X-Git-Url: https://git.rapsys.eu/youtubedl/blobdiff_plain/f46044c66663049e286c20ee015db99d47d9dd8a..d018d3313032e12968a6add6800e51d412e2f602:/youtube_dl/YoutubeDL.py?ds=inline diff --git a/youtube_dl/YoutubeDL.py b/youtube_dl/YoutubeDL.py index 74e4261..9a8c7da 100755 --- a/youtube_dl/YoutubeDL.py +++ b/youtube_dl/YoutubeDL.py @@ -4,8 +4,10 @@ from __future__ import absolute_import, unicode_literals import collections +import contextlib import datetime import errno +import fileinput import io import itertools import json @@ -19,6 +21,7 @@ import subprocess import socket import sys import time +import tokenize import traceback if os.name == 'nt': @@ -28,14 +31,16 @@ from .compat import ( compat_basestring, compat_cookiejar, compat_expanduser, + compat_get_terminal_size, compat_http_client, compat_kwargs, compat_str, + compat_tokenize_tokenize, compat_urllib_error, compat_urllib_request, + compat_urllib_request_DataHandler, ) from .utils import ( - escape_url, ContentTooShortError, date_from_str, DateRange, @@ -46,28 +51,31 @@ from .utils import ( ExtractorError, format_bytes, formatSeconds, - get_term_width, locked_file, make_HTTPS_handler, MaxDownloadsReached, PagedList, parse_filesize, + PerRequestProxyHandler, PostProcessingError, platform_name, preferredencoding, render_table, SameFileError, sanitize_filename, + sanitize_path, + sanitized_Request, std_headers, subtitles_filename, - takewhile_inclusive, UnavailableVideoError, url_basename, version_tuple, write_json_file, write_string, + YoutubeDLCookieProcessor, YoutubeDLHandler, prepend_extension, + replace_extension, args_to_str, age_restricted, ) @@ -114,7 +122,7 @@ 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. usenetrc: Use netrc for authentication instead. verbose: Print additional info to stdout. quiet: Do not print messages to stdout. @@ -131,10 +139,10 @@ class YoutubeDL(object): (or video) as a single JSON line. simulate: Do not download the video files. format: Video format code. See options.py for more information. - format_limit: Highest quality format to try. 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. @@ -150,7 +158,7 @@ class YoutubeDL(object): 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 @@ -181,6 +189,8 @@ class YoutubeDL(object): 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 + cn_verification_proxy: URL of the proxy to use for IP address verification + on Chinese sites. (Experimental) socket_timeout: Time to wait for unresponsive hosts, in seconds bidi_workaround: Work around buggy terminals without bidirectional text support, using fridibi @@ -247,15 +257,16 @@ class YoutubeDL(object): hls_prefer_native: Use the native HLS downloader instead of ffmpeg/avconv. 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, - xattr_set_filesize. + xattr_set_filesize, external_downloader_args. 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. - exec_cmd: Arbitrary command to run after downloading + postprocessor_args: A list of additional command-line arguments for the + postprocessor. """ params = None @@ -277,14 +288,18 @@ class YoutubeDL(object): self._num_downloads = 0 self._screen_file = [sys.stdout, sys.stderr][params.get('logtostderr', False)] self._err_file = sys.stderr - self.params = params + self.params = { + # Default parameters + 'nocheckcertificate': False, + } + self.params.update(params) self.cache = Cache(self) if params.get('bidi_workaround', False): try: import pty master, slave = pty.openpty() - width = get_term_width() + width = compat_get_terminal_size().columns if width is None: width_args = [] else: @@ -317,8 +332,10 @@ class YoutubeDL(object): 'Set the LC_ALL environment variable to fix this.') self.params['restrictfilenames'] = True - if '%(stitle)s' in self.params.get('outtmpl', ''): - self.report_warning('%(stitle)s is deprecated. Use the %(title)s and the --restrict-filenames flag(which also secures %(uploader)s et al) instead.') + if 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() @@ -565,7 +582,7 @@ class YoutubeDL(object): # to workaround encoding issues with subprocess on python2 @ Windows if sys.version_info < (3, 0) and sys.platform == 'win32': filename = encodeFilename(filename, True).decode(preferredencoding()) - return filename + return sanitize_path(filename) except ValueError as err: self.report_error('Error in output template: ' + str(err) + ' (encoding: ' + repr(preferredencoding()) + ')') return None @@ -619,12 +636,15 @@ class YoutubeDL(object): info_dict.setdefault(key, value) def extract_info(self, url, download=True, ie_key=None, extra_info={}, - process=True): + 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)] @@ -753,7 +773,9 @@ class YoutubeDL(object): if isinstance(ie_entries, list): n_all_entries = len(ie_entries) if playlistitems: - entries = [ie_entries[i - 1] for i in playlistitems] + entries = [ + ie_entries[i - 1] for i in playlistitems + if -n_all_entries <= i - 1 < n_all_entries] else: entries = ie_entries[playliststart:playlistend] n_entries = len(entries) @@ -813,6 +835,7 @@ 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( @@ -838,8 +861,8 @@ class YoutubeDL(object): else: raise Exception('Invalid result type: %s' % result_type) - def _apply_format_filter(self, format_spec, available_formats): - " Returns a tuple of the remaining format_spec and filtered formats " + def _build_format_filter(self, filter_spec): + " Returns a function to filter the formats according to the filter_spec " OPERATORS = { '<': operator.lt, @@ -849,13 +872,13 @@ class YoutubeDL(object): '=': operator.eq, '!