]> jfr.im git - yt-dlp.git/blobdiff - yt_dlp/YoutubeDL.py
[ie/loom] Add extractors (#8686)
[yt-dlp.git] / yt_dlp / YoutubeDL.py
index e65bef862c5025e20054da82ec7cd5cdc3e3c5ee..e3d1db3761e84c3dc5bd8b2bbe0af6b431f88911 100644 (file)
@@ -40,9 +40,9 @@
     NoSupportingHandlers,
     RequestError,
     SSLError,
-    _CompatHTTPError,
     network_exceptions,
 )
+from .networking.impersonate import ImpersonateRequestHandler
 from .plugins import directories as plugin_directories
 from .postprocessor import _PLUGIN_CLASSES as plugin_pps
 from .postprocessor import (
     get_postprocessor,
 )
 from .postprocessor.ffmpeg import resolve_mapping as resolve_recode_mapping
-from .update import REPOSITORY, _get_system_deprecation, _make_label, current_git_head, detect_variant
+from .update import (
+    REPOSITORY,
+    _get_system_deprecation,
+    _make_label,
+    current_git_head,
+    detect_variant,
+)
 from .utils import (
     DEFAULT_OUTTMPL,
     IDENTITY,
     SameFileError,
     UnavailableVideoError,
     UserNotLive,
+    YoutubeDLError,
     age_restricted,
     args_to_str,
     bug_reports_message,
@@ -397,6 +404,8 @@ class YoutubeDL:
                        - "detect_or_warn": check whether we can do anything
                                            about it, warn otherwise (default)
     source_address:    Client-side IP address to bind to.
+    impersonate:       Client to impersonate for requests.
+                       An ImpersonateTarget (from yt_dlp.networking.impersonate)
     sleep_interval_requests: Number of seconds to sleep between requests
                        during extraction
     sleep_interval:    Number of seconds to sleep before each download when
@@ -570,11 +579,18 @@ class YoutubeDL:
         'url', 'manifest_url', 'manifest_stream_number', 'ext', 'format', 'format_id', 'format_note',
         'width', 'height', 'aspect_ratio', 'resolution', 'dynamic_range', 'tbr', 'abr', 'acodec', 'asr', 'audio_channels',
         'vbr', 'fps', 'vcodec', 'container', 'filesize', 'filesize_approx', 'rows', 'columns',
-        'player_url', 'protocol', 'fragment_base_url', 'fragments', 'is_from_start',
+        'player_url', 'protocol', 'fragment_base_url', 'fragments', 'is_from_start', 'is_dash_periods', 'request_data',
         'preference', 'language', 'language_preference', 'quality', 'source_preference', 'cookies',
         'http_headers', 'stretched_ratio', 'no_resume', 'has_drm', 'extra_param_to_segment_url', 'hls_aes', 'downloader_options',
         'page_url', 'app', 'play_path', 'tc_url', 'flash_version', 'rtmp_live', 'rtmp_conn', 'rtmp_protocol', 'rtmp_real_time'
     }
+    _deprecated_multivalue_fields = {
+        'album_artist': 'album_artists',
+        'artist': 'artists',
+        'composer': 'composers',
+        'creator': 'creators',
+        'genre': 'genres',
+    }
     _format_selection_exts = {
         'audio': set(MEDIA_EXTENSIONS.common_audio),
         'video': set(MEDIA_EXTENSIONS.common_video + ('3gp', )),
@@ -678,7 +694,6 @@ def process_color_policy(stream):
         self.params['http_headers'] = HTTPHeaderDict(std_headers, self.params.get('http_headers'))
         self._load_cookies(self.params['http_headers'].get('Cookie'))  # compat
         self.params['http_headers'].pop('Cookie', None)
-        self._request_director = self.build_request_director(_REQUEST_HANDLERS.values(), _RH_PREFERENCES)
 
         if auto_init and auto_init != 'no_verbose_header':
             self.print_debug_header()
@@ -702,6 +717,13 @@ def check_deprecated(param, option, suggestion):
         for msg in self.params.get('_deprecation_warnings', []):
             self.deprecated_feature(msg)
 
+        if impersonate_target := self.params.get('impersonate'):
+            if not self._impersonate_target_available(impersonate_target):
+                raise YoutubeDLError(
+                    f'Impersonate target "{impersonate_target}" is not available. '
+                    f'Use --list-impersonate-targets to see available targets. '
+                    f'You may be missing dependencies required to support this target.')
+
         if 'list-formats' in self.params['compat_opts']:
             self.params['listformats_table'] = False
 
