X-Git-Url: https://git.rapsys.eu/youtubedl/blobdiff_plain/bc5059672e222ec4df4fe3983c399c0d0fca84d5..1f17a37b9b95db09a420a1f52cf18723ce4eb8b5:/youtube_dl/YoutubeDL.py diff --git a/youtube_dl/YoutubeDL.py b/youtube_dl/YoutubeDL.py old mode 100644 new mode 100755 index b68b110..97bd9c5 --- a/youtube_dl/YoutubeDL.py +++ b/youtube_dl/YoutubeDL.py @@ -1,11 +1,19 @@ #!/usr/bin/env python -# -*- coding: utf-8 -*- +# coding: utf-8 -from __future__ import absolute_import +from __future__ import absolute_import, unicode_literals +import collections +import contextlib +import copy +import datetime import errno +import fileinput import io +import itertools import json +import locale +import operator import os import platform import re @@ -14,45 +22,93 @@ import subprocess import socket import sys import time +import tokenize import traceback +import random -if os.name == 'nt': - import ctypes +from string import ascii_letters -from .utils import ( +from .compat import ( + compat_basestring, compat_cookiejar, + compat_get_terminal_size, compat_http_client, - compat_print, + compat_kwargs, + compat_numeric_types, + compat_os_name, compat_str, + compat_tokenize_tokenize, compat_urllib_error, compat_urllib_request, + compat_urllib_request_DataHandler, +) +from .utils import ( + age_restricted, + args_to_str, ContentTooShortError, date_from_str, DateRange, + DEFAULT_OUTTMPL, determine_ext, + determine_protocol, DownloadError, + encode_compat_str, encodeFilename, + error_to_compat_str, + expand_path, ExtractorError, format_bytes, + formatSeconds, + GeoRestrictedError, + int_or_none, + ISO3166Utils, locked_file, make_HTTPS_handler, MaxDownloadsReached, - PostProcessingError, + orderedSet, + PagedList, + parse_filesize, + PerRequestProxyHandler, platform_name, + PostProcessingError, preferredencoding, + prepend_extension, + register_socks_protocols, + render_table, + replace_extension, SameFileError, sanitize_filename, + sanitize_path, + sanitize_url, + sanitized_Request, + std_headers, subtitles_filename, - takewhile_inclusive, UnavailableVideoError, + url_basename, + version_tuple, write_json_file, write_string, + YoutubeDLCookieProcessor, YoutubeDLHandler, ) -from .extractor import get_info_extractor, gen_extractors -from .FileDownloader import FileDownloader +from .cache import Cache +from .extractor import get_info_extractor, gen_extractor_classes, _LAZY_LOADER +from .extractor.openload import PhantomJSwrapper +from .downloader import get_suitable_downloader +from .downloader.rtmp import rtmpdump_version +from .postprocessor import ( + FFmpegFixupM3u8PP, + FFmpegFixupM4aPP, + FFmpegFixupStretchedPP, + FFmpegMergerPP, + FFmpegPostProcessor, + get_postprocessor, +) from .version import __version__ +if compat_os_name == 'nt': + import ctypes + class YoutubeDL(object): """YoutubeDL class. @@ -83,26 +139,36 @@ class YoutubeDL(object): username: Username for authentication purposes. password: Password for authentication purposes. - videopassword: Password for acces a video. + videopassword: Password for accessing a video. + ap_mso: Adobe Pass multiple-system operator identifier. + ap_username: Multiple-system operator account username. + ap_password: Multiple-system operator account password. 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. + dump_single_json: Force printing the info_dict of the whole playlist + (or video) as a single JSON line. simulate: Do not download the video files. - format: Video format code. - format_limit: Highest quality format to try. + format: Video format code. See options.py for more information. outtmpl: Template for output names. restrictfilenames: Do not allow "&" and spaces in file names ignoreerrors: Do not stop on download errors. + force_generic_extractor: Force downloader to use the generic extractor nooverwrites: Prevent overwriting files. playliststart: Playlist item to start at. playlistend: Playlist item to end at. + playlist_items: Specific indices of playlist to download. + playlistreverse: Download playlist items in reverse order. + playlistrandom: Download playlist items in random order. matchtitle: Download only matching titles. rejecttitle: Reject downloads for matching titles. logger: Log messages to a logging.Logger instance. @@ -111,35 +177,153 @@ class YoutubeDL(object): writeinfojson: Write the video description to a .info.json file writeannotations: Write the video annotations to a .annotations.xml file writethumbnail: Write the thumbnail image to a file + write_all_thumbnails: Write all thumbnail formats to files writesubtitles: Write the video subtitles to a file - writeautomaticsub: Write the automatic subtitles to a file + writeautomaticsub: Write the automatically generated subtitles to a file allsubtitles: Downloads all the subtitles of the video (requires writesubtitles or writeautomaticsub) listsubtitles: Lists all available subtitles for the video - subtitlesformat: Subtitle format [srt/sbv/vtt] (default=srt) + subtitlesformat: The format code for subtitles subtitleslangs: List of languages of the subtitles to download keepvideo: Keep the video file after post-processing daterange: A DateRange object, download only if the upload_date is in the range. skip_download: Skip the actual download of the video file cachedir: Location of the cache files in the filesystem. - None to disable filesystem cache. + False to disable filesystem cache. 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 + geo_verification_proxy: URL of the proxy to use for IP address verification + on geo-restricted sites. (Experimental) 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. + extract_flat: Do not resolve URLs, return the immediate result. + Pass in 'in_playlist' to only show this behavior for + playlist items. + postprocessors: A list of dictionaries, each with an entry + * key: The name of the postprocessor. See + youtube_dl/postprocessor/__init__.py for a list. + as well as any further keyword arguments for the + postprocessor. + progress_hooks: A list of functions that get called on download + progress, with a dictionary with the entries + * status: One of "downloading", "error", or "finished". + Check this first and ignore unknown values. + + If status is one of "downloading", or "finished", the + following properties may also be present: + * filename: The final filename (always present) + * tmpfilename: The filename we're currently writing to + * downloaded_bytes: Bytes on disk + * total_bytes: Size of the whole file, None if unknown + * total_bytes_estimate: Guess of the eventual file size, + None if unavailable. + * elapsed: The number of seconds since download started. + * eta: The estimated time in seconds, None if unknown + * speed: The download speed in bytes/second, None if + unknown + * fragment_index: The counter of the currently + downloaded video fragment. + * fragment_count: The number of fragments (= individual + files that will be merged) + + Progress hooks are guaranteed to be called at least once + (with status "finished") if the download is successful. + merge_output_format: Extension to use when merging formats. + fixup: Automatically correct known faults of the file. + One of: + - "never": do nothing + - "warn": only emit a warning + - "detect_or_warn": check whether we can do anything + about it, warn otherwise (default) + source_address: (Experimental) Client-side IP address to bind to. + call_home: Boolean, true iff we are allowed to contact the + youtube-dl servers for debugging. + sleep_interval: Number of seconds to sleep before each download when + used alone or a lower bound of a range for randomized + sleep before each download (minimum possible number + of seconds to sleep) when used along with + max_sleep_interval. + max_sleep_interval:Upper bound of a range for randomized sleep before each + download (maximum possible number of seconds to sleep). + Must only be used along with sleep_interval. + Actual sleep time will be a random float from range + [sleep_interval; max_sleep_interval]. + listformats: Print an overview of available video formats and exit. + list_thumbnails: Print a table of all thumbnails and exit. + match_filter: A function that gets called with the info_dict of + every video. + If it returns a message, the video is ignored. + If it returns None, the video is downloaded. + match_filter_func in utils.py is one example for this. + no_color: Do not emit color codes in output. + geo_bypass: Bypass geographic restriction via faking X-Forwarded-For + HTTP header (experimental) + geo_bypass_country: + Two-letter ISO 3166-2 country code that will be used for + explicit geographic restriction bypassing via faking + X-Forwarded-For HTTP header (experimental) + + The following options determine which downloader is picked: + external_downloader: Executable of the external downloader to call. + None or unset for standard (built-in) downloader. + hls_prefer_native: Use the native HLS downloader instead of ffmpeg/avconv + if True, otherwise use ffmpeg/avconv if False, otherwise + use downloader suggested by extractor if None. The following parameters are not used by YoutubeDL itself, they are used by - the FileDownloader: + the downloader (see youtube_dl/downloader/common.py): nopart, updatetime, buffersize, ratelimit, min_filesize, max_filesize, test, - noresizebuffer, retries, continuedl, noprogress, consoletitle + noresizebuffer, retries, continuedl, noprogress, consoletitle, + xattr_set_filesize, external_downloader_args, hls_use_mpegts. + + 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. + postprocessor_args: A list of additional command-line arguments for the + postprocessor. + + The following options are used by the Youtube extractor: + youtube_include_dash_manifest: If True (default), DASH manifests and related + data will be downloaded and processed by extractor. + You can reduce network I/O by disabling it if you don't + care about DASH. """ + _NUMERIC_FIELDS = set(( + 'width', 'height', 'tbr', 'abr', 'asr', 'vbr', 'fps', 'filesize', 'filesize_approx', + 'timestamp', 'upload_year', 'upload_month', 'upload_day', + 'duration', 'view_count', 'like_count', 'dislike_count', 'repost_count', + 'average_rating', 'comment_count', 'age_limit', + 'start_time', 'end_time', + 'chapter_number', 'season_number', 'episode_number', + 'track_number', 'disc_number', 'release_year', + 'playlist_index', + )) + params = None _ies = [] _pps = [] @@ -147,8 +331,10 @@ class YoutubeDL(object): _num_downloads = None _screen_file = None - def __init__(self, params=None): + def __init__(self, params=None, auto_init=True): """Create a FileDownloader object with the given options.""" + if params is None: + params = {} self._ies = [] self._ies_instances = {} self._pps = [] @@ -156,30 +342,111 @@ 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 = { + # Default parameters + 'nocheckcertificate': False, + } + self.params.update(params) + self.cache = Cache(self) + + def check_deprecated(param, option, suggestion): + if self.params.get(param) is not None: + self.report_warning( + '%s is deprecated. Use %s instead.' % (option, suggestion)) + return True + return False - 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) + if check_deprecated('cn_verification_proxy', '--cn-verification-proxy', '--geo-verification-proxy'): + if self.params.get('geo_verification_proxy') is None: + self.params['geo_verification_proxy'] = self.params['cn_verification_proxy'] + + check_deprecated('autonumber_size', '--autonumber-size', 'output template with %(autonumber)0Nd, where N in the number of digits') + check_deprecated('autonumber', '--auto-number', '-o "%(autonumber)s-%(title)s.%(ext)s"') + check_deprecated('usetitle', '--title', '-o "%(title)s-%(id)s.%(ext)s"') + + if params.get('bidi_workaround', False): + try: + import pty + master, slave = pty.openpty() + width = compat_get_terminal_size().columns + 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 == errno.ENOENT: + 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 $PATH.') + else: + raise + + if (sys.platform != 'win32' and + sys.getfilesystemencoding() in ['ascii', 'ANSI_X3.4-1968'] and + not params.get('restrictfilenames', False)): + # Unicode filesystem API will throw errors (#1474, #13027) self.report_warning( - u'Assuming --restrict-filenames since file system encoding ' - u'cannot encode all charactes. ' - u'Set the LC_ALL environment variable to fix this.') + 'Assuming --restrict-filenames since file system encoding ' + 'cannot encode all characters. ' + 'Set the LC_ALL environment variable to fix this.') self.params['restrictfilenames'] = True - self.fd = FileDownloader(self, self.params) - - if '%(stitle)s' in self.params.get('outtmpl', ''): - self.report_warning(u'%(stitle)s is deprecated. Use the %(title)s and the --restrict-filenames flag(which also secures %(uploader)s et al) instead.') + if isinstance(params.get('outtmpl'), bytes): + self.report_warning( + 'Parameter outtmpl is bytes, but should be a unicode string. ' + 'Put from __future__ import unicode_literals at the top of your code file or consider switching to Python 3.x.') self._setup_opener() + if auto_init: + self.print_debug_header() + self.add_default_info_extractors() + + for pp_def_raw in self.params.get('postprocessors', []): + pp_class = get_postprocessor(pp_def_raw['key']) + pp_def = dict(pp_def_raw) + del pp_def['key'] + pp = pp_class(self, **compat_kwargs(pp_def)) + self.add_post_processor(pp) + + for ph in self.params.get('progress_hooks', []): + self.add_progress_hook(ph) + + register_socks_protocols() + + def warn_if_short_id(self, argv): + # short YouTube ID starting with dash? + idxs = [ + i for i, a in enumerate(argv) + if re.match(r'^-[0-9A-Za-z_-]{10}$', a)] + if idxs: + correct_argv = ( + ['youtube-dl'] + + [a for i, a in enumerate(argv) if i not in idxs] + + ['--'] + [argv[i] for i in idxs] + ) + self.report_warning( + 'Long argument string detected. ' + 'Use -- to separate parameters and URLs, like this:\n%s\n' % + args_to_str(correct_argv)) + def add_info_extractor(self, ie): """Add an InfoExtractor object to the end of the list.""" self._ies.append(ie) - self._ies_instances[ie.ie_key()] = ie - ie.set_downloader(self) + if not isinstance(ie, type): + self._ies_instances[ie.ie_key()] = ie + ie.set_downloader(self) def get_info_extractor(self, ie_key): """ @@ -197,7 +464,7 @@ class YoutubeDL(object): """ Add the InfoExtractors returned by gen_extractors to the end of the list """ - for ie in gen_extractors(): + for ie in gen_extractor_classes(): self.add_info_extractor(ie) def add_post_processor(self, pp): @@ -205,49 +472,75 @@ 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 isinstance(message, compat_str) + 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 isinstance(message, compat_str) 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): 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)) + if compat_os_name == 'nt': + if ctypes.windll.kernel32.GetConsoleWindow(): + # c_wchar_p() might not be necessary if `message` is + # already of type unicode() + ctypes.windll.kernel32.SetConsoleTitleW(ctypes.c_wchar_p(message)) elif 'TERM' in os.environ: - write_string(u'\033]0;%s\007' % message, self._screen_file) + 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: + if compat_os_name != 'nt' and '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: + if compat_os_name != 'nt' and '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 +548,7 @@ class YoutubeDL(object): def __exit__(self, *args): self.restore_console_title() - + if self.params.get('cookiefile') is not None: self.cookiejar.save() @@ -273,13 +566,13 @@ 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 += compat_str(traceback.format_exc()) + tb += ''.join(traceback.format_exception(*sys.exc_info()[1].exc_info)) + tb += encode_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 +587,36 @@ 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 not self.params.get('no_color') and self._err_file.isatty() and compat_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 not self.params.get('no_color') and self._err_file.isatty() and compat_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 +627,126 @@ 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' - 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['autonumber'] = self.params.get('autonumber_start', 1) - 1 + self._num_downloads + 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'] = '%dx?' % 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')) - template_dict = dict((k, sanitize(k, v)) - for k, v in template_dict.items()) - - 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 + is_id=(k == 'id' or k.endswith('_id'))) + template_dict = dict((k, v if isinstance(v, compat_numeric_types) else sanitize(k, v)) + for k, v in template_dict.items() + if v is not None and not isinstance(v, (list, tuple, dict))) + template_dict = collections.defaultdict(lambda: 'NA', template_dict) + + outtmpl = self.params.get('outtmpl', DEFAULT_OUTTMPL) + + # For fields playlist_index and autonumber convert all occurrences + # of %(field)s to %(field)0Nd for backward compatibility + field_size_compat_map = { + 'playlist_index': len(str(template_dict['n_entries'])), + 'autonumber': autonumber_size, + } + FIELD_SIZE_COMPAT_RE = r'(?autonumber|playlist_index)\)s' + mobj = re.search(FIELD_SIZE_COMPAT_RE, outtmpl) + if mobj: + outtmpl = re.sub( + FIELD_SIZE_COMPAT_RE, + r'%%(\1)0%dd' % field_size_compat_map[mobj.group('field')], + outtmpl) + + # Missing numeric fields used together with integer presentation types + # in format specification will break the argument substitution since + # string 'NA' is returned for missing fields. We will patch output + # template for missing fields to meet string presentation type. + for numeric_field in self._NUMERIC_FIELDS: + if numeric_field not in template_dict: + # As of [1] format syntax is: + # %[mapping_key][conversion_flags][minimum_width][.precision][length_modifier]type + # 1. https://docs.python.org/2/library/stdtypes.html#string-formatting + FORMAT_RE = r'''(?x) + (? max_views: + return 'Skipping %s, because it has exceeded the maximum view count (%d/%d)' % (video_title, view_count, max_views) + if age_restricted(info_dict.get('age_limit'), self.params.get('age_limit')): + return 'Skipping "%s" because it is age restricted' % video_title 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 + + if not incomplete: + match_filter = self.params.get('match_filter') + if match_filter is not None: + ret = match_filter(info_dict) + if ret is not None: + return ret + return None @staticmethod @@ -405,12 +755,16 @@ 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, force_generic_extractor=False): ''' 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 not ie_key and force_generic_extractor: + ie_key = 'Generic' if ie_key: ies = [self.get_info_extractor(ie_key)] @@ -421,13 +775,14 @@ class YoutubeDL(object): if not ie.suitable(url): continue + ie = self.get_info_extractor(ie.ie_key()) 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) - if ie_result is None: # Finished already (backwards compatibility; listformats and friends should be moved here) + if ie_result is None: # Finished already (backwards compatibility; listformats and friends should be moved here) break if isinstance(ie_result, list): # Backwards compatibility: old IE result format @@ -435,24 +790,40 @@ 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) - except ExtractorError as de: # An error we somewhat expected - self.report_error(compat_str(de), de.format_traceback()) + 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 GeoRestrictedError as e: + msg = e.msg + if e.countries: + msg += '\nThis video is available in %s.' % ', '.join( + map(ISO3166Utils.short2full, e.countries)) + msg += '\nYou might want to use a VPN or a proxy server (with --proxy) to workaround.' + self.report_error(msg) + break + except ExtractorError as e: # An error we somewhat expected + self.report_error(compat_str(e), e.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())) + self.report_error(error_to_compat_str(e), tb=encode_compat_str(traceback.format_exc())) break 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={}): """ @@ -462,8 +833,17 @@ class YoutubeDL(object): It will also download the videos if 'download'. Returns the resolved ie_result. """ + result_type = ie_result.get('_type', 'video') + + if result_type in ('url', 'url_transparent'): + ie_result['url'] = sanitize_url(ie_result['url']) + extract_flat = self.params.get('extract_flat', False) + if ((extract_flat == 'in_playlist' and 'playlist' in extra_info) or + extract_flat is True): + if self.params.get('forcejson', False): + self.to_stdout(json.dumps(ie_result)) + return ie_result - 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': self.add_extra_info(ie_result, extra_info) return self.process_video_result(ie_result, download=download) @@ -474,41 +854,139 @@ class YoutubeDL(object): download, ie_key=ie_result.get('ie_key'), extra_info=extra_info) - elif result_type == 'playlist': - + 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) + + # extract_info may return None when ignoreerrors is enabled and + # extraction failed with an error, don't crash and return early + # in this case + if not info: + return info + + force_properties = dict( + (k, v) for k, v in ie_result.items() if v is not None) + for f in ('_type', 'url', 'id', 'extractor', 'extractor_key', 'ie_key'): + if f in force_properties: + del force_properties[f] + new_result = info.copy() + new_result.update(force_properties) + + # Extracted info may not be a video result (i.e. + # info.get('_type', 'video') != video) but rather an url or + # url_transparent. In such cases outer metadata (from ie_result) + # should be propagated to inner one (info). For this to happen + # _type of info should be overridden with url_transparent. This + # fixes issue from https://github.com/rg3/youtube-dl/pull/11163. + if new_result.get('_type') == 'url': + new_result['_type'] = 'url_transparent' + + return self.process_ie_result( + new_result, download=download, extra_info=extra_info) + elif result_type in ('playlist', 'multi_video'): # 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) + playlist = ie_result.get('title') or ie_result.get('id') + 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') + # 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 + + playlistitems_str = self.params.get('playlist_items') + playlistitems = None + if playlistitems_str is not None: + def iter_playlistitems(format): + for string_segment in format.split(','): + if '-' in string_segment: + start, end = string_segment.split('-') + for item in range(int(start), int(end) + 1): + yield int(item) + else: + yield int(string_segment) + playlistitems = orderedSet(iter_playlistitems(playlistitems_str)) + + ie_entries = ie_result['entries'] + + def make_playlistitems_entries(list_ie_entries): + num_entries = len(list_ie_entries) + return [ + list_ie_entries[i - 1] for i in playlistitems + if -num_entries <= i - 1 < num_entries] + + def report_download(num_entries): + self.to_screen( + '[%s] playlist %s: Downloading %d videos' % + (ie_result['extractor'], playlist, num_entries)) + + if isinstance(ie_entries, list): + n_all_entries = len(ie_entries) + if playlistitems: + entries = make_playlistitems_entries(ie_entries) + else: + entries = ie_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)) + elif isinstance(ie_entries, PagedList): + if playlistitems: + entries = [] + for item in playlistitems: + entries.extend(ie_entries.getslice( + item - 1, item + )) + else: + entries = ie_entries.getslice( + playliststart, playlistend) + n_entries = len(entries) + report_download(n_entries) + else: # iterable + if playlistitems: + entries = make_playlistitems_entries(list(itertools.islice( + ie_entries, 0, max(playlistitems)))) + else: + entries = list(itertools.islice( + ie_entries, playliststart, playlistend)) + n_entries = len(entries) + report_download(n_entries) - n_entries = len(entries) + if self.params.get('playlistreverse', False): + entries = entries[::-1] - 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 self.params.get('playlistrandom', False): + random.shuffle(entries) + + x_forwarded_for = ie_result.get('__x_forwarded_for_ip') 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)) + # This __x_forwarded_for_ip thing is a bit ugly but requires + # minimal changes + if x_forwarded_for: + entry['__x_forwarded_for_ip'] = x_forwarded_for extra = { + 'n_entries': n_entries, 'playlist': playlist, + 'playlist_id': ie_result.get('id'), + 'playlist_title': ie_result.get('title'), + 'playlist_uploader': ie_result.get('uploader'), + 'playlist_uploader_id': ie_result.get('uploader_id'), '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) + reason = self._match_entry(entry, incomplete=True) if reason is not None: - self.to_screen(u'[download] ' + reason) + self.to_screen('[download] ' + reason) continue entry_result = self.process_ie_result(entry, @@ -516,15 +994,23 @@ class YoutubeDL(object): extra_info=extra) playlist_results.append(entry_result) ie_result['entries'] = playlist_results + self.to_screen('[download] Finished downloading playlist: %s' % playlist) return ie_result elif result_type == 'compat_list': + self.report_warning( + 'Extractor %s returned a compat_list result. ' + 'It needs to be updated.' % ie_result.get('extractor')) + def _fixup(r): - self.add_extra_info(r, + 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, extra_info) @@ -534,35 +1020,477 @@ class YoutubeDL(object): else: raise Exception('Invalid result type: %s' % result_type) - def select_format(self, format_spec, available_formats): - if format_spec == 'best' or format_spec is None: - return available_formats[-1] - elif format_spec == 'worst': - return available_formats[0] - else: - extensions = [u'mp4', u'flv', u'webm', u'3gp'] - if format_spec in extensions: - filter_f = lambda f: f['ext'] == format_spec - else: - filter_f = lambda f: f['format_id'] == format_spec - matches = list(filter(filter_f, available_formats)) - if matches: - return matches[-1] - return None + def _build_format_filter(self, filter_spec): + " Returns a function to filter the formats according to the filter_spec " + + OPERATORS = { + '<': operator.lt, + '<=': operator.le, + '>': operator.gt, + '>=': operator.ge, + '=': operator.eq, + '!=': operator.ne, + } + operator_rex = re.compile(r'''(?x)\s* + (?Pwidth|height|tbr|abr|vbr|asr|filesize|fps) + \s*(?P%s)(?P\s*\?)?\s* + (?P[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?) + $ + ''' % '|'.join(map(re.escape, OPERATORS.keys()))) + m = operator_rex.search(filter_spec) + if m: + try: + comparison_value = int(m.group('value')) + except ValueError: + comparison_value = parse_filesize(m.group('value')) + if comparison_value is None: + comparison_value = parse_filesize(m.group('value') + 'B') + if comparison_value is None: + raise ValueError( + 'Invalid value %r in format specification %r' % ( + m.group('value'), filter_spec)) + op = OPERATORS[m.group('op')] + + if not m: + STR_OPERATORS = { + '=': operator.eq, + '!=': operator.ne, + '^=': lambda attr, value: attr.startswith(value), + '$=': lambda attr, value: attr.endswith(value), + '*=': lambda attr, value: value in attr, + } + str_operator_rex = re.compile(r'''(?x) + \s*(?Pext|acodec|vcodec|container|protocol|format_id) + \s*(?P%s)(?P\s*\?)? + \s*(?P[a-zA-Z0-9._-]+) + \s*$ + ''' % '|'.join(map(re.escape, STR_OPERATORS.keys()))) + m = str_operator_rex.search(filter_spec) + if m: + comparison_value = m.group('value') + op = STR_OPERATORS[m.group('op')] + + if not m: + raise ValueError('Invalid filter specification %r' % filter_spec) + + def _filter(f): + actual_value = f.get(m.group('key')) + if actual_value is None: + return m.group('none_inclusive') + return op(actual_value, comparison_value) + return _filter + + def _default_format_spec(self, info_dict, download=True): + + def can_merge(): + merger = FFmpegMergerPP(self) + return merger.available and merger.can_merge() + + def prefer_best(): + if self.params.get('simulate', False): + return False + if not download: + return False + if self.params.get('outtmpl', DEFAULT_OUTTMPL) == '-': + return True + if info_dict.get('is_live'): + return True + if not can_merge(): + return True + return False + + req_format_list = ['bestvideo+bestaudio', 'best'] + if prefer_best(): + req_format_list.reverse() + return '/'.join(req_format_list) + + def build_format_selector(self, format_spec): + def syntax_error(note, start): + message = ( + 'Invalid format specification: ' + '{0}\n\t{1}\n\t{2}^'.format(note, format_spec, ' ' * start[1])) + return SyntaxError(message) + + PICKFIRST = 'PICKFIRST' + MERGE = 'MERGE' + SINGLE = 'SINGLE' + GROUP = 'GROUP' + FormatSelector = collections.namedtuple('FormatSelector', ['type', 'selector', 'filters']) + + def _parse_filter(tokens): + filter_parts = [] + for type, string, start, _, _ in tokens: + if type == tokenize.OP and string == ']': + return ''.join(filter_parts) + else: + filter_parts.append(string) + + def _remove_unused_ops(tokens): + # Remove operators that we don't use and join them with the surrounding strings + # for example: 'mp4' '-' 'baseline' '-' '16x9' is converted to 'mp4-baseline-16x9' + ALLOWED_OPS = ('/', '+', ',', '(', ')') + last_string, last_start, last_end, last_line = None, None, None, None + for type, string, start, end, line in tokens: + if type == tokenize.OP and string == '[': + if last_string: + yield tokenize.NAME, last_string, last_start, last_end, last_line + last_string = None + yield type, string, start, end, line + # everything inside brackets will be handled by _parse_filter + for type, string, start, end, line in tokens: + yield type, string, start, end, line + if type == tokenize.OP and string == ']': + break + elif type == tokenize.OP and string in ALLOWED_OPS: + if last_string: + yield tokenize.NAME, last_string, last_start, last_end, last_line + last_string = None + yield type, string, start, end, line + elif type in [tokenize.NAME, tokenize.NUMBER, tokenize.OP]: + if not last_string: + last_string = string + last_start = start + last_end = end + else: + last_string += string + if last_string: + yield tokenize.NAME, last_string, last_start, last_end, last_line + + def _parse_format_selection(tokens, inside_merge=False, inside_choice=False, inside_group=False): + selectors = [] + current_selector = None + for type, string, start, _, _ in tokens: + # ENCODING is only defined in python 3.x + if type == getattr(tokenize, 'ENCODING', None): + continue + elif type in [tokenize.NAME, tokenize.NUMBER]: + current_selector = FormatSelector(SINGLE, string, []) + elif type == tokenize.OP: + if string == ')': + if not inside_group: + # ')' will be handled by the parentheses group + tokens.restore_last_token() + break + elif inside_merge and string in ['/', ',']: + tokens.restore_last_token() + break + elif inside_choice and string == ',': + tokens.restore_last_token() + break + elif string == ',': + if not current_selector: + raise syntax_error('"," must follow a format selector', start) + selectors.append(current_selector) + current_selector = None + elif string == '/': + if not current_selector: + raise syntax_error('"/" must follow a format selector', start) + first_choice = current_selector + second_choice = _parse_format_selection(tokens, inside_choice=True) + current_selector = FormatSelector(PICKFIRST, (first_choice, second_choice), []) + elif string == '[': + if not current_selector: + current_selector = FormatSelector(SINGLE, 'best', []) + format_filter = _parse_filter(tokens) + current_selector.filters.append(format_filter) + elif string == '(': + if current_selector: + raise syntax_error('Unexpected "("', start) + group = _parse_format_selection(tokens, inside_group=True) + current_selector = FormatSelector(GROUP, group, []) + elif string == '+': + video_selector = current_selector + audio_selector = _parse_format_selection(tokens, inside_merge=True) + if not video_selector or not audio_selector: + raise syntax_error('"+" must be between two format selectors', start) + current_selector = FormatSelector(MERGE, (video_selector, audio_selector), []) + else: + raise syntax_error('Operator not recognized: "{0}"'.format(string), start) + elif type == tokenize.ENDMARKER: + break + if current_selector: + selectors.append(current_selector) + return selectors + + def _build_selector_function(selector): + if isinstance(selector, list): + fs = [_build_selector_function(s) for s in selector] + + def selector_function(ctx): + for f in fs: + for format in f(ctx): + yield format + return selector_function + elif selector.type == GROUP: + selector_function = _build_selector_function(selector.selector) + elif selector.type == PICKFIRST: + fs = [_build_selector_function(s) for s in selector.selector] + + def selector_function(ctx): + for f in fs: + picked_formats = list(f(ctx)) + if picked_formats: + return picked_formats + return [] + elif selector.type == SINGLE: + format_spec = selector.selector + + def selector_function(ctx): + formats = list(ctx['formats']) + if not formats: + return + if format_spec == 'all': + for f in formats: + yield f + elif format_spec in ['best', 'worst', None]: + format_idx = 0 if format_spec == 'worst' else -1 + audiovideo_formats = [ + f for f in formats + if f.get('vcodec') != 'none' and f.get('acodec') != 'none'] + if audiovideo_formats: + yield audiovideo_formats[format_idx] + # for extractors with incomplete formats (audio only (soundcloud) + # or video only (imgur)) we will fallback to best/worst + # {video,audio}-only format + elif ctx['incomplete_formats']: + yield formats[format_idx] + elif format_spec == 'bestaudio': + audio_formats = [ + f for f in formats + if f.get('vcodec') == 'none'] + if audio_formats: + yield audio_formats[-1] + elif format_spec == 'worstaudio': + audio_formats = [ + f for f in formats + if f.get('vcodec') == 'none'] + if audio_formats: + yield audio_formats[0] + elif format_spec == 'bestvideo': + video_formats = [ + f for f in formats + if f.get('acodec') == 'none'] + if video_formats: + yield video_formats[-1] + elif format_spec == 'worstvideo': + video_formats = [ + f for f in formats + if f.get('acodec') == 'none'] + if video_formats: + yield video_formats[0] + else: + extensions = ['mp4', 'flv', 'webm', '3gp', 'm4a', 'mp3', 'ogg', 'aac', 'wav'] + 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, formats)) + if matches: + yield matches[-1] + elif selector.type == MERGE: + def _merge(formats_info): + format_1, format_2 = [f['format_id'] for f in formats_info] + # The first format must contain the video and the + # second the audio + if formats_info[0].get('vcodec') == 'none': + self.report_error('The first format must ' + 'contain the video, try using ' + '"-f %s+%s"' % (format_2, format_1)) + return + # Formats must be opposite (video+audio) + if formats_info[0].get('acodec') == 'none' and formats_info[1].get('acodec') == 'none': + self.report_error( + 'Both formats %s and %s are video-only, you must specify "-f video+audio"' + % (format_1, format_2)) + return + output_ext = ( + formats_info[0]['ext'] + if self.params.get('merge_output_format') is None + else self.params['merge_output_format']) + return { + 'requested_formats': formats_info, + 'format': '%s+%s' % (formats_info[0].get('format'), + formats_info[1].get('format')), + 'format_id': '%s+%s' % (formats_info[0].get('format_id'), + formats_info[1].get('format_id')), + 'width': formats_info[0].get('width'), + 'height': formats_info[0].get('height'), + 'resolution': formats_info[0].get('resolution'), + 'fps': formats_info[0].get('fps'), + 'vcodec': formats_info[0].get('vcodec'), + 'vbr': formats_info[0].get('vbr'), + 'stretched_ratio': formats_info[0].get('stretched_ratio'), + 'acodec': formats_info[1].get('acodec'), + 'abr': formats_info[1].get('abr'), + 'ext': output_ext, + } + video_selector, audio_selector = map(_build_selector_function, selector.selector) + + def selector_function(ctx): + for pair in itertools.product( + video_selector(copy.deepcopy(ctx)), audio_selector(copy.deepcopy(ctx))): + yield _merge(pair) + + filters = [self._build_format_filter(f) for f in selector.filters] + + def final_selector(ctx): + ctx_copy = copy.deepcopy(ctx) + for _filter in filters: + ctx_copy['formats'] = list(filter(_filter, ctx_copy['formats'])) + return selector_function(ctx_copy) + return final_selector + + stream = io.BytesIO(format_spec.encode('utf-8')) + try: + tokens = list(_remove_unused_ops(compat_tokenize_tokenize(stream.readline))) + except tokenize.TokenError: + raise syntax_error('Missing closing/opening brackets or parenthesis', (0, len(format_spec))) + + class TokenIterator(object): + def __init__(self, tokens): + self.tokens = tokens + self.counter = 0 + + def __iter__(self): + return self + + def __next__(self): + if self.counter >= len(self.tokens): + raise StopIteration() + value = self.tokens[self.counter] + self.counter += 1 + return value + + next = __next__ + + def restore_last_token(self): + self.counter -= 1 + + parsed_selector = _parse_format_selection(iter(TokenIterator(tokens))) + return _build_selector_function(parsed_selector) + + def _calc_headers(self, info_dict): + res = std_headers.copy() + + add_headers = info_dict.get('http_headers') + if add_headers: + res.update(add_headers) + + cookies = self._calc_cookies(info_dict) + if cookies: + res['Cookie'] = cookies + + if 'X-Forwarded-For' not in res: + x_forwarded_for_ip = info_dict.get('__x_forwarded_for_ip') + if x_forwarded_for_ip: + res['X-Forwarded-For'] = x_forwarded_for_ip + + return res + + def _calc_cookies(self, info_dict): + pr = sanitized_Request(info_dict['url']) + self.cookiejar.add_cookie_header(pr) + return pr.get_header('Cookie') 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') + + def report_force_conversion(field, field_not, conversion): + self.report_warning( + '"%s" field is not %s - forcing %s conversion, there is an error in extractor' + % (field, field_not, conversion)) + + def sanitize_string_field(info, string_field): + field = info.get(string_field) + if field is None or isinstance(field, compat_str): + return + report_force_conversion(string_field, 'a string', 'string') + info[string_field] = compat_str(field) + + def sanitize_numeric_fields(info): + for numeric_field in self._NUMERIC_FIELDS: + field = info.get(numeric_field) + if field is None or isinstance(field, compat_numeric_types): + continue + report_force_conversion(numeric_field, 'numeric', 'int') + info[numeric_field] = int_or_none(field) + + sanitize_string_field(info_dict, 'id') + sanitize_numeric_fields(info_dict) + if 'playlist' not in info_dict: # It isn't part of a playlist info_dict['playlist'] = None info_dict['playlist_index'] = None - # This extractors handle format selection themselves - if info_dict['extractor'] in [u'youtube', u'Youku']: - if download: - self.process_info(info_dict) - return info_dict + thumbnails = info_dict.get('thumbnails') + if thumbnails is None: + thumbnail = info_dict.get('thumbnail') + if thumbnail: + info_dict['thumbnails'] = thumbnails = [{'url': thumbnail}] + if thumbnails: + thumbnails.sort(key=lambda t: ( + t.get('preference') if t.get('preference') is not None else -1, + t.get('width') if t.get('width') is not None else -1, + t.get('height') if t.get('height') is not None else -1, + t.get('id') if t.get('id') is not None else '', t.get('url'))) + for i, t in enumerate(thumbnails): + t['url'] = sanitize_url(t['url']) + if t.get('width') and t.get('height'): + t['resolution'] = '%dx%d' % (t['width'], t['height']) + if t.get('id') is None: + t['id'] = '%d' % i + + if self.params.get('list_thumbnails'): + self.list_thumbnails(info_dict) + return + + thumbnail = info_dict.get('thumbnail') + if thumbnail: + info_dict['thumbnail'] = sanitize_url(thumbnail) + elif thumbnails: + 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: + # Working around out-of-range timestamp values (e.g. negative ones on Windows, + # see http://bugs.python.org/issue1646728) + try: + upload_date = datetime.datetime.utcfromtimestamp(info_dict['timestamp']) + info_dict['upload_date'] = upload_date.strftime('%Y%m%d') + except (ValueError, OverflowError, OSError): + pass + + # Auto generate title fields corresponding to the *_number fields when missing + # in order to always have clean titles. This is very common for TV series. + for field in ('chapter', 'season', 'episode'): + if info_dict.get('%s_number' % field) is not None and not info_dict.get(field): + info_dict[field] = '%s %d' % (field.capitalize(), info_dict['%s_number' % field]) + + subtitles = info_dict.get('subtitles') + if subtitles: + for _, subtitle in subtitles.items(): + for subtitle_format in subtitle: + if subtitle_format.get('url'): + subtitle_format['url'] = sanitize_url(subtitle_format['url']) + if subtitle_format.get('ext') is None: + subtitle_format['ext'] = determine_ext(subtitle_format['url']).lower() + + if self.params.get('listsubtitles', False): + if 'automatic_captions' in info_dict: + self.list_subtitles(info_dict['id'], info_dict.get('automatic_captions'), 'automatic captions') + self.list_subtitles(info_dict['id'], subtitles, 'subtitles') + return + info_dict['requested_subtitles'] = self.process_subtitles( + info_dict['id'], subtitles, + info_dict.get('automatic_captions')) # We now pick which formats have to be downloaded if info_dict.get('formats') is None: @@ -571,62 +1499,123 @@ class YoutubeDL(object): else: formats = info_dict['formats'] + if not formats: + raise ExtractorError('No video formats found!') + + def is_wellformed(f): + url = f.get('url') + if