Merge 14a087ec62 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/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/extractors.py

@@ -1179,6 +1179,7 @@ from .southpark import (
 from .spankbang import (
     SpankBangIE,
     SpankBangPlaylistIE,
+    SpankBangListIE,
 )
 from .spankwire import SpankwireIE
 from .spiegel import SpiegelIE

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/extractor/spankbang.py

@@ -1,11 +1,17 @@
+# coding: utf-8
 from __future__ import unicode_literals
 
+import itertools
 import re
 
 from .common import InfoExtractor
+from ..compat import compat_urlparse
 from ..utils import (
     determine_ext,
+    extract_attributes,
     ExtractorError,
+    get_element_by_class,
+    get_element_by_id,
     merge_dicts,
     parse_duration,
     parse_resolution,
@@ -26,19 +32,24 @@ class SpankBangIE(InfoExtractor):
                         )
                     '''
     _TESTS = [{
-        'url': 'http://spankbang.com/3vvn/video/fantasy+solo',
-        'md5': '1cc433e1d6aa14bc376535b8679302f7',
+        'url': 'https://spankbang.com/56b3d/video/the+slut+maker+hmv',
+        'md5': '5039ba9d26f6124a7fdea6df2f21e765',
         'info_dict': {
-            'id': '3vvn',
+            'id': '56b3d',
             'ext': 'mp4',
-            'title': 'fantasy solo',
-            'description': 'dillion harper masturbates on a bed',
+            'title': 'The Slut Maker HMV',
+            'description': 'Girls getting converted into cock slaves.',
             'thumbnail': r're:^https?://.*\.jpg$',
-            'uploader': 'silly2587',
-            'timestamp': 1422571989,
-            'upload_date': '20150129',
+            'uploader': 'Mindself',
+            'uploader_id': 'mindself',
+            'timestamp': 1617109572,
+            'upload_date': '20210330',
             'age_limit': 18,
-        }
+        },
+        'params': {
+            # adaptive download
+            'skip_download': True,
+        },
     }, {
         # 480p only
         'url': 'http://spankbang.com/1vt0/video/solvane+gangbang',
@@ -134,15 +145,15 @@ class SpankBangIE(InfoExtractor):
         info = self._search_json_ld(webpage, video_id, default={})
 
         title = self._html_search_regex(
-            r'(?s)<h1[^>]*>(.+?)</h1>', webpage, 'title', default=None)
+            r'(?s)<h1[^>]+\btitle=["\']([^"]+)["\']>', webpage, 'title', default=None)
         description = self._search_regex(
             r'<div[^>]+\bclass=["\']bottom[^>]+>\s*<p>[^<]*</p>\s*<p>([^<]+)',
             webpage, 'description', default=None)
         thumbnail = self._og_search_thumbnail(webpage, default=None)
         uploader = self._html_search_regex(
-            (r'(?s)<li[^>]+class=["\']profile[^>]+>(.+?)</a>',
-             r'class="user"[^>]*><img[^>]+>([^<]+)'),
-            webpage, 'uploader', default=None)
+            r'<svg[^>]+\bclass="(?:[^"]*?user[^"]*?)">.*?</svg>([^<]+)', webpage, 'uploader', default=None)
+        uploader_id = self._html_search_regex(
+            r'<a[^>]+href="/profile/([^"]+)"', webpage, 'uploader_id', default=None)
         duration = parse_duration(self._search_regex(
             r'<div[^>]+\bclass=["\']right_side[^>]+>\s*<span>([^<]+)',
             webpage, 'duration', default=None))
@@ -157,6 +168,7 @@ class SpankBangIE(InfoExtractor):
             'description': description,
             'thumbnail': thumbnail,
             'uploader': uploader,
+            'uploader_id': uploader_id,
             'duration': duration,
             'view_count': view_count,
             'formats': formats,
@@ -167,32 +179,166 @@ class SpankBangIE(InfoExtractor):
 
 class SpankBangPlaylistIE(InfoExtractor):
     _VALID_URL = r'https?://(?:[^/]+\.)?spankbang\.com/(?P<id>[\da-z]+)/playlist/(?P<display_id>[^/]+)'
-    _TEST = {
+    _TESTS = [{
         'url': 'https://spankbang.com/ug0k/playlist/big+ass+titties',
         'info_dict': {
             'id': 'ug0k',
             'title': 'Big Ass Titties',
+            'description': 'md5:65b01bb13a9276cf172a67a41304bafd',
         },
-        'playlist_mincount': 40,
-    }
+        'playlist_mincount': 35,
+    }, {
+        # pagination required
+        'url': 'https://spankbang.com/51wxk/playlist/dance',
+        'info_dict': {
+            'id': '51wxk',
+            'title': 'Dance',
+            'description': 'md5:7aae6991c65d561a9319ecab31f857e2',
+        },
+        'playlist_mincount': 60,
+    }]
+
+    def _entries(self, url, playlist_id, webpage=None):
+        for ii in itertools.count(1):
+            if not webpage:
+                webpage = self._download_webpage(
+                    url, playlist_id,
+                    note='Downloading playlist page %d' % (ii, ),
+                    fatal=False)
+            if not webpage:
+                break
+            # search <main id="container">...</main>.innerHTML
+            for mobj in re.finditer(
+                    r'''<a\b[^>]*?\bclass\s*=\s*('|")(?:(?:(?!\1).)+?\s)?\s*thumb\b[^>]*>''',
+                    get_element_by_id('container', webpage) or webpage):
+                item_url = extract_attributes(mobj.group(0)).get('href')
+                if item_url:
+                    yield urljoin(url, item_url)
+            next_url = self._search_regex(
+                r'''\bhref\s*=\s*(["'])(?P<path>(?!\1).+?)/?\1''',
+                get_element_by_class('next', webpage) or '',
+                'continuation page', group='path', default=None)
+            if next_url is None or next_url in url:
+                break
+            url, webpage = urljoin(url, next_url), None
+            p_url = compat_urlparse.urlparse(url)
+            url = compat_urlparse.urlunparse(p_url._replace(path=p_url.path + '/'))
+
+    def _get_title(self, list_id, webpage, url):
+        return self._html_search_regex(
+            r'<h1>([^<]+)\s+playlist\s*<', webpage, 'playlist title',
+            fatal=False) or re.sub(r'(\w)\+(\w)', r'\1 \2', list_id).title()
 
     def _real_extract(self, url):
-        mobj = re.match(self._VALID_URL, url)
-        playlist_id = mobj.group('id')
-        display_id = mobj.group('display_id')
+        mobj = re.match(self._VALID_URL, url).groupdict()
+        playlist_id = mobj['id']
+        display_id = mobj.get('display_id') or playlist_id
+        related = mobj.get('related')
 
-        webpage = self._download_webpage(
-            url, playlist_id, headers={'Cookie': 'country=US; mobile=on'})
+        webpage = self._download_webpage(url, playlist_id)
 
-        entries = [self.url_result(
-            urljoin(url, mobj.group('path')),
-            ie=SpankBangIE.ie_key(), video_id=mobj.group('id'))
-            for mobj in re.finditer(
-                r'<a[^>]+\bhref=(["\'])(?P<path>/?[\da-z]+-(?P<id>[\da-z]+)/playlist/%s(?:(?!\1).)*)\1'
-                % re.escape(display_id), webpage)]
+        def get_title():
+            t = self._get_title(display_id, webpage, url)
+            if related:
+                t = '%s of %s' % (related.title(), t, )
+            return t
 
-        title = self._html_search_regex(
-            r'<h1>([^<]+)\s+playlist\s*<', webpage, 'playlist title',
-            fatal=False)
+        result = self.playlist_from_matches(self._entries(url, playlist_id, webpage), playlist_id, get_title(), ie=SpankBangIE.ie_key())
+        description = self._html_search_meta(('description', 'og:description'), webpage)
+        if description:
+            result['description'] = description
+        return result
 