@@ -951,7 +973,9 @@ def __exit__(self, *args):
 
     def close(self):
         self.save_cookies()
-        self._request_director.close()
+        if '_request_director' in self.__dict__:
+            self._request_director.close()
+            del self._request_director
 
     def trouble(self, message=None, tb=None, is_error=True):
         """Determine action to take when a download problem appears.
@@ -1179,6 +1203,7 @@ def prepare_outtmpl(self, outtmpl, info_dict, sanitize=False):
         MATH_FUNCTIONS = {
             '+': float.__add__,
             '-': float.__sub__,
+            '*': float.__mul__,
         }
         # Field is of the form key1.key2...
         # where keys (except first) can be string, int, slice or "{field, ...}"
@@ -1200,6 +1225,15 @@ def prepare_outtmpl(self, outtmpl, info_dict, sanitize=False):
                 (?:\|(?P<default>.*?))?
             )$''')
 
+        def _from_user_input(field):
+            if field == ':':
+                return ...
+            elif ':' in field:
+                return slice(*map(int_or_none, field.split(':')))
+            elif int_or_none(field) is not None:
+                return int(field)
+            return field
+
         def _traverse_infodict(fields):
             fields = [f for x in re.split(r'\.({.+?})\.?', fields)
                       for f in ([x] if x.startswith('{') else x.split('.'))]
@@ -1209,11 +1243,12 @@ def _traverse_infodict(fields):
 
             for i, f in enumerate(fields):
                 if not f.startswith('{'):
+                    fields[i] = _from_user_input(f)
                     continue
                 assert f.endswith('}'), f'No closing brace for {f} in {fields}'
-                fields[i] = {k: k.split('.') for k in f[1:-1].split(',')}
+                fields[i] = {k: list(map(_from_user_input, k.split('.'))) for k in f[1:-1].split(',')}
 
-            return traverse_obj(info_dict, fields, is_user_input=True, traverse_string=True)
+            return traverse_obj(info_dict, fields, traverse_string=True)
 
         def get_value(mdict):
             # Object traversal
@@ -2203,7 +2238,7 @@ def _parse_format_selection(tokens, inside_merge=False, inside_choice=False, ins
             selectors = []
             current_selector = None
             for type, string_, start, _, _ in tokens:
-                # ENCODING is only defined in python 3.x
+                # ENCODING is only defined in Python 3.x
                 if type == getattr(tokenize, 'ENCODING', None):
                     continue
                 elif type in [tokenize.NAME, tokenize.NUMBER]:
@@ -2435,7 +2470,7 @@ def selector_function(ctx):
                                 # for extractors with incomplete formats (audio only (soundcloud)
                                 # or video only (imgur)) best/worst will fallback to
                                 # best/worst {video,audio}-only format
-                                matches = formats
+                                matches = list(filter(lambda f: f.get('vcodec') != 'none' or f.get('acodec') != 'none', formats))
                             elif seperate_fallback and not ctx['has_merged_format']:
                                 # for compatibility with youtube-dl when there is no pre-merged format
                                 matches = list(filter(seperate_fallback, formats))
@@ -2454,9 +2489,16 @@ def final_selector(ctx):
                 return selector_function(ctx_copy)
             return final_selector
 
-        stream = io.BytesIO(format_spec.encode())
+        # HACK: Python 3.12 changed the underlying parser, rendering '7_a' invalid
+        #       Prefix numbers with random letters to avoid it being classified as a number
+        #       See: https://github.com/yt-dlp/yt-dlp/pulls/8797
+        # TODO: Implement parser not reliant on tokenize.tokenize
+        prefix = ''.join(random.choices(string.ascii_letters, k=32))
+        stream = io.BytesIO(re.sub(r'\d[_\d]*', rf'{prefix}\g<0>', format_spec).encode())
         try:
-            tokens = list(_remove_unused_ops(tokenize.tokenize(stream.readline)))
+            tokens = list(_remove_unused_ops(
+                token._replace(string=token.string.replace(prefix, ''))
+                for token in tokenize.tokenize(stream.readline)))
         except tokenize.TokenError:
             raise syntax_error('Missing closing/opening brackets or parenthesis', (0, len(format_spec)))
 
@@ -2617,6 +2659,15 @@ def _fill_common_fields(self, info_dict, final=True):
             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])
 
+        for old_key, new_key in self._deprecated_multivalue_fields.items():
+            if new_key in info_dict and old_key in info_dict:
+                if '_version' not in info_dict:  # HACK: Do not warn when using --load-info-json
+                    self.deprecation_warning(f'Do not return {old_key!r} when {new_key!r} is present')
+            elif old_value := info_dict.get(old_key):
+                info_dict[new_key] = old_value.split(', ')
+            elif new_value := info_dict.get(new_key):
+                info_dict[old_key] = ', '.join(v.replace(',', '\N{FULLWIDTH COMMA}') for v in new_value)
+
     def _raise_pending_errors(self, info):
         err = info.pop('__pending_error', None)
         if err:
@@ -3460,7 +3511,8 @@ def ffmpeg_fixup(cndn, msg, cls):
                                      or info_dict.get('is_live') and self.params.get('hls_use_mpegts') is None,
                                      'Possible MPEG-TS in MP4 container or malformed AAC timestamps',
                                      FFmpegFixupM3u8PP)
-                        ffmpeg_fixup(info_dict.get('is_live') and downloader == 'dashsegments',
+                        ffmpeg_fixup(downloader == 'dashsegments'
+                                     and (info_dict.get('is_live') or info_dict.get('is_dash_periods')),
                                      'Possible duplicate MOOV atoms', FFmpegFixupDuplicateMoovPP)
 
                     ffmpeg_fixup(downloader == 'web_socket_fragment', 'Malformed timestamps detected', FFmpegFixupTimestampPP)
@@ -3537,6 +3589,8 @@ def download_with_info_file(self, info_filename):
                     raise
                 self.report_warning(f'The info failed to download: {e}; trying with URL {webpage_url}')
                 self.download([webpage_url])
+            except ExtractorError as e:
+                self.report_error(e)
         return self._download_retcode
 
     @staticmethod
@@ -4034,6 +4088,22 @@ def _opener(self):
         handler = self._request_director.handlers['Urllib']
         return handler._get_instance(cookiejar=self.cookiejar, proxies=self.proxies)
 
+    def _get_available_impersonate_targets(self):
+        # todo(future): make available as public API
+        return [
+            (target, rh.RH_NAME)
+            for rh in self._request_director.handlers.values()
+            if isinstance(rh, ImpersonateRequestHandler)
+            for target in rh.supported_targets
+        ]
+
+    def _impersonate_target_available(self, target):
+        # todo(future): make available as public API
+        return any(
+            rh.is_supported_target(target)
+            for rh in self._request_director.handlers.values()
+            if isinstance(rh, ImpersonateRequestHandler))
+
     def urlopen(self, req):
         """ Start an HTTP download """
         if isinstance(req, str):
@@ -4065,9 +4135,13 @@ def urlopen(self, req):
                     raise RequestError(
                         'file:// URLs are disabled by default in yt-dlp for security reasons. '
                         'Use --enable-file-urls to enable at your own risk.', cause=ue) from ue
-                if 'unsupported proxy type: "https"' in ue.msg.lower():
+                if (
+                    'unsupported proxy type: "https"' in ue.msg.lower()
+                    and 'requests' not in self._request_director.handlers
+                    and 'curl_cffi' not in self._request_director.handlers
+                ):
                     raise RequestError(
-                        'To use an HTTPS proxy for this request, one of the following dependencies needs to be installed: requests')
+                        'To use an HTTPS proxy for this request, one of the following dependencies needs to be installed: requests, curl_cffi')
 
                 elif (
                     re.match(r'unsupported url scheme: "wss?"', ue.msg.lower())
@@ -4077,6 +4151,13 @@ def urlopen(self, req):
                         'This request requires WebSocket support. '
                         'Ensure one of the following dependencies are installed: websockets',
                         cause=ue) from ue
+
+                elif re.match(r'unsupported (?:extensions: impersonate|impersonate target)', ue.msg.lower()):
+                    raise RequestError(
+                        f'Impersonate target "{req.extensions["impersonate"]}" is not available.'
+                        f' See --list-impersonate-targets for available targets.'
+                        f' This request requires browser impersonation, however you may be missing dependencies'
+                        f' required to support this target.')
             raise
         except SSLError as e:
             if 'UNSAFE_LEGACY_RENEGOTIATION_DISABLED' in str(e):
@@ -4086,8 +4167,6 @@ def urlopen(self, req):
                     'SSLV3_ALERT_HANDSHAKE_FAILURE: The server may not support the current cipher list. '
                     'Try using --legacy-server-connect', cause=e) from e
             raise
-        except HTTPError as e:  # TODO: Remove in a future release
-            raise _CompatHTTPError(e) from e
 
     def build_request_director(self, handlers, preferences=None):
         logger = _YDLLogger(self)
@@ -4111,6 +4190,7 @@ def build_request_director(self, handlers, preferences=None):
                     'timeout': 'socket_timeout',
                     'legacy_ssl_support': 'legacyserverconnect',
                     'enable_file_urls': 'enable_file_urls',
+                    'impersonate': 'impersonate',
                     'client_cert': {
                         'client_certificate': 'client_certificate',
                         'client_certificate_key': 'client_certificate_key',
@@ -4123,6 +4203,10 @@ def build_request_director(self, handlers, preferences=None):
             director.preferences.add(lambda rh, _: 500 if rh.RH_KEY == 'Urllib' else 0)
         return director
 
+    @functools.cached_property
+    def _request_director(self):
+        return self.build_request_director(_REQUEST_HANDLERS.values(), _RH_PREFERENCES)
+
     def encode(self, s):
         if isinstance(s, bytes):
             return s  # Already encoded