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()) - - filename = self.params['outtmpl'] % template_dict + is_id=(k == 'id')) + template_dict = dict((k, sanitize(k, v)) + 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']) + 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 """ - 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 + '"' + 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 '"' + title + '" title did not match pattern "' + matchtitle + '"' + rejecttitle = self.params.get('rejecttitle', False) + if rejecttitle: + if re.search(rejecttitle, title, re.IGNORECASE): + 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'%(title)s has already been recorded in archive' - % info_dict) + return '%s has already been recorded in archive' % video_title return None - - def extract_info(self, url, download=True, ie_key=None, extra_info={}): + + @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={}, + process=True): ''' 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: @@ -346,8 +480,8 @@ 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) @@ -355,17 +489,21 @@ 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, } + self.add_extra_info(ie_result, + { + 'extractor': ie.IE_NAME, + 'webpage_url': url, + 'webpage_url_basename': url_basename(url), + 'extractor_key': ie.ie_key(), + }) + if process: + return self.process_ie_result(ie_result, download, extra_info) 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) + return ie_result except ExtractorError as de: # An error we somewhat expected self.report_error(compat_str(de), de.format_traceback()) break @@ -376,8 +514,8 @@ class YoutubeDL(object): else: raise else: - self.report_error(u'no suitable InfoExtractor: %s' % url) - + self.report_error('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 +527,8 @@ 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 @@ -398,38 +536,69 @@ 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 + entries = ie_result['entries'][playliststart:playlistend] n_entries = len(entries) - self.to_screen(u"[%s] playlist '%s': Collected %d video ids (downloading %d of them)" % + self.to_screen( + "[%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('[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'], + '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('[download] ' + reason) + continue + entry_result = self.process_ie_result(entry, download=download, extra_info=extra) @@ -438,16 +607,38 @@ 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'], + 'webpage_url_basename': url_basename(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 = ['mp4', 'flv', 'webm', '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 +648,9 @@ 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 ['Youku']: + if download: + self.process_info(info_dict) return info_dict # We now pick which formats have to be downloaded @@ -470,59 +662,59 @@ 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'] = '{id} - {res}{note}'.format( + id=format['format_id'], + res=self.format_resolution(format), + 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']) + format_limit = self.params.get('format_limit', None) + if format_limit: + formats = list(takewhile_inclusive( + lambda f: f['format_id'] != format_limit, formats + )) + + # 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 - format_limit = self.params.get('format_limit', None) - if format_limit: - formats = [f for f in formats if f['format_id'] <= format_limit] - 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') + 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 - # the first that is availble, starting from left + # 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('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) @@ -540,7 +732,7 @@ class YoutubeDL(object): 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'] @@ -550,7 +742,7 @@ 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') @@ -562,20 +754,25 @@ class YoutubeDL(object): # Forced printings if self.params.get('forcetitle', False): - compat_print(info_dict['title']) + 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'')) - if self.params.get('forcethumbnail', False) and 'thumbnail' in info_dict: - compat_print(info_dict['thumbnail']) - if self.params.get('forcedescription', False) and 'description' in info_dict: - compat_print(info_dict['description']) + 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: + self.to_stdout(info_dict['thumbnail']) + if self.params.get('forcedescription', False) and info_dict.get('description') is not None: + 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): + info_dict['_filename'] = filename + self.to_stdout(json.dumps(info_dict)) # Do nothing else if in simulate mode if self.params.get('simulate', False): @@ -589,121 +786,155 @@ class YoutubeDL(object): 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')]) - 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: 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 ' + descfn) return if self.params.get('writeinfojson', False): - infofn = filename + 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 = compat_urllib_request.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) + fd = get_suitable_downloader(info_dict)(self, self.params) + for ph in self._progress_hooks: + fd.add_progress_hook(ph) + success = fd.download(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.""" - if len(url_list) > 1 and self.fixed_template(): + if (len(url_list) > 1 and + '%' not in self.params['outtmpl'] + and self.params.get('max_downloads') != 1): raise SameFileError(self.params['outtmpl']) 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') + 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) @@ -711,7 +942,7 @@ 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 @@ -722,16 +953,31 @@ 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 + # and backwards compatibility with prior versions + extractor = info_dict.get('extractor_key') + if extractor is None: + if 'id' in info_dict: + extractor = info_dict.get('ie_key') # key in a playlist + if extractor is None: + return None # Incomplete video information + return extractor.lower() + ' ' + info_dict['id'] def in_download_archive(self, info_dict): fn = self.params.get('download_archive') if fn is None: return False - vid_id = info_dict['extractor'] + u' ' + info_dict['id'] + + vid_id = self._make_archive_id(info_dict) + if vid_id is None: + return False # Incomplete video information + try: with locked_file(fn, 'r', encoding='utf-8') as archive_file: for line in archive_file: @@ -746,20 +992,154 @@ class YoutubeDL(object): fn = self.params.get('download_archive') if fn is None: return - vid_id = info_dict['extractor'] + u' ' + info_dict['id'] + 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('height') is not None: + if format.get('width') is not None: + res = '%sx%s' % (format['width'], format['height']) + else: + 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): - 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 = '' + 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('vcodec') is not None and + fdict.get('vcodec') != 'none'): + res += '%-5s@' % fdict['vcodec'] + 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 += ', ' + 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('filesize') is not None: + if res: + res += ', ' + res += format_bytes(fdict['filesize']) + return res + + def line(format, idlen=20): + return (('%-' + compat_str(idlen + 1) + '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('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': '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) + + def print_debug_header(self): + if not self.params.get('verbose'): + return + write_string('[debug] youtube-dl version ' + __version__ + '\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('[debug] Git HEAD: ' + out + '\n') + except: + try: + sys.exc_clear() + except: + pass + 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('[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) + + 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) + + debuglevel = 1 if self.params.get('debug_printtraffic') else 0 + https_handler = make_HTTPS_handler( + self.params.get('nocheckcertificate', False), debuglevel=debuglevel) + ydlh = YoutubeDLHandler(debuglevel=debuglevel) + opener = compat_urllib_request.build_opener( + 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)