-        return self.playlist_result(entries, playlist_id, title)
+
+class SpankBangListIE(SpankBangPlaylistIE):
+    _VALID_URL = r'''(?x)
+            https?://(?:[^/]+\.)?spankbang\.com/
+                (?:
+                    category|tag|s|
+                    [^/]+/channel|
+                    [^/]+/pornstar|
+                    (?P<profile>profile)
+                )/(?P<id>[^/?#]+)(?(profile)/(?P<related>videos|likes|playlists)?|)'''
+    _TESTS = [{
+        'url': 'https://spankbang.com/category/striptease/?p=d',
+        'info_dict': {
+            'id': 'striptease',
+            'title': 'Category: Striptease Porn (Jun 2022) - Joi & Strip (today)',
+            'description': 'md5:d4580b5fc1e8fae9f627178964989959',
+        },
+        'playlist_mincount': 1,
+    }, {
+        'url': 'https://spankbang.com/tag/monty/',
+        'info_dict': {
+            'id': 'monty',
+            'title': 'Tag: Monty Porn - Jenni Lee & British',
+            'description': 'md5:93535d53455f10b934034bfb9562de35',
+        },
+        'playlist_mincount': 40,
+    }, {
+        'url': 'https://spankbang.com/s/silly/?q=fhd&d=10',
+        'info_dict': {
+            'id': 'silly',
+            'title': 'Search: Silly HD Porn - Chubby Teens & Cum Swallow (FHD, 10+ mins)',
+            'description': 'md5:a241b24847f1efd7cc3cfc4ef2b22429',
+        },
+        'playlist_mincount': 20,
+    }, {
+        'url': 'https://spankbang.com/er/channel/purexxxfilms/?o=new',
+        'info_dict': {
+            'id': 'purexxxfilms',
+            'title': 'PureXXXFilms Channel (new)',
+            'description': 'md5:bb61fbf523511f3bd15aea0360bdbdc0',
+        },
+        'playlist_mincount': 65,
+    }, {
+        'url': 'https://spankbang.com/5pj/pornstar/promise/',
+        'info_dict': {
+            'id': 'promise',
+            'title': 'Promise Pornstar Page',
+            'description': 'md5:ff08a6ac2d6cd1225f0fae74ac896d63',
+        },
+        'playlist_mincount': 90,
+    }, {
+        'url': 'https://spankbang.com/profile/migouche/likes',
+        'info_dict': {
+            'id': 'migouche',
+            'title': 'Likes of migouche Profile',
+            'description': 'md5:b97540209879a62cadf2af6b0934bbc9',
+        },
+        'playlist_mincount': 2,
+    },
+    ]
+
+    def _get_title(self, list_id, webpage, url):
+        title = (
+            self._og_search_title(webpage, default=None)
+            or self._html_search_regex(
+                r'<title\b[^>]*>([^<]+)</title', webpage, 'channel title',
+                fatal=False))
+        if title:
+            title = re.sub(r'\s*(?:&\s+(?:Porn\s+))?(?:Videos\s+)?[@-]\s*SpankBang\s*$', '', title)
+        title = title or re.sub(r'(\w)\+(\w)', r'\1 \2', list_id).title()
+        qs = compat_urlparse.parse_qs(compat_urlparse.urlparse(url).query)
+        quals = []
+        for q in ('o', 'q', 'd', 'p'):
+            v = qs.get(q, [None])[-1]
+            if not v:
+                continue
+            if q == 'q':
+                v = v.upper()
+            elif q == 'd':
+                v += '+ mins'
+            elif q == 'p':
+                v = {'d': 'today', 'w': 'this week', 'm': 'this month', 'y': 'this year'}.get(v, v)
+            quals.append(v)
+        quals = ', '.join(quals)
+        if quals:
+            title += ' (%s)' % (quals, )
+        m = re.search(r'/(category|tag|s)/', url)
+        if m:
+            pfx = m.group(1)
+            if pfx == 's':
+                pfx = 'search'
+            title = '%s: %s' % (pfx.title(), title)
+        return title

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

@@ -2406,7 +2406,7 @@ class ExtractorError(YoutubeDLError):
         """ 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: