Merge d7b502a727 into 4d05f84325

* no space after @ in decorator

test/test_jsinterp.py

@@ -577,9 +577,11 @@ class TestJSInterpreter(unittest.TestCase):
     def test_unary_operators(self):
         jsi = JSInterpreter('function f(){return 2  -  - - 2;}')
         self.assertEqual(jsi.call_function('f'), 0)
-        # fails
-        # jsi = JSInterpreter('function f(){return 2 + - + - - 2;}')
-        # self.assertEqual(jsi.call_function('f'), 0)
+        jsi = JSInterpreter('function f(){return 2 + - + - - 2;}')
+        self.assertEqual(jsi.call_function('f'), 0)
+        # https://github.com/ytdl-org/youtube-dl/issues/32815
+        jsi = JSInterpreter('function f(){return 0  - 7 * - 6;}')
+        self.assertEqual(jsi.call_function('f'), 42)
 
     """ # fails so far
     def test_packed(self):

test/test_youtube_signature.py

@@ -158,6 +158,10 @@ _NSIG_TESTS = [
         'https://www.youtube.com/s/player/b7910ca8/player_ias.vflset/en_US/base.js',
         '_hXMCwMt9qE310D', 'LoZMgkkofRMCZQ',
     ),
+    (
+        'https://www.youtube.com/s/player/590f65a6/player_ias.vflset/en_US/base.js',
+        '1tm7-g_A9zsI8_Lay_', 'xI4Vem4Put_rOg',
+    ),
 ]
 
 

youtube_dl/compat.py

@@ -2448,12 +2448,12 @@ try:
 except ImportError:
     import BaseHTTPServer as compat_http_server
 
+# urllib.parse
 try:
     from urllib.parse import unquote_to_bytes as compat_urllib_parse_unquote_to_bytes
     from urllib.parse import unquote as compat_urllib_parse_unquote
     from urllib.parse import unquote_plus as compat_urllib_parse_unquote_plus
     from urllib.parse import urlencode as compat_urllib_parse_urlencode
-    from urllib.parse import parse_qs as compat_parse_qs
 except ImportError:  # Python 2
     _asciire = (compat_urllib_parse._asciire if hasattr(compat_urllib_parse, '_asciire')
                 else re.compile(r'([\x00-\x7f]+)'))
@@ -2543,60 +2543,80 @@ except ImportError:  # Python 2
 
         return compat_urllib_parse._urlencode(encode_elem(query), doseq=doseq)
 
-    # HACK: The following is the correct parse_qs implementation from cpython 3's stdlib.
-    # Python 2's version is apparently totally broken
-    def _parse_qsl(qs, keep_blank_values=False, strict_parsing=False,
-                   encoding='utf-8', errors='replace'):
-        qs, _coerce_result = qs, compat_str
-        pairs = [s2 for s1 in qs.split('&') for s2 in s1.split(';')]
-        r = []
-        for name_value in pairs:
-            if not name_value and not strict_parsing:
-                continue
-            nv = name_value.split('=', 1)
-            if len(nv) != 2:
-                if strict_parsing:
-                    raise ValueError('bad query field: %r' % (name_value,))
-                # Handle case of a control-name with no equal sign
-                if keep_blank_values:
-                    nv.append('')
-                else:
-                    continue
-            if len(nv[1]) or keep_blank_values:
-                name = nv[0].replace('+', ' ')
-                name = compat_urllib_parse_unquote(
-                    name, encoding=encoding, errors=errors)
-                name = _coerce_result(name)
-                value = nv[1].replace('+', ' ')
-                value = compat_urllib_parse_unquote(
-                    value, encoding=encoding, errors=errors)
-                value = _coerce_result(value)
-                r.append((name, value))
-        return r
-
-    def compat_parse_qs(qs, keep_blank_values=False, strict_parsing=False,
-                        encoding='utf-8', errors='replace'):
-        parsed_result = {}
-        pairs = _parse_qsl(qs, keep_blank_values, strict_parsing,
-                           encoding=encoding, errors=errors)
-        for name, value in pairs:
-            if name in parsed_result:
-                parsed_result[name].append(value)
-            else:
-                parsed_result[name] = [value]
-        return parsed_result
-
     setattr(compat_urllib_parse, '_urlencode',
             getattr(compat_urllib_parse, 'urlencode'))
     for name, fix in (
             ('unquote_to_bytes', compat_urllib_parse_unquote_to_bytes),
             ('parse_unquote', compat_urllib_parse_unquote),
             ('unquote_plus', compat_urllib_parse_unquote_plus),
-            ('urlencode', compat_urllib_parse_urlencode),
-            ('parse_qs', compat_parse_qs)):
+            ('urlencode', compat_urllib_parse_urlencode)):
         setattr(compat_urllib_parse, name, fix)
+finally:
+    try:
+        # arguments changed in 3.8 and 3.10
+        from urllib.parse import parse_qs as _parse_qs
+        _parse_qs('a=b', separator='&')
+        compat_parse_qs = _parse_qs
+    except (ImportError, TypeError):  # Python 2, < 3.10
 