not url: + self.report_warning( + '"url" field is missing or empty - skipping format, ' + 'there is an error in extractor') + return False + if isinstance(url, bytes): + sanitize_string_field(f, 'url') + return True + + # Filter out malformed formats for better extraction robustness + formats = list(filter(is_wellformed, formats)) + + formats_dict = {} + # We check that all the formats have the format and format_id fields - for (i, format) in enumerate(formats): - if format.get('format_id') is None: + for i, format in enumerate(formats): + sanitize_string_field(format, 'format_id') + sanitize_numeric_fields(format) + format['url'] = sanitize_url(format['url']) + if not format.get('format_id'): format['format_id'] = compat_str(i) + else: + # Sanitize format_id from characters used in format selector expression + format['format_id'] = re.sub(r'[\s,/+\[\]()]', '_', format['format_id']) + format_id = format['format_id'] + if format_id not in formats_dict: + formats_dict[format_id] = [] + formats_dict[format_id].append(format) + + # Make sure all formats have unique format_id + for format_id, ambiguous_formats in formats_dict.items(): + if len(ambiguous_formats) > 1: + for i, format in enumerate(ambiguous_formats): + format['format_id'] = '%s-%d' % (format_id, i) + + for i, format in enumerate(formats): 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): + if format.get('ext') is None: + format['ext'] = determine_ext(format['url']).lower() + # Automatically determine protocol if missing (useful for format + # selection purposes) + if format.get('protocol') is None: + format['protocol'] = determine_protocol(format) + # Add HTTP headers, so that external programs can use them from the + # json output + full_format_info = info_dict.copy() + full_format_info.update(format) + format['http_headers'] = self._calc_headers(full_format_info) + # Remove private housekeeping stuff + if '__x_forwarded_for_ip' in info_dict: + del info_dict['__x_forwarded_for_ip'] + + # 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, + # which can't be exported to json + info_dict['formats'] = formats + if self.params.get('listformats'): self.list_formats(info_dict) return - 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') + req_format = self.params.get('format') if req_format is None: - req_format = 'best' - formats_to_download = [] - # The -1 is for supporting YoutubeIE - if req_format in ('-1', 'all'): - formats_to_download = formats - else: - # 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: - selected_format = self.select_format(rf, formats) - if selected_format is not None: - formats_to_download = [selected_format] - break + req_format = self._default_format_spec(info_dict, download=download) + if self.params.get('verbose'): + self.to_stdout('[debug] Default format spec: %s' % req_format) + + format_selector = self.build_format_selector(req_format) + + # While in format selection we may need to have an access to the original + # format set in order to calculate some metrics or do some processing. + # For now we need to be able to guess whether original formats provided + # by extractor are incomplete or not (i.e. whether extractor provides only + # video-only or audio-only formats) for proper formats selection for + # extractors with such incomplete formats (see + # https://github.com/rg3/youtube-dl/pull/5556). + # Since formats may be filtered during format selection and may not match + # the original formats the results may be incorrect. Thus original formats + # or pre-calculated metrics should be passed to format selection routines + # as well. + # We will pass a context object containing all necessary additional data + # instead of just formats. + # This fixes incorrect format selection issue (see + # https://github.com/rg3/youtube-dl/issues/10083). + incomplete_formats = ( + # All formats are video-only or + all(f.get('vcodec') != 'none' and f.get('acodec') == 'none' for f in formats) or + # all formats are audio-only + all(f.get('vcodec') == 'none' and f.get('acodec') != 'none' for f in formats)) + + ctx = { + 'formats': formats, + 'incomplete_formats': incomplete_formats, + } + + formats_to_download = list(format_selector(ctx)) 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) @@ -635,53 +1624,105 @@ class YoutubeDL(object): info_dict.update(formats_to_download[-1]) return info_dict + def process_subtitles(self, video_id, normal_subtitles, automatic_captions): + """Select the requested subtitles and their format""" + available_subs = {} + if normal_subtitles and self.params.get('writesubtitles'): + available_subs.update(normal_subtitles) + if automatic_captions and self.params.get('writeautomaticsub'): + for lang, cap_info in automatic_captions.items(): + if lang not in available_subs: + available_subs[lang] = cap_info + + if (not self.params.get('writesubtitles') and not + self.params.get('writeautomaticsub') or not + available_subs): + return None + + if self.params.get('allsubtitles', False): + requested_langs = available_subs.keys() + else: + if self.params.get('subtitleslangs', False): + requested_langs = self.params.get('subtitleslangs') + elif 'en' in available_subs: + requested_langs = ['en'] + else: + requested_langs = [list(available_subs.keys())[0]] + + formats_query = self.params.get('subtitlesformat', 'best') + formats_preference = formats_query.split('/') if formats_query else [] + subs = {} + for lang in requested_langs: + formats = available_subs.get(lang) + if formats is None: + self.report_warning('%s subtitles not available for %s' % (lang, video_id)) + continue + for ext in formats_preference: + if ext == 'best': + f = formats[-1] + break + matches = list(filter(lambda f: f['ext'] == ext, formats)) + if matches: + f = matches[-1] + break + else: + f = formats[-1] + self.report_warning( + 'No subtitle format found matching "%s" for language %s, ' + 'using %s' % (formats_query, lang, f['ext'])) + subs[lang] = f + return subs + def process_info(self, info_dict): """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'] - - if not 'format' in info_dict: + if 'format' not in info_dict: info_dict['format'] = info_dict['ext'] - reason = self._match_entry(info_dict) + reason = self._match_entry(info_dict, incomplete=False) 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) + info_dict['_filename'] = 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'')) + if info_dict.get('requested_formats') is not None: + for f in info_dict['requested_formats']: + self.to_stdout(f['url'] + f.get('play_path', '')) + else: + # For RTMP URLs, also include the playpath + 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)) + self.to_stdout(json.dumps(info_dict)) # Do nothing else if in simulate mode if self.params.get('simulate', False): @@ -690,150 +1731,316 @@ class YoutubeDL(object): if filename is None: return - try: - dn = os.path.dirname(encodeFilename(filename)) - 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)) + def ensure_dir_exists(path): + try: + dn = os.path.dirname(path) + if dn and not os.path.exists(dn): + os.makedirs(dn) + return True + except (OSError, IOError) as err: + self.report_error('unable to create directory ' + error_to_compat_str(err)) + return False + + if not ensure_dir_exists(sanitize_path(encodeFilename(filename))): 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 = replace_extension(filename, 'description', info_dict.get('ext')) + if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(descfn)): + self.to_screen('[info] Video description is already present') + elif info_dict.get('description') is None: + self.report_warning('There\'s no description to write.') + 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 (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 = replace_extension(filename, 'annotations.xml', info_dict.get('ext')) + 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 info_dict.get('requested_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', 'srt') - for sub_lang in subtitles.keys(): - sub = subtitles[sub_lang] - if sub is None: - continue + subtitles = info_dict['requested_subtitles'] + ie = self.get_info_extractor(info_dict['extractor_key']) + for sub_lang, sub_info in subtitles.items(): + sub_format = sub_info['ext'] + sub_filename = subtitles_filename(filename, sub_lang, sub_format) + 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) + if sub_info.get('data') is not None: + try: + # Use newline='' to prevent conversion of newline characters + # See https://github.com/rg3/youtube-dl/issues/10268 + with io.open(encodeFilename(sub_filename), 'w', encoding='utf-8', newline='') as subfile: + subfile.write(sub_info['data']) + except (OSError, IOError): + self.report_error('Cannot write subtitles file ' + sub_filename) + return + else: + try: + sub_data = ie._request_webpage( + sub_info['url'], info_dict['id'], note=False).read() + with io.open(encodeFilename(sub_filename), 'wb') as subfile: + subfile.write(sub_data) + except (ExtractorError, IOError, OSError, ValueError) as err: + self.report_warning('Unable to download subtitle for "%s": %s' % + (sub_lang, error_to_compat_str(err))) + continue + + if self.params.get('writeinfojson', False): + infofn = replace_extension(filename, 'info.json', info_dict.get('ext')) + 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: - 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) + write_json_file(self.filter_requested_info(info_dict), infofn) except (OSError, IOError): - self.report_error(u'Cannot write subtitles file ' + descfn) + self.report_error('Cannot write metadata to JSON file ' + infofn) return - if self.params.get('writeinfojson', False): - infofn = os.path.splitext(filename)[0] + u'.info.json' - self.report_writeinfojson(infofn) + self._write_thumbnails(info_dict, filename) + + if not self.params.get('skip_download', False): 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) + def dl(name, info): + fd = get_suitable_downloader(info, self.params)(self, self.params) + for ph in self._progress_hooks: + fd.add_progress_hook(ph) + if self.params.get('verbose'): + self.to_stdout('[debug] Invoking downloader on %r' % info.get('url')) + return fd.download(name, info) + + if info_dict.get('requested_formats') is not None: + downloaded = [] + success = True + merger = FFmpegMergerPP(self) + if not merger.available: + 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] + + def compatible_formats(formats): + video, audio = formats + # Check extension + video_ext, audio_ext = audio.get('ext'), video.get('ext') + if video_ext and audio_ext: + COMPATIBLE_EXTS = ( + ('mp3', 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'ismv', 'isma'), + ('webm') + ) + for exts in COMPATIBLE_EXTS: + if video_ext in exts and audio_ext in exts: + return True + # TODO: Check acodec/vcodec + return False + + filename_real_ext = os.path.splitext(filename)[1][1:] + filename_wo_ext = ( + os.path.splitext(filename)[0] + if filename_real_ext == info_dict['ext'] + else filename) + requested_formats = info_dict['requested_formats'] + if self.params.get('merge_output_format') is None and not compatible_formats(requested_formats): + info_dict['ext'] = 'mkv' + self.report_warning( + 'Requested formats are incompatible for merge and will be merged into mkv.') + # Ensure filename always has a correct extension for successful merge + filename = '%s.%s' % (filename_wo_ext, info_dict['ext']) + if os.path.exists(encodeFilename(filename)): + self.to_screen( + '[download] %s has already been downloaded and ' + 'merged' % filename) + else: + for f in requested_formats: + new_info = dict(info_dict) + new_info.update(f) + fname = prepend_extension( + self.prepare_filename(new_info), + 'f%s' % f['format_id'], new_info['ext']) + if not ensure_dir_exists(fname): + return + 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('unable to download video data: %s' % error_to_compat_str(err)) + return + except (OSError, IOError) as err: + raise UnavailableVideoError(err) + except (ContentTooShortError, ) as err: + self.report_error('content too short (expected %s bytes and served %s)' % (err.expected, err.downloaded)) 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))) + if success and filename != '-': + # Fixup content + fixup_policy = self.params.get('fixup') + if fixup_policy is None: + fixup_policy = 'detect_or_warn' + + INSTALL_FFMPEG_MESSAGE = 'Install ffmpeg or avconv to fix this automatically.' + + stretched_ratio = info_dict.get('stretched_ratio') + if stretched_ratio is not None and stretched_ratio != 1: + if fixup_policy == 'warn': + self.report_warning('%s: Non-uniform pixel ratio (%s)' % ( + info_dict['id'], stretched_ratio)) + elif fixup_policy == 'detect_or_warn': + stretched_pp = FFmpegFixupStretchedPP(self) + if stretched_pp.available: + info_dict.setdefault('__postprocessors', []) + info_dict['__postprocessors'].append(stretched_pp) + else: + self.report_warning( + '%s: Non-uniform pixel ratio (%s). %s' + % (info_dict['id'], stretched_ratio, INSTALL_FFMPEG_MESSAGE)) + else: + assert fixup_policy in ('ignore', 'never') + + if (info_dict.get('requested_formats') is None and + info_dict.get('container') == 'm4a_dash'): + if fixup_policy == 'warn': + self.report_warning( + '%s: writing DASH m4a. ' + 'Only some players support this container.' + % info_dict['id']) + elif fixup_policy == 'detect_or_warn': + fixup_pp = FFmpegFixupM4aPP(self) + if fixup_pp.available: + info_dict.setdefault('__postprocessors', []) + info_dict['__postprocessors'].append(fixup_pp) + else: + self.report_warning( + '%s: writing DASH m4a. ' + 'Only some players support this container. %s' + % (info_dict['id'], INSTALL_FFMPEG_MESSAGE)) + else: + assert fixup_policy in ('ignore', 'never') + + if (info_dict.get('protocol') == 'm3u8_native' or + info_dict.get('protocol') == 'm3u8' and + self.params.get('hls_prefer_native')): + if fixup_policy == 'warn': + self.report_warning('%s: malformed AAC bitstream detected.' % ( + info_dict['id'])) + elif fixup_policy == 'detect_or_warn': + fixup_pp = FFmpegFixupM3u8PP(self) + if fixup_pp.available: + info_dict.setdefault('__postprocessors', []) + info_dict['__postprocessors'].append(fixup_pp) + else: + self.report_warning( + '%s: malformed AAC bitstream detected. %s' + % (info_dict['id'], INSTALL_FFMPEG_MESSAGE)) + else: + assert fixup_policy in ('ignore', 'never') - 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) - 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)) - 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)) - 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) + 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'] - and self.params.get('max_downloads') != 1): - raise SameFileError(self.params['outtmpl']) + outtmpl != '-' and + '%' not in outtmpl and + self.params.get('max_downloads') != 1): + raise SameFileError(outtmpl) for url in url_list: try: - #It also downloads the videos - self.extract_info(url) + # It also downloads the videos + res = self.extract_info( + url, force_generic_extractor=self.params.get('force_generic_extractor', False)) 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 + else: + if self.params.get('dump_single_json', False): + self.to_stdout(json.dumps(res)) return self._download_retcode + def download_with_info_file(self, info_filename): + with contextlib.closing(fileinput.FileInput( + [info_filename], mode='r', + openhook=fileinput.hook_encoded('utf-8'))) as f: + # FileInput doesn't have a read method, we can't call json.load + info = self.filter_requested_info(json.loads('\n'.join(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 + + @staticmethod + def filter_requested_info(info_dict): + return dict( + (k, v) for k, v in info_dict.items() + if k not in ['requested_formats', 'requested_subtitles']) + 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: + files_to_delete = [] try: - keep_video_wish, new_info = pp.run(info) - if keep_video_wish is not None: - if keep_video_wish: - keep_video = keep_video_wish - elif keep_video is None: - # No clear decision yet, let IE decide - keep_video = keep_video_wish + files_to_delete, info = pp.run(info) except PostProcessingError as e: 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) - os.remove(encodeFilename(filename)) - except (IOError, OSError): - self.report_warning(u'Unable to remove downloaded video file') + if files_to_delete and not self.params.get('keepvideo', False): + for old_filename in files_to_delete: + self.to_screen('Deleting original file %s (pass -k to keep)' % old_filename) + try: + os.remove(encodeFilename(old_filename)) + except (IOError, OSError): + self.report_warning('Unable to remove downloaded original file') def _make_archive_id(self, info_dict): # Future-proof against any change in case @@ -844,7 +2051,7 @@ 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 +2079,146 @@ 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 = '%dx?' % 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('language'): + if res: + res += ' ' + res += '[%s] ' % fdict['language'] + 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 - - def line(format, idlen=20): - return ((u'%-' + compat_str(idlen + 1) + u's%-10s%-12s%s') % ( - format['format_id'], - format['ext'], - self.format_resolution(format), - format_note(format), - )) + 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('fps') is not None: + if res: + res += ', ' + res += '%sfps' % fdict['fps'] + 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']) + elif fdict.get('filesize_approx') is not None: + if res: + res += ', ' + res += '~' + format_bytes(fdict['filesize_approx']) + return res + def list_formats(self, info_dict): formats = info_dict.get('formats', [info_dict]) - idlen = max(len(u'format code'), - max(len(f['format_id']) for f in formats)) - formats_s = [line(f, idlen) for f in formats] + table = [ + [f['format_id'], f['ext'], self.format_resolution(f), self._format_note(f)] + for f in formats + if f.get('preference') is None or f['preference'] >= -1000] if len(formats) > 1: - formats_s[0] += (' ' if format_note(formats[0]) else '') + '(worst)' - formats_s[-1] += (' ' if format_note(formats[-1]) else '') + '(best)' + table[-1][-1] += (' ' if table[-1][-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))) + header_line = ['format code', 'extension', 'resolution', 'note'] + self.to_screen( + '[info] Available formats for %s:\n%s' % + (info_dict['id'], render_table(header_line, table))) + + def list_thumbnails(self, info_dict): + thumbnails = info_dict.get('thumbnails') + if not thumbnails: + self.to_screen('[info] No thumbnails present for %s' % info_dict['id']) + return + + self.to_screen( + '[info] Thumbnails for %s:' % info_dict['id']) + self.to_screen(render_table( + ['ID', 'width', 'height', 'URL'], + [[t['id'], t.get('width', 'unknown'), t.get('height', 'unknown'), t['url']] for t in thumbnails])) + + def list_subtitles(self, video_id, subtitles, name='subtitles'): + if not subtitles: + self.to_screen('%s has no %s' % (video_id, name)) + return + self.to_screen( + 'Available %s for %s:' % (name, video_id)) + self.to_screen(render_table( + ['Language', 'formats'], + [[lang, ', '.join(f['ext'] for f in reversed(formats))] + for lang, formats in subtitles.items()])) def urlopen(self, req): """ Start an HTTP download """ - return self._opener.open(req) + if isinstance(req, compat_basestring): + req = sanitized_Request(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') + + if type('') is not compat_str: + # Python 2.6 on SLES11 SP1 (https://github.com/rg3/youtube-dl/issues/3326) + self.report_warning( + 'Your Python is broken! Update to a newer and supported version') + + stdout_encoding = getattr( + sys.stdout, 'encoding', 'missing (%s)' % type(sys.stdout).__name__) + encoding_str = ( + '[debug] Encodings: locale %s, fs %s, out %s, pref %s\n' % ( + locale.getpreferredencoding(), + sys.getfilesystemencoding(), + stdout_encoding, + self.get_encoding())) + write_string(encoding_str, encoding=None) + + self._write_string('[debug] youtube-dl version ' + __version__ + '\n') + if _LAZY_LOADER: + self._write_string('[debug] Lazy loading extractors enabled' + '\n') try: sp = subprocess.Popen( ['git', 'rev-parse', '--short', 'HEAD'], @@ -955,24 +2227,55 @@ 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') - except: + self._write_string('[debug] Git HEAD: ' + out + '\n') + except Exception: try: sys.exc_clear() - except: + except Exception: pass - write_string(u'[debug] Python version %s - %s' % - (platform.python_version(), platform_name()) + u'\n') + + def python_implementation(): + impl_name = platform.python_implementation() + if impl_name == 'PyPy' and hasattr(sys, 'pypy_version_info'): + return impl_name + ' version %d.%d.%d' % sys.pypy_version_info[:3] + return impl_name + + self._write_string('[debug] Python version %s (%s) - %s\n' % ( + platform.python_version(), python_implementation(), + platform_name())) + + exe_versions = FFmpegPostProcessor.get_versions(self) + exe_versions['rtmpdump'] = rtmpdump_version() + exe_versions['phantomjs'] = PhantomJSwrapper._version() + exe_str = ', '.join( + '%s %s' % (exe, v) + for exe, v in sorted(exe_versions.items()) + if v + ) + if not exe_str: + exe_str = 'none' + self._write_string('[debug] exe versions: %s\n' % exe_str) 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') + + if self.params.get('call_home', False): + ipaddr = self.urlopen('https://yt-dl.org/ip').read().decode('utf-8') + self._write_string('[debug] Public IP address: %s\n' % ipaddr) + latest_version = self.urlopen( + 'https://yt-dl.org/latest/version').read().decode('utf-8') + if version_tuple(latest_version) > version_tuple(__version__): + self.report_warning( + 'You are using an outdated version (newest version: %s)! ' + 'See https://yt-dl.org/update if you need help updating.' % + latest_version) 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') @@ -980,13 +2283,13 @@ class YoutubeDL(object): if opts_cookiefile is None: self.cookiejar = compat_cookiejar.CookieJar() else: + opts_cookiefile = expand_path(opts_cookiefile) 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) + cookie_processor = YoutubeDLCookieProcessor(self.cookiejar) if opts_proxy is not None: if opts_proxy == '': proxies = {} @@ -997,17 +2300,80 @@ class YoutubeDL(object): # Set HTTPS proxy to HTTP one if given (https://github.com/rg3/youtube-dl/issues/805) if 'http' in proxies and 'https' not in proxies: proxies['https'] = proxies['http'] - proxy_handler = compat_urllib_request.ProxyHandler(proxies) - https_handler = make_HTTPS_handler( - self.params.get('nocheckcertificate', False)) + proxy_handler = PerRequestProxyHandler(proxies) + + debuglevel = 1 if self.params.get('debug_printtraffic') else 0 + https_handler = make_HTTPS_handler(self.params, debuglevel=debuglevel) + ydlh = YoutubeDLHandler(self.params, debuglevel=debuglevel) + data_handler = compat_urllib_request_DataHandler() + + # When passing our own FileHandler instance, build_opener won't add the + # default FileHandler and allows us to disable the file protocol, which + # can be used for malicious purposes (see + # https://github.com/rg3/youtube-dl/issues/8227) + file_handler = compat_urllib_request.FileHandler() + + def file_open(*args, **kwargs): + raise compat_urllib_error.URLError('file:// scheme is explicitly disabled in youtube-dl for security reasons') + file_handler.file_open = file_open + opener = compat_urllib_request.build_opener( - https_handler, proxy_handler, cookie_processor, YoutubeDLHandler()) + proxy_handler, https_handler, cookie_processor, ydlh, data_handler, file_handler) + # 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 + + def _write_thumbnails(self, info_dict, filename): + if self.params.get('writethumbnail', False): + thumbnails = info_dict.get('thumbnails') + if thumbnails: + thumbnails = [thumbnails[-1]] + elif self.params.get('write_all_thumbnails', False): + thumbnails = info_dict.get('thumbnails') + else: + return + + if not thumbnails: + # No thumbnails present, so return immediately + return + + for t in thumbnails: + thumb_ext = determine_ext(t['url'], 'jpg') + suffix = '_%s' % t['id'] if len(thumbnails) > 1 else '' + thumb_display_id = '%s ' % t['id'] if len(thumbnails) > 1 else '' + t['filename'] = thumb_filename = os.path.splitext(filename)[0] + suffix + '.' + thumb_ext + + if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(thumb_filename)): + self.to_screen('[%s] %s: Thumbnail %sis already present' % + (info_dict['extractor'], info_dict['id'], thumb_display_id)) + else: + self.to_screen('[%s] %s: Downloading thumbnail %s...' % + (info_dict['extractor'], info_dict['id'], thumb_display_id)) + try: + uf = self.urlopen(t['url']) + with open(encodeFilename(thumb_filename), 'wb') as thumbf: + shutil.copyfileobj(uf, thumbf) + self.to_screen('[%s] %s: Writing thumbnail %sto: %s' % + (info_dict['extractor'], info_dict['id'], thumb_display_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' % + (t['url'], error_to_compat_str(err)))