X-Git-Url: https://jfr.im/git/yt-dlp.git/blobdiff_plain/48c8424bd9e03fdfd5c4c4495de233e896eb1f16..d5d1df8afdd532cc889f9d95be0740668a0776fe:/yt_dlp/YoutubeDL.py diff --git a/yt_dlp/YoutubeDL.py b/yt_dlp/YoutubeDL.py index 99db8be92..42780e794 100644 --- a/yt_dlp/YoutubeDL.py +++ b/yt_dlp/YoutubeDL.py @@ -108,6 +108,7 @@ get_domain, int_or_none, iri_to_uri, + is_path_like, join_nonempty, locked_file, make_archive_id, @@ -251,8 +252,8 @@ class YoutubeDL: matchtitle: Download only matching titles. rejecttitle: Reject downloads for matching titles. logger: Log messages to a logging.Logger instance. - logtostderr: Log messages to stderr instead of stdout. - consoletitle: Display progress in console window's titlebar. + logtostderr: Print everything to stderr instead of stdout. + consoletitle: Display progress in console window's titlebar. writedescription: Write the video description to a .description file writeinfojson: Write the video description to a .info.json file clean_infojson: Remove private fields from the infojson @@ -293,9 +294,8 @@ class YoutubeDL: 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. + download_archive: A set, or the name of a file where all downloads are recorded. + Videos already present in the file are not downloaded again. break_on_existing: Stop the download process after attempting to download a file that is in the archive. break_on_reject: Stop the download process when encountering a video that @@ -548,7 +548,7 @@ class YoutubeDL: # NB: Keep in sync with the docstring of extractor/common.py 'url', 'manifest_url', 'manifest_stream_number', 'ext', 'format', 'format_id', 'format_note', 'width', 'height', 'resolution', 'dynamic_range', 'tbr', 'abr', 'acodec', 'asr', 'audio_channels', - 'vbr', 'fps', 'vcodec', 'container', 'filesize', 'filesize_approx', + 'vbr', 'fps', 'vcodec', 'container', 'filesize', 'filesize_approx', 'rows', 'columns', 'player_url', 'protocol', 'fragment_base_url', 'fragments', 'is_from_start', 'preference', 'language', 'language_preference', 'quality', 'source_preference', 'http_headers', 'stretched_ratio', 'no_resume', 'has_drm', 'downloader_options', @@ -723,21 +723,23 @@ def check_deprecated(param, option, suggestion): def preload_download_archive(fn): """Preload the archive, if any is specified""" + archive = set() if fn is None: - return False + return archive + elif not is_path_like(fn): + return fn + self.write_debug(f'Loading archive file {fn!r}') try: with locked_file(fn, 'r', encoding='utf-8') as archive_file: for line in archive_file: - self.archive.add(line.strip()) + archive.add(line.strip()) except OSError as ioe: if ioe.errno != errno.ENOENT: raise - return False - return True + return archive - self.archive = set() - preload_download_archive(self.params.get('download_archive')) + self.archive = preload_download_archive(self.params.get('download_archive')) def warn_if_short_id(self, argv): # short YouTube ID starting with dash? @@ -844,7 +846,7 @@ def to_stdout(self, message, skip_eol=False, quiet=None): 'Use "YoutubeDL.to_screen" instead') self._write_string(f'{self._bidi_workaround(message)}\n', self._out_files.out) - def to_screen(self, message, skip_eol=False, quiet=None): + def to_screen(self, message, skip_eol=False, quiet=None, only_once=False): """Print message to screen if not in quiet mode""" if self.params.get('logger'): self.params['logger'].debug(message) @@ -853,7 +855,7 @@ def to_screen(self, message, skip_eol=False, quiet=None): return self._write_string( '%s%s' % (self._bidi_workaround(message), ('' if skip_eol else '\n')), - self._out_files.screen) + self._out_files.screen, only_once=only_once) def to_stderr(self, message, only_once=False): """Print message to stderr""" @@ -1245,9 +1247,11 @@ def create_key(outer_mobj): delim = '\n' if '#' in flags else ', ' value, fmt = delim.join(map(str, variadic(value, allowed_types=(str, bytes)))), str_fmt elif fmt[-1] == 'j': # json - value, fmt = json.dumps(value, default=_dumpjson_default, indent=4 if '#' in flags else None), str_fmt + value, fmt = json.dumps( + value, default=_dumpjson_default, + indent=4 if '#' in flags else None, ensure_ascii='+' not in flags), str_fmt elif fmt[-1] == 'h': # html - value, fmt = escapeHTML(value), str_fmt + value, fmt = escapeHTML(str(value)), str_fmt elif fmt[-1] == 'q': # quoted value = map(str, variadic(value) if '#' in flags else [value]) value, fmt = ' '.join(map(compat_shlex_quote, value)), str_fmt @@ -1419,18 +1423,19 @@ def add_extra_info(info_dict, extra_info): def extract_info(self, url, download=True, ie_key=None, extra_info=None, process=True, force_generic_extractor=False): """ - Return a list with a dictionary for each video extracted. + Extract and return the information dictionary of the URL Arguments: - url -- URL to extract + @param url URL to extract Keyword arguments: - download -- whether to download videos during extraction - ie_key -- extractor key hint - extra_info -- dictionary containing the extra values to add to each result - process -- whether to resolve all unresolved references (URLs, playlist items), - must be True for download to work. - force_generic_extractor -- force using the generic extractor + @param download Whether to download videos + @param process Whether to resolve all unresolved references (URLs, playlist items). + Must be True for download to work + @param ie_key Use only the extractor with this key + + @param extra_info Dictionary containing the extra values to add to the info (For internal use only) + @force_generic_extractor Force using the generic extractor (Deprecated; use ie_key='Generic') """ if extra_info is None: @@ -1616,6 +1621,7 @@ def process_ie_result(self, ie_result, download=True, extra_info=None): self.add_default_extra_info(info_copy, ie, ie_result['url']) self.add_extra_info(info_copy, extra_info) info_copy, _ = self.pre_process(info_copy) + self._fill_common_fields(info_copy, False) self.__forced_printings(info_copy, self.prepare_filename(info_copy), incomplete=True) self._raise_pending_errors(info_copy) if self.params.get('force_write_download_archive', False): @@ -1682,8 +1688,8 @@ def process_ie_result(self, ie_result, download=True, extra_info=None): elif result_type in ('playlist', 'multi_video'): # Protect from infinite recursion due to recursively nested playlists # (see https://github.com/ytdl-org/youtube-dl/issues/27833) - webpage_url = ie_result['webpage_url'] - if webpage_url in self._playlist_urls: + webpage_url = ie_result.get('webpage_url') # Playlists maynot have webpage_url + if webpage_url and webpage_url in self._playlist_urls: self.to_screen( '[download] Skipping already downloaded playlist: %s' % ie_result.get('title') or ie_result.get('id')) @@ -1737,14 +1743,17 @@ def _playlist_infodict(ie_result, strict=False, **kwargs): } if strict: return info + if ie_result.get('webpage_url'): + info.update({ + 'webpage_url': ie_result['webpage_url'], + 'webpage_url_basename': url_basename(ie_result['webpage_url']), + 'webpage_url_domain': get_domain(ie_result['webpage_url']), + }) return { **info, 'playlist_index': 0, '__last_playlist_index': max(ie_result['requested_entries'] or (0, 0)), 'extractor': ie_result['extractor'], - 'webpage_url': ie_result['webpage_url'], - 'webpage_url_basename': url_basename(ie_result['webpage_url']), - 'webpage_url_domain': get_domain(ie_result['webpage_url']), 'extractor_key': ie_result['extractor_key'], } @@ -2371,10 +2380,9 @@ def check_thumbnails(thumbnails): else: info_dict['thumbnails'] = thumbnails - def _fill_common_fields(self, info_dict, is_video=True): + def _fill_common_fields(self, info_dict, final=True): # TODO: move sanitization here - if is_video: - # playlists are allowed to lack "title" + if final: title = info_dict.get('title', NO_DEFAULT) if title is NO_DEFAULT: raise ExtractorError('Missing "title" field in extractor result', @@ -2418,11 +2426,13 @@ def _fill_common_fields(self, info_dict, is_video=True): for key in live_keys: if info_dict.get(key) is None: info_dict[key] = (live_status == key) + if live_status == 'post_live': + info_dict['was_live'] = True # 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): + if final and 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]) def _raise_pending_errors(self, info): @@ -2515,21 +2525,17 @@ def sanitize_numeric_fields(info): info_dict['requested_subtitles'] = self.process_subtitles( info_dict['id'], subtitles, automatic_captions) - if info_dict.get('formats') is None: - # There's only one format available - formats = [info_dict] - else: - formats = info_dict['formats'] + formats = self._get_formats(info_dict) # or None ensures --clean-infojson removes it info_dict['_has_drm'] = any(f.get('has_drm') for f in formats) or None if not self.params.get('allow_unplayable_formats'): formats = [f for f in formats if not f.get('has_drm')] - if info_dict['_has_drm'] and formats and all( - f.get('acodec') == f.get('vcodec') == 'none' for f in formats): - self.report_warning( - 'This video is DRM protected and only images are available for download. ' - 'Use --list-formats to see them') + + if formats and all(f.get('acodec') == f.get('vcodec') == 'none' for f in formats): + self.report_warning( + f'{"This video is DRM protected and " if info_dict["_has_drm"] else ""}' + 'only images are available for download. Use --list-formats to see them'.capitalize()) get_from_start = not info_dict.get('is_live') or bool(self.params.get('live_from_start')) if not get_from_start: @@ -2634,7 +2640,7 @@ def is_wellformed(f): info_dict, _ = self.pre_process(info_dict, 'after_filter') # The pre-processors may have modified the formats - formats = info_dict.get('formats', [info_dict]) + formats = self._get_formats(info_dict) list_only = self.params.get('simulate') is None and ( self.params.get('list_thumbnails') or self.params.get('listformats') or self.params.get('listsubtitles')) @@ -2692,31 +2698,30 @@ def is_wellformed(f): # Process what we can, even without any available formats. formats_to_download = [{}] - requested_ranges = self.params.get('download_ranges') - if requested_ranges: - requested_ranges = tuple(requested_ranges(info_dict, self)) - + requested_ranges = tuple(self.params.get('download_ranges', lambda *_: [{}])(info_dict, self)) best_format, downloaded_formats = formats_to_download[-1], [] if download: - if best_format: + if best_format and requested_ranges: def to_screen(*msg): self.to_screen(f'[info] {info_dict["id"]}: {" ".join(", ".join(variadic(m)) for m in msg)}') to_screen(f'Downloading {len(formats_to_download)} format(s):', (f['format_id'] for f in formats_to_download)) - if requested_ranges: + if requested_ranges != ({}, ): to_screen(f'Downloading {len(requested_ranges)} time ranges:', - (f'{int(c["start_time"])}-{int(c["end_time"])}' for c in requested_ranges)) + (f'{c["start_time"]:.1f}-{c["end_time"]:.1f}' for c in requested_ranges)) max_downloads_reached = False - for fmt, chapter in itertools.product(formats_to_download, requested_ranges or [{}]): + for fmt, chapter in itertools.product(formats_to_download, requested_ranges): new_info = self._copy_infodict(info_dict) new_info.update(fmt) offset, duration = info_dict.get('section_start') or 0, info_dict.get('duration') or float('inf') + end_time = offset + min(chapter.get('end_time', duration), duration) if chapter or offset: new_info.update({ 'section_start': offset + chapter.get('start_time', 0), - 'section_end': offset + min(chapter.get('end_time', duration), duration), + # duration may not be accurate. So allow deviations <1sec + 'section_end': end_time if end_time <= offset + duration + 1 else None, 'section_title': chapter.get('title'), 'section_number': chapter.get('index'), }) @@ -3464,8 +3469,7 @@ def _make_archive_id(self, info_dict): return make_archive_id(extractor, video_id) def in_download_archive(self, info_dict): - fn = self.params.get('download_archive') - if fn is None: + if not self.archive: return False vid_ids = [self._make_archive_id(info_dict)] @@ -3478,9 +3482,11 @@ def record_download_archive(self, info_dict): return vid_id = self._make_archive_id(info_dict) assert vid_id + self.write_debug(f'Adding to archive: {vid_id}') - with locked_file(fn, 'a', encoding='utf-8') as archive_file: - archive_file.write(vid_id + '\n') + if is_path_like(fn): + with locked_file(fn, 'a', encoding='utf-8') as archive_file: + archive_file.write(vid_id + '\n') self.archive.add(vid_id) @staticmethod @@ -3562,11 +3568,17 @@ def _format_note(self, fdict): res += '~' + format_bytes(fdict['filesize_approx']) return res - def render_formats_table(self, info_dict): - if not info_dict.get('formats') and not info_dict.get('url'): - return None + def _get_formats(self, info_dict): + if info_dict.get('formats') is None: + if info_dict.get('url') and info_dict.get('_type', 'video') == 'video': + return [info_dict] + return [] + return info_dict['formats'] - formats = info_dict.get('formats', [info_dict]) + def render_formats_table(self, info_dict): + formats = self._get_formats(info_dict) + if not formats: + return if not self.params.get('listformats_table', True) is not False: table = [ [ @@ -3574,7 +3586,7 @@ def render_formats_table(self, info_dict): format_field(f, 'ext'), self.format_resolution(f), self._format_note(f) - ] for f in formats if f.get('preference') is None or f['preference'] >= -1000] + ] for f in formats if (f.get('preference') or 0) >= -1000] return render_table(['format code', 'extension', 'resolution', 'note'], table, extra_gap=1) def simplified_codec(f, field): @@ -3633,7 +3645,7 @@ def render_thumbnails_table(self, info_dict): return None return render_table( self._list_format_headers('ID', 'Width', 'Height', 'URL'), - [[t.get('id'), t.get('width', 'unknown'), t.get('height', 'unknown'), t['url']] for t in thumbnails]) + [[t.get('id'), t.get('width') or 'unknown', t.get('height') or 'unknown', t['url']] for t in thumbnails]) def render_subtitles_table(self, video_id, subtitles): def _row(lang, formats): @@ -3676,6 +3688,8 @@ def print_debug_header(self): if not self.params.get('verbose'): return + from . import _IN_CLI # Must be delayed import + # These imports can be slow. So import them only as needed from .extractor.extractors import _LAZY_LOADER from .extractor.extractors import _PLUGIN_CLASSES as plugin_extractors @@ -3712,6 +3726,7 @@ def get_encoding(stream): __version__, f'[{RELEASE_GIT_HEAD}]' if RELEASE_GIT_HEAD else '', '' if source == 'unknown' else f'({source})', + '' if _IN_CLI else 'API', delim=' ')) if not _LAZY_LOADER: if os.environ.get('YTDLP_NO_LAZY_EXTRACTORS'):