-compat_urllib_parse_parse_qs = compat_parse_qs
+        # HACK: The following is the correct parse_qs implementation from cpython 3's stdlib.
+        # Python 2's version is apparently totally broken
+        # Also use this implementation for Py < 3.10
+        # * support only default separator '&', not r'[&;]', like 3.10+
+        # * support max_num_fields, like 3.8+
+        def _parse_qsl(qs, keep_blank_values=False, strict_parsing=False,
+                       encoding='utf-8', errors='replace',
+                       max_num_fields=None, separator='&'):
+            if not isinstance(separator, (compat_str, str)):
+                raise ValueError('Separator must be of type string or bytes')
+            # DoS protection, if anyone cares
+            if qs and max_num_fields is not None and qs.count(separator) >= max_num_fields:
+                raise ValueError('Too many fields')
+            _coerce_result = compat_str
+            r = []
+            for name_value in qs.split(separator):
+                if not name_value and not strict_parsing:
+                    continue
+                nv = name_value.split('=', 1)
+                if len(nv) != 2:
+                    if strict_parsing:
+                        raise ValueError('bad query field: %r' % (name_value,))
+                    # Handle case of a control-name with no equal sign
+                    if keep_blank_values:
+                        nv.append('')
+                    else:
+                        continue
+                if len(nv[1]) or keep_blank_values:
+                    name = nv[0].replace('+', ' ')
+                    name = compat_urllib_parse_unquote(
+                        name, encoding=encoding, errors=errors)
+                    name = _coerce_result(name)
+                    value = nv[1].replace('+', ' ')
+                    value = compat_urllib_parse_unquote(
+                        value, encoding=encoding, errors=errors)
+                    value = _coerce_result(value)
+                    r.append((name, value))
+            return r
+
+        def compat_parse_qs(qs, keep_blank_values=False, strict_parsing=False,
+                            encoding='utf-8', errors='replace',
+                            max_num_fields=None, separator='&'):
+            parsed_result = {}
+            pairs = _parse_qsl(qs, keep_blank_values, strict_parsing,
+                               encoding, errors, max_num_fields, separator)
+            for name, value in pairs:
+                if name in parsed_result:
+                    parsed_result[name].append(value)
+                else:
+                    parsed_result[name] = [value]
+            return parsed_result
+
+        for name, fix in (
+                ('parse_qs', compat_parse_qs),
+                ('parse_qsl', _parse_qsl)):
+            setattr(compat_urllib_parse, name, fix)
+
+    compat_urllib_parse_parse_qs = compat_parse_qs
 
 try:
     from urllib.request import DataHandler as compat_urllib_request_DataHandler

youtube_dl/extractor/bfi.py

@@ -4,7 +4,12 @@ from __future__ import unicode_literals
 import re
 
 from .common import InfoExtractor
-from ..utils import extract_attributes
+from ..utils import (
+    extract_attributes,
+    parse_qs,
+    remove_start,
+    smuggle_url,
+)
 
 
 class BFIPlayerIE(InfoExtractor):
@@ -12,26 +17,39 @@ class BFIPlayerIE(InfoExtractor):
     _VALID_URL = r'https?://player\.bfi\.org\.uk/[^/]+/film/watch-(?P<id>[\w-]+)-online'
     _TEST = {
         'url': 'https://player.bfi.org.uk/free/film/watch-computer-doctor-1974-online',
-        'md5': 'e8783ebd8e061ec4bc6e9501ed547de8',
+        'md5': '15598bdd6a413ce9363970754f054d76',
         'info_dict': {
             'id': 'htNnhlZjE60C9VySkQEIBtU-cNV1Xx63',
             'ext': 'mp4',
             'title': 'Computer Doctor',
             'description': 'md5:fb6c240d40c4dbe40428bdd62f78203b',
+            'timestamp': 1564424975,
+            'upload_date': '20190729',
+            'uploader_id': '6057949427001',
         },
-        'skip': 'BFI Player films cannot be played outside of the UK',
+        # 'skip': 'BFI Player films cannot be played outside of the UK',
     }
+    _BRIGHTCOVE_ACCOUNT_ID = '6057949427001'
 
     def _real_extract(self, url):
         video_id = self._match_id(url)
         webpage = self._download_webpage(url, video_id)
-        entries = []
-        for player_el in re.findall(r'(?s)<[^>]+class="player"[^>]*>', webpage):
-            player_attr = extract_attributes(player_el)
-            ooyala_id = player_attr.get('data-video-id')
-            if not ooyala_id:
-                continue
-            entries.append(self.url_result(
-                'ooyala:' + ooyala_id, 'Ooyala',
-                ooyala_id, player_attr.get('data-label')))
-        return self.playlist_result(entries)
+
+        film_only = 'play-film' in parse_qs(url, keep_blank_values=True)
+
+        def entries():
+            for player_el in re.finditer(r'(?s)<video-js\b[^>]+>', webpage):
+                player_attr = extract_attributes(player_el.group(0))
+                bcv_id, account_id, player_id, embed = (
+                    player_attr.get(x) for x in ('data-ref-id', 'data-acid', 'data-pid', 'data-embed'))
+                if not bcv_id:
+                    continue
+                if film_only and player_attr.get('data-video-type') != 'film':
+                    continue
+                bc_url = 'brightcove:new:%s:%s:%s:video:ref:%s' % (
+                    account_id or self._BRIGHTCOVE_ACCOUNT_ID, player_id or 'default', embed or 'default', bcv_id)
+
+                yield self.url_result(smuggle_url(
+                    bc_url, {'referrer': url, 'force_videoid': remove_start(bcv_id, 'ref:')}), ie='BrightcoveNew', video_id=video_id)
+
+        return self.playlist_result(entries())

youtube_dl/extractor/brightcove.py

@@ -340,7 +340,7 @@ class BrightcoveLegacyIE(InfoExtractor):
 
 class BrightcoveNewIE(AdobePassIE):
     IE_NAME = 'brightcove:new'
-    _VALID_URL = r'https?://players\.brightcove\.net/(?P<account_id>\d+)/(?P<player_id>[^/]+)_(?P<embed>[^/]+)/index\.html\?.*(?P<content_type>video|playlist)Id=(?P<video_id>\d+|ref:[^&]+)'
+    _VALID_URL = r'(?:brightcove:new|(?P<u>https?)):(?(u)//players\.brightcove\.net/)(?P<account_id>\d+)(?(u)/|:)(?P<player_id>[^/]+)(?(u)_|:)(?P<embed>[^/]+)(?(u)/index\.html\?.*|:)(?P<content_type>video|playlist)(?(u)Id=|:)(?P<video_id>\d+|ref:[^&]+)'
     _TESTS = [{
         'url': 'http://players.brightcove.net/929656772001/e41d32dc-ec74-459e-a845-6c69f7b724ea_default/index.html?videoId=4463358922001',
         'md5': 'c8100925723840d4b0d243f7025703be',
@@ -593,7 +593,7 @@ class BrightcoveNewIE(AdobePassIE):
             'ip_blocks': smuggled_data.get('geo_ip_blocks'),
         })
 
-        account_id, player_id, embed, content_type, video_id = re.match(self._VALID_URL, url).groups()
+        account_id, player_id, embed, content_type, video_id = re.match(self._VALID_URL, url).groups()[1:]
 
         policy_key_id = '%s_%s' % (account_id, player_id)
         policy_key = self._downloader.cache.load('brightcove', policy_key_id)
@@ -678,4 +678,4 @@ class BrightcoveNewIE(AdobePassIE):
                 json_data.get('description'))
 
         return self._parse_brightcove_metadata(
-            json_data, video_id, headers=headers)
+            json_data, smuggled_data.get('force_videoid') or video_id, headers=headers)

youtube_dl/extractor/common.py

@@ -3033,7 +3033,6 @@ class InfoExtractor(object):
             transform_source=transform_source, default=None)
 
     def _extract_jwplayer_data(self, webpage, video_id, *args, **kwargs):
-
         # allow passing `transform_source` through to _find_jwplayer_data()
         transform_source = kwargs.pop('transform_source', None)
         kwfind = compat_kwargs({'transform_source': transform_source}) if transform_source else {}

youtube_dl/extractor/palcomp3.py

@@ -8,7 +8,7 @@ from ..compat import compat_str
 from ..utils import (
     int_or_none,
     str_or_none,
-    try_get,
+    traverse_obj,
 )
 
 
@@ -109,7 +109,7 @@ class PalcoMP3ArtistIE(PalcoMP3BaseIE):
     }
     name'''
 
-    @ classmethod
+    @classmethod
     def suitable(cls, url):
         return False if re.match(PalcoMP3IE._VALID_URL, url) else super(PalcoMP3ArtistIE, cls).suitable(url)
 
@@ -118,7 +118,8 @@ class PalcoMP3ArtistIE(PalcoMP3BaseIE):
         artist = self._call_api(artist_slug, self._ARTIST_FIELDS_TMPL)['artist']
 
         def entries():
-            for music in (try_get(artist, lambda x: x['musics']['nodes'], list) or []):
+            for music in traverse_obj(artist, (
+                    'musics', 'nodes', lambda _, m: m['musicID'])):
                 yield self._parse_music(music)
 
         return self.playlist_result(
@@ -137,7 +138,7 @@ class PalcoMP3VideoIE(PalcoMP3BaseIE):
             'title': 'Maiara e Maraisa - Você Faz Falta Aqui - DVD Ao Vivo Em Campo Grande',
             'description': 'md5:7043342c09a224598e93546e98e49282',
             'upload_date': '20161107',
-            'uploader_id': 'maiaramaraisaoficial',
+            'uploader_id': '@maiaramaraisaoficial',
             'uploader': 'Maiara e Maraisa',
         }
     }]

youtube_dl/jsinterp.py

@@ -14,6 +14,7 @@ from .utils import (
     remove_quotes,
     unified_timestamp,
     variadic,
+    write_string,
 )
 from .compat import (
     compat_basestring,
@@ -53,15 +54,16 @@ def wraps_op(op):
 
 # NB In principle NaN cannot be checked by membership.
 # Here all NaN values are actually this one, so _NaN is _NaN,
-# although _NaN != _NaN.
+# although _NaN != _NaN. Ditto Infinity.
 
 _NaN = float('nan')
+_Infinity = float('inf')
 
 
 def _js_bit_op(op):
 
     def zeroise(x):
-        return 0 if x in (None, JS_Undefined, _NaN) else x
+        return 0 if x in (None, JS_Undefined, _NaN, _Infinity) else x
 
     @wraps_op(op)
     def wrapped(a, b):
@@ -84,7 +86,7 @@ def _js_arith_op(op):
 def _js_div(a, b):
     if JS_Undefined in (a, b) or not (a or b):
         return _NaN
-    return operator.truediv(a or 0, b) if b else float('inf')
+    return operator.truediv(a or 0, b) if b else _Infinity
 
 
 def _js_mod(a, b):
@@ -220,6 +222,42 @@ class LocalNameSpace(ChainMap):
         return 'LocalNameSpace%s' % (self.maps, )
 
 
+class Debugger(object):
+    ENABLED = False
+
+    @staticmethod
+    def write(*args, **kwargs):
+        level = kwargs.get('level', 100)
+
+        def truncate_string(s, left, right=0):
+            if s is None or len(s) <= left + right:
+                return s
+            return '...'.join((s[:left - 3], s[-right:] if right else ''))
+
+        write_string('[debug] JS: {0}{1}\n'.format(
+            '  ' * (100 - level),
+            ' '.join(truncate_string(compat_str(x), 50, 50) for x in args)))
+
+    @classmethod
+    def wrap_interpreter(cls, f):
+        def interpret_statement(self, stmt, local_vars, allow_recursion, *args, **kwargs):
+            if cls.ENABLED and stmt.strip():
+                cls.write(stmt, level=allow_recursion)
+            try:
+                ret, should_ret = f(self, stmt, local_vars, allow_recursion, *args, **kwargs)
+            except Exception as e:
+                if cls.ENABLED:
+                    if isinstance(e, ExtractorError):
+                        e = e.orig_msg
+                    cls.write('=> Raises:', e, '<-|', stmt, level=allow_recursion)
+                raise
+            if cls.ENABLED and stmt.strip():
+                if should_ret or not repr(ret) == stmt:
+                    cls.write(['->', '=>'][should_ret], repr(ret), '<-|', stmt, level=allow_recursion)
+            return ret, should_ret
+        return interpret_statement
+
+
 class JSInterpreter(object):
     __named_object_counter = 0
 
@@ -307,8 +345,7 @@ class JSInterpreter(object):
     def __op_chars(cls):
         op_chars = set(';,[')
         for op in cls._all_operators():
-            for c in op[0]:
-                op_chars.add(c)
+            op_chars.update(op[0])
         return op_chars
 
     def _named_object(self, namespace, obj):
@@ -326,9 +363,8 @@ class JSInterpreter(object):
         # collections.Counter() is ~10% slower in both 2.7 and 3.9
         counters = dict((k, 0) for k in _MATCHING_PARENS.values())
         start, splits, pos, delim_len = 0, 0, 0, len(delim) - 1
-        in_quote, escaping, skipping = None, False, 0
-        after_op, in_regex_char_group = True, False
-
+        in_quote, escaping, after_op, in_regex_char_group = None, False, True, False
+        skipping = 0
         for idx, char in enumerate(expr):
             paren_delta = 0
             if not in_quote:
@@ -382,10 +418,12 @@ class JSInterpreter(object):
         return separated[0][1:].strip(), separated[1].strip()
 
     @staticmethod
-    def _all_operators():
-        return itertools.chain(
-            # Ref: https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Operators/Operator_Precedence
-            _SC_OPERATORS, _LOG_OPERATORS, _COMP_OPERATORS, _OPERATORS)
+    def _all_operators(_cached=[]):
+        if not _cached:
+            _cached.extend(itertools.chain(
+                # Ref: https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Operators/Operator_Precedence
+                _SC_OPERATORS, _LOG_OPERATORS, _COMP_OPERATORS, _OPERATORS))
+        return _cached
 
     def _operator(self, op, left_val, right_expr, expr, local_vars, allow_recursion):
         if op in ('||', '&&'):
@@ -416,7 +454,7 @@ class JSInterpreter(object):
         except Exception as e:
             if allow_undefined:
                 return JS_Undefined
-            raise self.Exception('Cannot get index {idx:.100}'.format(**locals()), expr=repr(obj), cause=e)
+            raise self.Exception('Cannot get index {idx!r:.100}'.format(**locals()), expr=repr(obj), cause=e)
 
     def _dump(self, obj, namespace):
         try:
@@ -438,6 +476,7 @@ class JSInterpreter(object):
     _FINALLY_RE = re.compile(r'finally\s*\{')
     _SWITCH_RE = re.compile(r'switch\s*\(')
 
+    @Debugger.wrap_interpreter
     def interpret_statement(self, stmt, local_vars, allow_recursion=100):
         if allow_recursion < 0:
             raise self.Exception('Recursion limit reached')
@@ -511,7 +550,6 @@ class JSInterpreter(object):
                 expr = self._dump(inner, local_vars) + outer
 
         if expr.startswith('('):
-
             m = re.match(r'\((?P<d>[a-z])%(?P<e>[a-z])\.length\+(?P=e)\.length\)%(?P=e)\.length', expr)
             if m:
                 # short-cut eval of frequently used `(d%e.length+e.length)%e.length`, worth ~6% on `pytest -k test_nsig`
@@ -693,7 +731,7 @@ class JSInterpreter(object):
                 (?P<op>{_OPERATOR_RE})?
                 =(?!=)(?P<expr>.*)$
             )|(?P<return>
-                (?!if|return|true|false|null|undefined)(?P<name>{_NAME_RE})$
+                (?!if|return|true|false|null|undefined|NaN|Infinity)(?P<name>{_NAME_RE})$
             )|(?P<indexing>
                 (?P<in>{_NAME_RE})\[(?P<idx>.+)\]$
             )|(?P<attribute>
@@ -727,11 +765,12 @@ class JSInterpreter(object):
             raise JS_Break()
         elif expr == 'continue':
             raise JS_Continue()
-
         elif expr == 'undefined':
             return JS_Undefined, should_return
         elif expr == 'NaN':
             return _NaN, should_return
+        elif expr == 'Infinity':
+            return _Infinity, should_return
 
         elif md.get('return'):
             return local_vars[m.group('name')], should_return
@@ -760,18 +799,28 @@ class JSInterpreter(object):
             right_expr = separated.pop()
             # handle operators that are both unary and binary, minimal BODMAS
             if op in ('+', '-'):
+                # simplify/adjust consecutive instances of these operators
                 undone = 0
                 while len(separated) > 1 and not separated[-1].strip():
                     undone += 1
                     separated.pop()
                 if op == '-' and undone % 2 != 0:
                     right_expr = op + right_expr
+                elif op == '+':
+                    while len(separated) > 1 and separated[-1].strip() in self.OP_CHARS:
+                        right_expr = separated.pop() + right_expr
+                # hanging op at end of left => unary + (strip) or - (push right)
                 left_val = separated[-1]
                 for dm_op in ('*', '%', '/', '**'):
                     bodmas = tuple(self._separate(left_val, dm_op, skip_delims=skip_delim))
                     if len(bodmas) > 1 and not bodmas[-1].strip():
                         expr = op.join(separated) + op + right_expr
-                        right_expr = None
+                        if len(separated) > 1:
+                            separated.pop()
+                            right_expr = op.join((left_val, right_expr))
+                        else:
+                            separated = [op.join((left_val, right_expr))]
+                            right_expr = None
                         break
                 if right_expr is None:
                     continue
@@ -797,6 +846,8 @@ class JSInterpreter(object):
 
             def eval_method():
                 if (variable, member) == ('console', 'debug'):
+                    if Debugger.ENABLED:
+                        Debugger.write(self.interpret_expression('[{}]'.format(arg_str), local_vars, allow_recursion))
                     return
                 types = {
                     'String': compat_str,

youtube_dl/utils.py

@@ -2402,11 +2402,11 @@ class YoutubeDLError(Exception):
 class ExtractorError(YoutubeDLError):
     """Error during info extraction."""
 
-    def __init__(self, msg, tb=None, expected=False, cause=None, video_id=None):
+    def __init__(self, msg, tb=None, expected=False, cause=None, video_id=None, ie=None):
         """ tb, if given, is the original traceback (so that it can be printed out).
         If expected is set, this is a normal error message and most likely not a bug in youtube-dl.
         """
-
+        self.orig_msg = msg
         if sys.exc_info()[0] in (compat_urllib_error.URLError, socket.timeout, UnavailableVideoError):
             expected = True
         if video_id is not None:
@@ -2421,6 +2421,7 @@ class ExtractorError(YoutubeDLError):
         self.exc_info = sys.exc_info()  # preserve original exception
         self.cause = cause
         self.video_id = video_id
+        self.ie = ie
 
     def format_traceback(self):
         if self.traceback is None:
@@ -6561,3 +6562,24 @@ def join_nonempty(*values, **kwargs):
     if from_dict is not None:
         values = (traverse_obj(from_dict, variadic(v)) for v in values)
     return delim.join(map(compat_str, filter(None, values)))
+
+
+class classproperty(object):
+    """property access for class methods with optional caching"""
+    def __new__(cls, *args, **kwargs):
+        func = args[0] if len(args) > 0 else kwargs.get('func')
+        if not func:
+            return functools.partial(cls, *args, **kwargs)
+        return super(classproperty, cls).__new__(cls)
+
+    def __init__(self, func, cache=False):
+        functools.update_wrapper(self, func)
+        self.func = func
+        self._cache = {} if cache else None
+
+    def __get__(self, _, cls):
+        if self._cache is None:
+            return self.func(cls)
+        elif cls not in self._cache:
+            self._cache[cls] = self.func(cls)
+        return self._cache[cls]