X-Git-Url: http://git.bitcoin.ninja/index.cgi?a=blobdiff_plain;ds=sidebyside;f=youtube_dl%2Fextractor%2Fcommon.py;h=b9014fc23e53eaf335d65ee56c3db560d218d642;hb=4b0f45f6675f13c8226ddcef8d1a77ef7739e54a;hp=3ae5d52127f1014b235447692bb4ed24615178bd;hpb=2ddcd88129e633ead8c5faee6c930b66dddb1b32;p=youtube-dl diff --git a/youtube_dl/extractor/common.py b/youtube_dl/extractor/common.py index 3ae5d5212..b9014fc23 100644 --- a/youtube_dl/extractor/common.py +++ b/youtube_dl/extractor/common.py @@ -22,18 +22,20 @@ from ..compat import ( compat_str, ) from ..utils import ( + NO_DEFAULT, age_restricted, bug_reports_message, clean_html, compiled_regex_type, + determine_ext, ExtractorError, + fix_xml_ampersands, float_or_none, int_or_none, RegexNotFoundError, sanitize_filename, unescapeHTML, ) -_NO_DEFAULT = object() class InfoExtractor(object): @@ -523,7 +525,7 @@ class InfoExtractor(object): video_info['description'] = playlist_description return video_info - def _search_regex(self, pattern, string, name, default=_NO_DEFAULT, fatal=True, flags=0, group=None): + def _search_regex(self, pattern, string, name, default=NO_DEFAULT, fatal=True, flags=0, group=None): """ Perform a regex search on the given string, using a single or a list of patterns returning the first matching group. @@ -549,7 +551,7 @@ class InfoExtractor(object): return next(g for g in mobj.groups() if g is not None) else: return mobj.group(group) - elif default is not _NO_DEFAULT: + elif default is not NO_DEFAULT: return default elif fatal: raise RegexNotFoundError('Unable to extract %s' % _name) @@ -557,7 +559,7 @@ class InfoExtractor(object): self._downloader.report_warning('unable to extract %s' % _name + bug_reports_message()) return None - def _html_search_regex(self, pattern, string, name, default=_NO_DEFAULT, fatal=True, flags=0, group=None): + def _html_search_regex(self, pattern, string, name, default=NO_DEFAULT, fatal=True, flags=0, group=None): """ Like _search_regex, but strips HTML tags and unescapes entities. """ @@ -705,6 +707,25 @@ class InfoExtractor(object): return self._html_search_meta('twitter:player', html, 'twitter card player') + @staticmethod + def _hidden_inputs(html): + return dict([ + (input.group('name'), input.group('value')) for input in re.finditer( + r'''(?x) + ["\'])hidden(?P=q_hidden)\s+ + name=(?P["\'])(?P.+?)(?P=q_name)\s+ + (?:id=(?P["\']).+?(?P=q_id)\s+)? + value=(?P["\'])(?P.*?)(?P=q_value) + ''', html) + ]) + + def _form_hidden_inputs(self, form_id, html): + form = self._search_regex( + r'(?s)]+?id=(["\'])%s\1[^>]*>(?P
.+?)
' % form_id, + html, '%s form' % form_id, group='form') + return self._hidden_inputs(form) + def _sort_formats(self, formats, field_preference=None): if not formats: raise ExtractorError('No video formats found') @@ -764,7 +785,7 @@ class InfoExtractor(object): f.get('fps') if f.get('fps') is not None else -1, f.get('filesize_approx') if f.get('filesize_approx') is not None else -1, f.get('source_preference') if f.get('source_preference') is not None else -1, - f.get('format_id'), + f.get('format_id') if f.get('format_id') is not None else '', ) formats.sort(key=_formats_key) @@ -786,8 +807,8 @@ class InfoExtractor(object): return True except ExtractorError as e: if isinstance(e.cause, compat_HTTPError): - self.report_warning( - '%s URL is invalid, skipping' % item, video_id) + self.to_screen( + '%s: %s URL is invalid, skipping' % (video_id, item)) return False raise @@ -815,10 +836,14 @@ class InfoExtractor(object): self.to_screen(msg) time.sleep(timeout) - def _extract_f4m_formats(self, manifest_url, video_id, preference=None, f4m_id=None): + def _extract_f4m_formats(self, manifest_url, video_id, preference=None, f4m_id=None, + transform_source=lambda s: fix_xml_ampersands(s).strip()): manifest = self._download_xml( manifest_url, video_id, 'Downloading f4m manifest', - 'Unable to download f4m manifest') + 'Unable to download f4m manifest', + # Some manifests may be malformed, e.g. prosiebensat1 generated manifests + # (see https://github.com/rg3/youtube-dl/issues/6215#issuecomment-121704244) + transform_source=transform_source) formats = [] manifest_version = '1.0' @@ -828,8 +853,19 @@ class InfoExtractor(object): media_nodes = manifest.findall('{http://ns.adobe.com/f4m/2.0}media') for i, media_el in enumerate(media_nodes): if manifest_version == '2.0': - manifest_url = ('/'.join(manifest_url.split('/')[:-1]) + '/' + - (media_el.attrib.get('href') or media_el.attrib.get('url'))) + media_url = media_el.attrib.get('href') or media_el.attrib.get('url') + if not media_url: + continue + manifest_url = ( + media_url if media_url.startswith('http://') or media_url.startswith('https://') + else ('/'.join(manifest_url.split('/')[:-1]) + '/' + media_url)) + # If media_url is itself a f4m manifest do the recursive extraction + # since bitrates in parent manifest (this one) and media_url manifest + # may differ leading to inability to resolve the format by requested + # bitrate in f4m downloader + if determine_ext(manifest_url) == 'f4m': + formats.extend(self._extract_f4m_formats(manifest_url, video_id, preference, f4m_id)) + continue tbr = int_or_none(media_el.attrib.get('bitrate')) formats.append({ 'format_id': '-'.join(filter(None, [f4m_id, compat_str(i if tbr is None else tbr)])), @@ -846,7 +882,8 @@ class InfoExtractor(object): def _extract_m3u8_formats(self, m3u8_url, video_id, ext=None, entry_protocol='m3u8', preference=None, - m3u8_id=None): + m3u8_id=None, note=None, errnote=None, + fatal=True): formats = [{ 'format_id': '-'.join(filter(None, [m3u8_id, 'meta'])), @@ -865,8 +902,11 @@ class InfoExtractor(object): m3u8_doc = self._download_webpage( m3u8_url, video_id, - note='Downloading m3u8 information', - errnote='Failed to download m3u8 information') + note=note or 'Downloading m3u8 information', + errnote=errnote or 'Failed to download m3u8 information', + fatal=fatal) + if m3u8_doc is False: + return m3u8_doc last_info = None last_media = None kv_rex = re.compile( @@ -896,7 +936,7 @@ class InfoExtractor(object): format_id = [] if m3u8_id: format_id.append(m3u8_id) - last_media_name = last_media.get('NAME') if last_media else None + last_media_name = last_media.get('NAME') if last_media and last_media.get('TYPE') != 'SUBTITLES' else None format_id.append(last_media_name if last_media_name else '%d' % (tbr if tbr else len(formats))) f = { 'format_id': '-'.join(format_id), @@ -956,7 +996,7 @@ class InfoExtractor(object): def _parse_smil_video(self, video, video_id, base, rtmp_count): src = video.get('src') if not src: - return ([], rtmp_count) + return [], rtmp_count bitrate = int_or_none(video.get('system-bitrate') or video.get('systemBitrate'), 1000) width = int_or_none(video.get('width')) height = int_or_none(video.get('height')) @@ -969,7 +1009,7 @@ class InfoExtractor(object): proto = 'http' ext = video.get('ext') if proto == 'm3u8': - return (self._extract_m3u8_formats(src, video_id, ext), rtmp_count) + return self._extract_m3u8_formats(src, video_id, ext), rtmp_count elif proto == 'rtmp': rtmp_count += 1 streamer = video.get('streamer') or base @@ -1072,9 +1112,6 @@ class InfoExtractor(object): def _get_automatic_captions(self, *args, **kwargs): raise NotImplementedError("This method must be implemented by subclasses") - def _subtitles_timecode(self, seconds): - return '%02d:%02d:%02d.%03d' % (seconds / 3600, (seconds % 3600) / 60, seconds % 60, (seconds % 1) * 1000) - class SearchInfoExtractor(InfoExtractor): """