Merge 43fe4b597f 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/ciscolive.py

@@ -1,8 +1,6 @@
 # coding: utf-8
 from __future__ import unicode_literals
 
-import itertools
-
 from .common import InfoExtractor
 from ..compat import (
     compat_parse_qs,
@@ -10,6 +8,7 @@ from ..compat import (
 )
 from ..utils import (
     clean_html,
+    ExtractorError,
     float_or_none,
     int_or_none,
     try_get,
@@ -21,12 +20,13 @@ class CiscoLiveBaseIE(InfoExtractor):
     # These appear to be constant across all Cisco Live presentations
     # and are not tied to any user session or event
     RAINFOCUS_API_URL = 'https://events.rainfocus.com/api/%s'
-    RAINFOCUS_API_PROFILE_ID = 'Na3vqYdAlJFSxhYTYQGuMbpafMqftalz'
-    RAINFOCUS_WIDGET_ID = 'n6l4Lo05R8fiy3RpUBm447dZN8uNWoye'
+    RAINFOCUS_API_PROFILE_ID = ''  # if blank will be fetched at runtime from site javascript
+    RAINFOCUS_WIDGET_ID = ''  # if blank will be fetched at runtime from site javascript
+    RAINFOCUS_TOKENS_URL = 'https://cdn-events.rainfocus.com/pages/cisco/clondemand/catalog.js'
     BRIGHTCOVE_URL_TEMPLATE = 'http://players.brightcove.net/5647924234001/SyK2FdqjM_default/index.html?videoId=%s'
 
     HEADERS = {
-        'Origin': 'https://ciscolive.cisco.com',
+        'Origin': 'https://ciscolive.com',
         'rfApiProfileId': RAINFOCUS_API_PROFILE_ID,
         'rfWidgetId': RAINFOCUS_WIDGET_ID,
     }
@@ -34,9 +34,25 @@ class CiscoLiveBaseIE(InfoExtractor):
     def _call_api(self, ep, rf_id, query, referrer, note=None):
         headers = self.HEADERS.copy()
         headers['Referer'] = referrer
-        return self._download_json(
+        if not self.RAINFOCUS_API_PROFILE_ID or not self.RAINFOCUS_WIDGET_ID:
+            rf_token_js = self._download_webpage(self.RAINFOCUS_TOKENS_URL, 'catalog.js', headers=headers)
+            for token in ('apiToken', 'widgetId'):
+                if token not in rf_token_js:
+                    raise ExtractorError(
+                        'Unable to fetch ' + token, expected=True)
+            api_token = self._html_search_regex(
+                r'''apiToken:\s+["'](\w+)''', rf_token_js, 'apiToken')
+            widget_id = self._html_search_regex(
+                r'''widgetId:\s+["'](\w+)''', rf_token_js, 'widgetId')
+            headers['rfApiProfileId'] = api_token
+            headers['rfWidgetId'] = widget_id
+        rf_result = self._download_json(
             self.RAINFOCUS_API_URL % ep, rf_id, note=note,
             data=urlencode_postdata(query), headers=headers)
+        if int(rf_result['responseCode']) != 0:
+            raise ExtractorError(
+                'Rainfocus %s api returned a non success responseCode: %s. api keys might be invalid' % (ep, rf_result['responseCode']), expected=True)
+        return rf_result
 
     def _parse_rf_item(self, rf_item):
         event_name = rf_item.get('eventName')
@@ -67,17 +83,16 @@ class CiscoLiveBaseIE(InfoExtractor):
 class CiscoLiveSessionIE(CiscoLiveBaseIE):
     _VALID_URL = r'https?://(?:www\.)?ciscolive(?:\.cisco)?\.com/[^#]*#/session/(?P<id>[^/?&]+)'
     _TESTS = [{
-        'url': 'https://ciscolive.cisco.com/on-demand-library/?#/session/1423353499155001FoSs',
-        'md5': 'c98acf395ed9c9f766941c70f5352e22',
+        'url': 'https://www.ciscolive.com/on-demand/on-demand-library.html?search=#/session/16360600004400017rMx',
+        'md5': 'e0f5b0b2927b586ebff619294fec6926',
         'info_dict': {
-            'id': '5803694304001',
+            'id': '6128601216001',
             'ext': 'mp4',
-            'title': '13 Smart Automations to Monitor Your Cisco IOS Network',
-            'description': 'md5:ec4a436019e09a918dec17714803f7cc',
-            'timestamp': 1530305395,
-            'upload_date': '20180629',
+            'title': 'A Deeper Dive into the Telco Cloud Architecture Evolution to support 5G and MEC - BRKSPG-1565',
+            'description': 'md5:04bc54ec8ede2bbd9d21264bfaf16cb3',
+            'timestamp': 1580474594,
+            'upload_date': '20200131',
             'uploader_id': '5647924234001',
-            'location': '16B Mezz.',
         },
     }, {
         'url': 'https://www.ciscolive.com/global/on-demand-library.html?search.event=ciscoliveemea2019#/session/15361595531500013WOU',
@@ -85,6 +100,9 @@ class CiscoLiveSessionIE(CiscoLiveBaseIE):
     }, {
         'url': 'https://www.ciscolive.com/global/on-demand-library.html?#/session/1490051371645001kNaS',
         'only_matching': True,
+    }, {
+        'url': 'https://www.ciscolive.com/on-demand/on-demand-library.html?#/session/16360600774720017Iby',
+        'only_matching': True,
     }]
 