=': operator.ne, } - operator_rex = re.compile(r'''(?x)\s*\[ + operator_rex = re.compile(r'''(?x)\s* (?Pwidth|height|tbr|abr|vbr|asr|filesize|fps) \s*(?P%s)(?P\s*\?)?\s* (?P[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?) - \]$ + $ ''' % '|'.join(map(re.escape, OPERATORS.keys()))) - m = operator_rex.search(format_spec) + m = operator_rex.search(filter_spec) if m: try: comparison_value = int(m.group('value')) @@ -866,7 +889,7 @@ class YoutubeDL(object): if comparison_value is None: raise ValueError( 'Invalid value %r in format specification %r' % ( - m.group('value'), format_spec)) + m.group('value'), filter_spec)) op = OPERATORS[m.group('op')] if not m: @@ -874,78 +897,283 @@ class YoutubeDL(object): '=': operator.eq, '!=': operator.ne, } - str_operator_rex = re.compile(r'''(?x)\s*\[ + str_operator_rex = re.compile(r'''(?x) \s*(?Pext|acodec|vcodec|container|protocol) \s*(?P%s)(?P\s*\?)? \s*(?P[a-zA-Z0-9_-]+) - \s*\]$ + \s*$ ''' % '|'.join(map(re.escape, STR_OPERATORS.keys()))) - m = str_operator_rex.search(format_spec) + 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 format specification %r' % format_spec) + 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) - new_formats = [f for f in available_formats if _filter(f)] + return _filter + + 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(formats): + for f in fs: + for format in f(formats): + 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(formats): + for f in fs: + picked_formats = list(f(formats)) + if picked_formats: + return picked_formats + return [] + elif selector.type == SINGLE: + format_spec = selector.selector + + def selector_function(formats): + formats = list(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 audio only (soundcloud) or video only (imgur) urls, select the best/worst audio format + elif (all(f.get('acodec') != 'none' for f in formats) or + all(f.get('vcodec') != 'none' for f in 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 + 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) - new_format_spec = format_spec[:-len(m.group(0))] - if not new_format_spec: - new_format_spec = 'best' + def selector_function(formats): + formats = list(formats) + for pair in itertools.product(video_selector(formats), audio_selector(formats)): + yield _merge(pair) - return (new_format_spec, new_formats) + filters = [self._build_format_filter(f) for f in selector.filters] - def select_format(self, format_spec, available_formats): - while format_spec.endswith(']'): - format_spec, available_formats = self._apply_format_filter( - format_spec, available_formats) - if not available_formats: - return None + def final_selector(formats): + for _filter in filters: + formats = list(filter(_filter, formats)) + return selector_function(formats) + return final_selector - if format_spec == 'best' or format_spec is None: - return available_formats[-1] - elif format_spec == 'worst': - return available_formats[0] - elif format_spec == 'bestaudio': - audio_formats = [ - f for f in available_formats - if f.get('vcodec') == 'none'] - if audio_formats: - return audio_formats[-1] - elif format_spec == 'worstaudio': - audio_formats = [ - f for f in available_formats - if f.get('vcodec') == 'none'] - if audio_formats: - return audio_formats[0] - elif format_spec == 'bestvideo': - video_formats = [ - f for f in available_formats - if f.get('acodec') == 'none'] - if video_formats: - return video_formats[-1] - elif format_spec == 'worstvideo': - video_formats = [ - f for f in available_formats - if f.get('acodec') == 'none'] - if video_formats: - return video_formats[0] - else: - extensions = ['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, available_formats)) - if matches: - return matches[-1] - return None + 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() @@ -961,7 +1189,7 @@ class YoutubeDL(object): return res def _calc_cookies(self, info_dict): - pr = compat_urllib_request.Request(info_dict['url']) + pr = sanitized_Request(info_dict['url']) self.cookiejar.add_cookie_header(pr) return pr.get_header('Cookie') @@ -988,7 +1216,7 @@ class YoutubeDL(object): t.get('preference'), t.get('width'), t.get('height'), t.get('id'), t.get('url'))) for i, t in enumerate(thumbnails): - if 'width' in t and 'height' in t: + 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 @@ -1000,29 +1228,30 @@ class YoutubeDL(object): 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 negative timestamps in Windows - # (see http://bugs.python.org/issue1646728) - if info_dict['timestamp'] < 0 and os.name == 'nt': - info_dict['timestamp'] = 0 - upload_date = datetime.datetime.utcfromtimestamp( - info_dict['timestamp']) - info_dict['upload_date'] = upload_date.strftime('%Y%m%d') + # 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 + + subtitles = info_dict.get('subtitles') + if subtitles: + for _, subtitle in subtitles.items(): + for subtitle_format in subtitle: + if 'ext' not in subtitle_format: + 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'], info_dict.get('subtitles'), 'subtitles') + self.list_subtitles(info_dict['id'], subtitles, 'subtitles') return info_dict['requested_subtitles'] = self.process_subtitles( - info_dict['id'], info_dict.get('subtitles'), + info_dict['id'], subtitles, info_dict.get('automatic_captions')) - # This extractors handle format selection themselves - if info_dict['extractor'] in ['Youku']: - if download: - self.process_info(info_dict) - return info_dict - # We now pick which formats have to be downloaded if info_dict.get('formats') is None: # There's only one format available @@ -1033,6 +1262,8 @@ class YoutubeDL(object): if not formats: raise ExtractorError('No video formats found!') + formats_dict = {} + # We check that all the formats have the format and format_id fields for i, format in enumerate(formats): if 'url' not in format: @@ -1040,6 +1271,18 @@ class YoutubeDL(object): if format.get('format_id') is None: format['format_id'] = compat_str(i) + 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'] = '{id} - {res}{note}'.format( id=format['format_id'], @@ -1055,12 +1298,6 @@ class YoutubeDL(object): full_format_info.update(format) format['http_headers'] = self._calc_headers(full_format_info) - format_limit = self.params.get('format_limit', None) - if format_limit: - formats = list(takewhile_inclusive( - lambda f: f['format_id'] != format_limit, formats - )) - # TODO Central sorting goes here if formats[0] is not info_dict: @@ -1078,58 +1315,17 @@ class YoutubeDL(object): 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: - for rfstr in req_format.split(','): - # We can accept formats requested in the format: 34/5/best, we pick - # the first that is available, starting from left - req_formats = rfstr.split('/') - for rf in req_formats: - if re.match(r'.+?\+.+?', rf) is not None: - # Two formats have been requested like '137+139' - format_1, format_2 = rf.split('+') - formats_info = (self.select_format(format_1, formats), - self.select_format(format_2, formats)) - if all(formats_info): - # 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 - output_ext = ( - formats_info[0]['ext'] - if self.params.get('merge_output_format') is None - else self.params['merge_output_format']) - selected_format = { - '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, - } - else: - selected_format = None - else: - selected_format = self.select_format(rf, formats) - if selected_format is not None: - formats_to_download.append(selected_format) - break + req_format_list = [] + if (self.params.get('outtmpl', DEFAULT_OUTTMPL) != '-' and + info_dict['extractor'] in ['youtube', 'ted'] and + not info_dict.get('is_live')): + merger = FFmpegMergerPP(self) + if merger.available and merger.can_merge(): + req_format_list.append('bestvideo+bestaudio') + req_format_list.append('best') + req_format = '/'.join(req_format_list) + format_selector = self.build_format_selector(req_format) + formats_to_download = list(format_selector(formats)) if not formats_to_download: raise ExtractorError('requested format not available', expected=True) @@ -1208,9 +1404,6 @@ class YoutubeDL(object): if len(info_dict['title']) > 200: info_dict['title'] = info_dict['title'][:197] + '...' - # Keep for backwards compatibility - info_dict['stitle'] = info_dict['title'] - if 'format' not in info_dict: info_dict['format'] = info_dict['ext'] @@ -1256,7 +1449,7 @@ class YoutubeDL(object): return try: - dn = os.path.dirname(encodeFilename(filename)) + dn = os.path.dirname(sanitize_path(encodeFilename(filename))) if dn and not os.path.exists(dn): os.makedirs(dn) except (OSError, IOError) as err: @@ -1264,7 +1457,7 @@ class YoutubeDL(object): return if self.params.get('writedescription', False): - descfn = filename + '.description' + 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: @@ -1279,7 +1472,7 @@ class YoutubeDL(object): return if self.params.get('writeannotations', False): - annofn = filename + '.annotations.xml' + 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: @@ -1326,13 +1519,13 @@ class YoutubeDL(object): return if self.params.get('writeinfojson', False): - infofn = os.path.splitext(filename)[0] + '.info.json' + 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: - write_json_file(info_dict, infofn) + write_json_file(self.filter_requested_info(info_dict), infofn) except (OSError, IOError): self.report_error('Cannot write metadata to JSON file ' + infofn) return @@ -1352,24 +1545,57 @@ class YoutubeDL(object): if info_dict.get('requested_formats') is not None: downloaded = [] success = True - merger = FFmpegMergerPP(self, not self.params.get('keepvideo')) + 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') + ' The formats won\'t be merged.') else: postprocessors = [merger] - for f in info_dict['requested_formats']: - new_info = dict(info_dict) - new_info.update(f) - fname = self.prepare_filename(new_info) - fname = prepend_extension(fname, 'f%s' % f['format_id']) - downloaded.append(fname) - partial_success = dl(fname, new_info) - success = success and partial_success - info_dict['__postprocessors'] = postprocessors - info_dict['__files_to_merge'] = downloaded + + 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'), + ('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 = self.prepare_filename(new_info) + fname = prepend_extension(fname, 'f%s' % f['format_id'], new_info['ext']) + 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) @@ -1439,7 +1665,8 @@ class YoutubeDL(object): for url in url_list: try: # It also downloads the videos - res = self.extract_info(url) + res = self.extract_info( + url, force_generic_extractor=self.params.get('force_generic_extractor', False)) except UnavailableVideoError: self.report_error('unable to download video') except MaxDownloadsReached: @@ -1452,8 +1679,11 @@ class YoutubeDL(object): return self._download_retcode def download_with_info_file(self, info_filename): - with io.open(info_filename, 'r', encoding='utf-8') as f: - info = json.load(f) + 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: @@ -1465,6 +1695,12 @@ class YoutubeDL(object): 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) @@ -1474,24 +1710,18 @@ class YoutubeDL(object): pps_chain.extend(ie_info['__postprocessors']) pps_chain.extend(self._pps) for pp in pps_chain: - keep_video = None - old_filename = info['filepath'] + files_to_delete = [] try: - keep_video_wish, 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: + 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) - os.remove(encodeFilename(old_filename)) - except (IOError, OSError): - self.report_warning('Unable to remove downloaded video file') + 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 @@ -1643,26 +1873,8 @@ class YoutubeDL(object): def urlopen(self, req): """ Start an HTTP download """ - - # According to RFC 3986, URLs can not contain non-ASCII characters, however this is not - # always respected by websites, some tend to give out URLs with non percent-encoded - # non-ASCII characters (see telemb.py, ard.py [#3412]) - # urllib chokes on URLs with non-ASCII characters (see http://bugs.python.org/issue3991) - # To work around aforementioned issue we will replace request's original URL with - # percent-encoded one - req_is_string = isinstance(req, compat_basestring) - url = req if req_is_string else req.get_full_url() - url_escaped = escape_url(url) - - # Substitute URL if any change after escaping - if url != url_escaped: - if req_is_string: - req = url_escaped - else: - req = compat_urllib_request.Request( - url_escaped, data=req.data, headers=req.headers, - origin_req_host=req.origin_req_host, unverifiable=req.unverifiable) - + if isinstance(req, compat_basestring): + req = sanitized_Request(req) return self._opener.open(req, timeout=self._socket_timeout) def print_debug_header(self): @@ -1694,10 +1906,10 @@ class YoutubeDL(object): out = out.decode().strip() if re.match('[0-9a-f]+', out): self._write_string('[debug] Git HEAD: ' + out + '\n') - except: + except Exception: try: sys.exc_clear() - except: + except Exception: pass self._write_string('[debug] Python version %s - %s\n' % ( platform.python_version(), platform_name())) @@ -1745,8 +1957,7 @@ class YoutubeDL(object): 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 = {} @@ -1757,13 +1968,15 @@ 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) + 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() opener = compat_urllib_request.build_opener( - https_handler, proxy_handler, cookie_processor, ydlh) + proxy_handler, https_handler, cookie_processor, ydlh, data_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) @@ -1804,7 +2017,7 @@ class YoutubeDL(object): 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 '' - thumb_filename = os.path.splitext(filename)[0] + suffix + '.' + thumb_ext + 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' % @@ -1814,7 +2027,7 @@ class YoutubeDL(object): (info_dict['extractor'], info_dict['id'], thumb_display_id)) try: uf = self.urlopen(t['url']) - with open(thumb_filename, 'wb') as thumbf: + 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))