     def _real_extract(self, url):
@@ -94,16 +112,15 @@ class CiscoLiveSessionIE(CiscoLiveBaseIE):
 
 
 class CiscoLiveSearchIE(CiscoLiveBaseIE):
-    _VALID_URL = r'https?://(?:www\.)?ciscolive(?:\.cisco)?\.com/(?:global/)?on-demand-library(?:\.html|/)'
+    _VALID_URL = r'https?://(?:www\.)?ciscolive(?:\.cisco)?\.com/(?:global/|on-demand/)?on-demand-library(?:\.html|/)'
     _TESTS = [{
-        'url': 'https://ciscolive.cisco.com/on-demand-library/?search.event=ciscoliveus2018&search.technicallevel=scpsSkillLevel_aintroductory&search.focus=scpsSessionFocus_designAndDeployment#/',
+        'url': 'https://www.ciscolive.com/on-demand/on-demand-library.html?search.technology=1614262524988009b6j9&search.technicallevel=scpsSkillLevel_bintermediate#/',
         'info_dict': {
             'title': 'Search query',
         },
-        'playlist_count': 5,
+        'playlist_count': 6,
     }, {
-        'url': 'https://ciscolive.cisco.com/on-demand-library/?search.technology=scpsTechnology_applicationDevelopment&search.technology=scpsTechnology_ipv6&search.focus=scpsSessionFocus_troubleshootingTroubleshooting#/',
-        'only_matching': True,
+        'url': 'https://www.ciscolive.com/on-demand/on-demand-library.html?search.technology=scpsTechnology_automation&search.technicallevel=scpsSkillLevel_cadvanced#/',
     }, {
         'url': 'https://www.ciscolive.com/global/on-demand-library.html?search.technicallevel=scpsSkillLevel_aintroductory&search.event=ciscoliveemea2019&search.technology=scpsTechnology_dataCenter&search.focus=scpsSessionFocus_bestPractices#/',
         'only_matching': True,
@@ -118,31 +135,25 @@ class CiscoLiveSearchIE(CiscoLiveBaseIE):
         return int_or_none(try_get(rf_item, lambda x: x['videos'][0]['url'])) is not None
 
     def _entries(self, query, url):
-        query['size'] = 50
-        query['from'] = 0
-        for page_num in itertools.count(1):
-            results = self._call_api(
-                'search', None, query, url,
-                'Downloading search JSON page %d' % page_num)
-            sl = try_get(results, lambda x: x['sectionList'][0], dict)
-            if sl:
-                results = sl
-            items = results.get('items')
-            if not items or not isinstance(items, list):
-                break
-            for item in items:
-                if not isinstance(item, dict):
-                    continue
-                if not self._check_bc_id_exists(item):
-                    continue
-                yield self._parse_rf_item(item)
-            size = int_or_none(results.get('size'))
-            if size is not None:
-                query['size'] = size
-            total = int_or_none(results.get('total'))
-            if total is not None and query['from'] + query['size'] > total:
-                break
-            query['from'] += query['size']
+        results = self._call_api(
+            'search', None, query, url,
+            'Downloading search JSON')
+        if int(results['totalSearchItems']) == 0:
+            raise ExtractorError(
+                'Search api returned no items (if matches are expected rfApiProfileId may be invalid)',
+                expected=True)
+        sl = try_get(results, lambda x: x['sectionList'], list)
+        if sl is not None:
+            for s in sl:
+                items = s.get('items')
+                if not items or not isinstance(items, list):
+                    break
+                for item in items:
+                    if not isinstance(item, dict):
+                        continue
+                    if not self._check_bc_id_exists(item):
+                        continue
+                    yield self._parse_rf_item(item)
 
     def _real_extract(self, url):
         query = compat_parse_qs(compat_urllib_parse_urlparse(url).query)

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

@@ -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: