Merge 51f781c75b into c5098961b0

Now also succeeds with player b12cc44b

test/test_YoutubeDL.py

@@ -652,6 +652,10 @@ class TestYoutubeDL(unittest.TestCase):
             'height': 1080,
             'title1': '$PATH',
             'title2': '%PATH%',
+            'track': '-track with initial hyphen',
+            'display_id': '-initial_hyphen',
+            'episode': '.episode with initial period',
+            'episode_id': '.initial_period',
         }
 
         def fname(templ, na_placeholder='NA'):
@@ -667,7 +671,9 @@ class TestYoutubeDL(unittest.TestCase):
         self.assertEqual(fname(NA_TEST_OUTTMPL), 'NA-NA-1234.mp4')
         # Or by provided placeholder
         self.assertEqual(fname(NA_TEST_OUTTMPL, na_placeholder='none'), 'none-none-1234.mp4')
-        self.assertEqual(fname(NA_TEST_OUTTMPL, na_placeholder=''), '--1234.mp4')
+        self.assertEqual(fname(NA_TEST_OUTTMPL, na_placeholder=''), '_-1234.mp4')
+        NA_TEST_OUTTMPL = '%(uploader_date)s+%(width)d+%(id)s.%(ext)s'
+        self.assertEqual(fname(NA_TEST_OUTTMPL, na_placeholder=''), '++1234.mp4')
         self.assertEqual(fname('%(height)d.%(ext)s'), '1080.mp4')
         self.assertEqual(fname('%(height)6d.%(ext)s'), '  1080.mp4')
         self.assertEqual(fname('%(height)-6d.%(ext)s'), '1080  .mp4')
@@ -685,6 +691,16 @@ class TestYoutubeDL(unittest.TestCase):
         self.assertEqual(fname('%%(width)06d.%(ext)s'), '%(width)06d.mp4')
         self.assertEqual(fname('Hello %(title1)s'), 'Hello $PATH')
         self.assertEqual(fname('Hello %(title2)s'), 'Hello %PATH%')
+        self.assertEqual(fname('%(track)s at start changes hyphen'),
+                         '_track with initial hyphen at start changes hyphen')
+        self.assertEqual(fname('medial %(track)s doesn\'t change hyphen'),
+                         'medial -track with initial hyphen doesn\'t change hyphen')
+        self.assertEqual(fname('%(display_id)s at start doesn\'t change hyphen'),
+                         '-initial_hyphen at start doesn\'t change hyphen')
+        self.assertEqual(fname('%(episode)s at start changes period'),
+                         '_episode with initial period at start changes period')
+        self.assertEqual(fname('%(episode_id)s at start doesn\'t change period'),
+                         '.initial_period at start doesn\'t change period')
 
     def test_format_note(self):
         ydl = YoutubeDL()

test/test_jsinterp.py

@@ -425,6 +425,34 @@ class TestJSInterpreter(unittest.TestCase):
             self._test(jsi, [''], args=['', '-'])
             self._test(jsi, [], args=['', ''])
 
+    def test_slice(self):
+        self._test('function f(){return [0, 1, 2, 3, 4, 5, 6, 7, 8].slice()}', [0, 1, 2, 3, 4, 5, 6, 7, 8])
+        self._test('function f(){return [0, 1, 2, 3, 4, 5, 6, 7, 8].slice(0)}', [0, 1, 2, 3, 4, 5, 6, 7, 8])
+        self._test('function f(){return [0, 1, 2, 3, 4, 5, 6, 7, 8].slice(5)}', [5, 6, 7, 8])
+        self._test('function f(){return [0, 1, 2, 3, 4, 5, 6, 7, 8].slice(99)}', [])
+        self._test('function f(){return [0, 1, 2, 3, 4, 5, 6, 7, 8].slice(-2)}', [7, 8])
+        self._test('function f(){return [0, 1, 2, 3, 4, 5, 6, 7, 8].slice(-99)}', [0, 1, 2, 3, 4, 5, 6, 7, 8])
+        self._test('function f(){return [0, 1, 2, 3, 4, 5, 6, 7, 8].slice(0, 0)}', [])
+        self._test('function f(){return [0, 1, 2, 3, 4, 5, 6, 7, 8].slice(1, 0)}', [])
+        self._test('function f(){return [0, 1, 2, 3, 4, 5, 6, 7, 8].slice(0, 1)}', [0])
+        self._test('function f(){return [0, 1, 2, 3, 4, 5, 6, 7, 8].slice(3, 6)}', [3, 4, 5])
+        self._test('function f(){return [0, 1, 2, 3, 4, 5, 6, 7, 8].slice(1, -1)}', [1, 2, 3, 4, 5, 6, 7])
+        self._test('function f(){return [0, 1, 2, 3, 4, 5, 6, 7, 8].slice(-1, 1)}', [])
+        self._test('function f(){return [0, 1, 2, 3, 4, 5, 6, 7, 8].slice(-3, -1)}', [6, 7])
+        self._test('function f(){return "012345678".slice()}', '012345678')
+        self._test('function f(){return "012345678".slice(0)}', '012345678')
+        self._test('function f(){return "012345678".slice(5)}', '5678')
+        self._test('function f(){return "012345678".slice(99)}', '')
+        self._test('function f(){return "012345678".slice(-2)}', '78')
+        self._test('function f(){return "012345678".slice(-99)}', '012345678')
+        self._test('function f(){return "012345678".slice(0, 0)}', '')
+        self._test('function f(){return "012345678".slice(1, 0)}', '')
+        self._test('function f(){return "012345678".slice(0, 1)}', '0')
+        self._test('function f(){return "012345678".slice(3, 6)}', '345')
+        self._test('function f(){return "012345678".slice(1, -1)}', '1234567')
+        self._test('function f(){return "012345678".slice(-1, 1)}', '')
+        self._test('function f(){return "012345678".slice(-3, -1)}', '67')
+
 
 if __name__ == '__main__':
     unittest.main()

test/test_utils.py

@@ -156,10 +156,8 @@ class TestUtil(unittest.TestCase):
             sanitize_filename('New World record at 0:12:34'),
             'New World record at 0_12_34')
 
-        self.assertEqual(sanitize_filename('--gasdgf'), '_-gasdgf')
-        self.assertEqual(sanitize_filename('--gasdgf', is_id=True), '--gasdgf')
-        self.assertEqual(sanitize_filename('.gasdgf'), 'gasdgf')
-        self.assertEqual(sanitize_filename('.gasdgf', is_id=True), '.gasdgf')
+        self.assertEqual(sanitize_filename('__gas__dgf_'), 'gas_dgf')
+        self.assertEqual(sanitize_filename('__gas__dgf_', is_id=True), '__gas__dgf_')
 
         forbidden = '"\0\\/'
         for fc in forbidden:

test/test_youtube_signature.py

@@ -174,6 +174,14 @@ _NSIG_TESTS = [
         'https://www.youtube.com/s/player/5604538d/player_ias.vflset/en_US/base.js',
         '7X-he4jjvMx7BCX', 'sViSydX8IHtdWA',
     ),
+    (
+        'https://www.youtube.com/s/player/20dfca59/player_ias.vflset/en_US/base.js',
+        '-fLCxedkAk4LUTK2', 'O8kfRq1y1eyHGw',
+    ),
+    (
+        'https://www.youtube.com/s/player/b12cc44b/player_ias.vflset/en_US/base.js',
+        'keLa5R2U00sR9SQK', 'N1OGyujjEwMnLw',
+    ),
 ]
 
 

youtube_dl/YoutubeDL.py

@@ -694,6 +694,20 @@ class YoutubeDL(object):
         except UnicodeEncodeError:
             self.to_screen('[download] The file has already been downloaded')
 
+    # As of [1] format syntax is:
+    #  %[mapping_key][conversion_flags][minimum_width][.precision][length_modifier]type
+    # 1. https://docs.python.org/2/library/stdtypes.html#string-formatting
+    _FORMAT_RE = r'''(?x)
+        (?<!%)
+        %
+        \({0}\)  # mapping key
+        (?:[#0\-+ ]+)?  # conversion flags (optional)
+        (?:\d+)?  # minimum field width (optional)
+        (?:\.\d+)?  # precision (optional)
+        [hlL]?  # length modifier (optional)
+        [diouxXeEfFgGcrs%]  # conversion type
+    '''
+
     def prepare_filename(self, info_dict):
         """Generate the output filename."""
         try:
@@ -712,10 +726,11 @@ class YoutubeDL(object):
                 elif template_dict.get('width'):
                     template_dict['resolution'] = '%dx?' % template_dict['width']
 
+            is_id = lambda k: k == 'id' or k.endswith('_id')
             sanitize = lambda k, v: sanitize_filename(
                 compat_str(v),
                 restricted=self.params.get('restrictfilenames'),
-                is_id=(k == 'id' or k.endswith('_id')))
+                is_id=is_id(k))
             template_dict = dict((k, v if isinstance(v, compat_numeric_types) else sanitize(k, v))
                                  for k, v in template_dict.items()
                                  if v is not None and not isinstance(v, (list, tuple, dict)))
@@ -743,21 +758,8 @@ class YoutubeDL(object):
             # output template for missing fields to meet string presentation type.
             for numeric_field in self._NUMERIC_FIELDS:
                 if numeric_field not in template_dict:
-                    # As of [1] format syntax is:
-                    #  %[mapping_key][conversion_flags][minimum_width][.precision][length_modifier]type
-                    # 1. https://docs.python.org/2/library/stdtypes.html#string-formatting
-                    FORMAT_RE = r'''(?x)
-                        (?<!%)
-                        %
-                        \({0}\)  # mapping key
-                        (?:[#0\-+ ]+)?  # conversion flags (optional)
-                        (?:\d+)?  # minimum field width (optional)
-                        (?:\.\d+)?  # precision (optional)
-                        [hlL]?  # length modifier (optional)
-                        [diouxXeEfFgGcrs%]  # conversion type
-                    '''
                     outtmpl = re.sub(
-                        FORMAT_RE.format(numeric_field),
+                        self._FORMAT_RE.format(numeric_field),
                         r'%({0})s'.format(numeric_field), outtmpl)
 
             # expand_path translates '%%' into '%' and '$$' into '$'
@@ -771,7 +773,35 @@ class YoutubeDL(object):
             # because meta fields may contain env variables we don't want to
             # be expanded. For example, for outtmpl "%(title)s.%(ext)s" and
             # title "Hello $PATH", we don't want `$PATH` to be expanded.
-            filename = expand_path(outtmpl).replace(sep, '') % template_dict
+            filename = expand_path(outtmpl).replace(sep, '')
+
+            # avoid non-mandatory (#571) initial '-' (#5035): convert to '_'
+            # also remove initial '.'
+
+            # first split the filename around the format fields
+            FORMAT_SPLIT_RE = (
+                '%s(%s)' % tuple(
+                    re.split(r'(\(\?[iLmsux]+\))',
+                             self._FORMAT_RE.format(r'\w+'), maxsplit=1)[1:]))
+            filename = re.split(FORMAT_SPLIT_RE, filename)
+            # tag any filename fragment that is an ID format
+            FORMAT_FIELD_RE = (
+                '%s(%s)' % tuple(
+                    re.split(r'(\(\?[iLmsux]+\))',
+                             self._FORMAT_RE.format(r'(?P<field>\w+)'), maxsplit=1)[1:]))
+            field_is_id = (
+                lambda x:
+                    (lambda y: y and is_id(y.group('field')))(re.match(FORMAT_FIELD_RE, x)))
+            # list formatted fragments and tags
+            tagged_fn = [(x % template_dict, field_is_id(x)) for x in filename]
+            # list the initial fragments that can be munged from [.-] to _, ie
+            # those that not IDs, or empty
+            mungable = itertools.takewhile(lambda x: not(x[1] and x[0]), tagged_fn)
+            col0 = lambda l: next(iter(zip(*l)), [])
+            mungable = col0(mungable)
+            # finally combine the munged initial part and the rest
+            filename = (re.sub(r'^[.-]', '_', ''.join(mungable))
+                        + ''.join(col0(tagged_fn[len(mungable):])))
 
             # Temporary fix for #4787
             # 'Treat' all problem characters by passing filename through preferredencoding

youtube_dl/extractor/youtube.py

@@ -1659,17 +1659,46 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
     def _extract_n_function_name(self, jscode):
         func_name, idx = self._search_regex(
             # new: (b=String.fromCharCode(110),c=a.get(b))&&c=nfunc[idx](c)
-            # or:  (b="nn"[+a.D],c=a.get(b))&&(c=nfunc[idx](c)s
-            # old: .get("n"))&&(b=nfunc[idx](b)
-            # older: .get("n"))&&(b=nfunc(b)
+            # or:  (b="nn"[+a.D],c=a.get(b))&&(c=nfunc[idx](c)
+            # or:  (PL(a),b=a.j.n||null)&&(b=nfunc[idx](b)
+            # or:  (b="nn"[+a.D],vL(a),c=a.j[b]||null)&&(c=narray[idx](c),a.set(b,c),narray.length||nfunc("")
+            # old: (b=a.get("n"))&&(b=nfunc[idx](b)(?P<c>[a-z])\s*=\s*[a-z]\s*
+            # older: (b=a.get("n"))&&(b=nfunc(b)
             r'''(?x)
-                (?:\(\s*(?P<b>[a-z])\s*=\s*(?:
-                    String\s*\.\s*fromCharCode\s*\(\s*110\s*\)|
-                    "n+"\[\s*\+?s*[\w$.]+\s*]
-                )\s*,(?P<c>[a-z])\s*=\s*[a-z]\s*)?
-                \.\s*get\s*\(\s*(?(b)(?P=b)|"n{1,2}")(?:\s*\)){2}\s*&&\s*\(\s*(?(c)(?P=c)|b)\s*=\s*
-                (?P<nfunc>[a-zA-Z_$][\w$]*)(?:\s*\[(?P<idx>\d+)\])?\s*\(\s*[\w$]+\s*\)
-            ''', jscode, 'Initial JS player n function name', group=('nfunc', 'idx'))
+                \((?:[\w$()\s]+,)*?\s*      # (
+                (?P<b>[a-z])\s*=\s*         # b=
+                (?:
+                    (?:                     # expect ,c=a.get(b) (etc)
+                        String\s*\.\s*fromCharCode\s*\(\s*110\s*\)|
+                        "n+"\[\s*\+?s*[\w$.]+\s*]
+                    )\s*(?:,[\w$()\s]+(?=,))*|
+                       (?P<old>[\w$]+)      # a (old[er])
+                   )\s*
+                   (?(old)
+                                            # b.get("n")
+                       (?:\.\s*[\w$]+\s*|\[\s*[\w$]+\s*]\s*)*?
+                       (?:\.\s*n|\[\s*"n"\s*]|\.\s*get\s*\(\s*"n"\s*\))
+                       |                    # ,c=a.get(b)
+                       ,\s*(?P<c>[a-z])\s*=\s*[a-z]\s*
+                       (?:\.\s*[\w$]+\s*|\[\s*[\w$]+\s*]\s*)*?
+                       (?:\[\s*(?P=b)\s*]|\.\s*get\s*\(\s*(?P=b)\s*\))
+                   )
+                                            # interstitial junk
+                   \s*(?:\|\|\s*null\s*)?(?:\)\s*)?&&\s*(?:\(\s*)?
+               (?(c)(?P=c)|(?P=b))\s*=\s*   # [c|b]=
+                                            # nfunc|nfunc[idx]
+                   (?P<nfunc>[a-zA-Z_$][\w$]*)(?:\s*\[(?P<idx>\d+)\])?\s*\(\s*[\w$]+\s*\)
+            ''', jscode, 'Initial JS player n function name', group=('nfunc', 'idx'),
+            default=(None, None))
+        # thx bashonly: yt-dlp/yt-dlp/pull/10611
+        if not func_name:
+            self.report_warning('Falling back to generic n function search')
+            return self._search_regex(
+                r'''(?xs)
+                    (?:(?<=[^\w$])|^)       # instead of \b, which ignores $
+                    (?P<name>(?!\d)[a-zA-Z\d_$]+)\s*=\s*function\((?!\d)[a-zA-Z\d_$]+\)
+                    \s*\{(?:(?!};).)+?["']enhanced_except_
+                ''', jscode, 'Initial JS player n function name', group='name')
         if not idx:
             return func_name
 

youtube_dl/jsinterp.py

@@ -925,9 +925,16 @@ class JSInterpreter(object):
                     obj.reverse()
                     return obj
                 elif member == 'slice':
-                    assertion(isinstance(obj, list), 'must be applied on a list')
-                    assertion(len(argvals) == 1, 'takes exactly one argument')
-                    return obj[argvals[0]:]
+                    assertion(isinstance(obj, (list, compat_str)), 'must be applied on a list or string')
+                    # From [1]:
+                    # .slice() - like [:]
+                    # .slice(n) - like [n:] (not [slice(n)]
+                    # .slice(m, n) - like [m:n] or [slice(m, n)]
+                    # [1] https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Global_Objects/Array/slice
+                    assertion(len(argvals) <= 2, 'takes between 0 and 2 arguments')
+                    if len(argvals) < 2:
+                        argvals += (None,)
+                    return obj[slice(*argvals)]
                 elif member == 'splice':
                     assertion(isinstance(obj, list), 'must be applied on a list')
                     assertion(argvals, 'takes one or more arguments')

youtube_dl/utils.py

@@ -2113,15 +2113,11 @@ def sanitize_filename(s, restricted=False, is_id=False):
     s = re.sub(r'[0-9]+(?::[0-9]+)+', lambda m: m.group(0).replace(':', '_'), s)
     result = ''.join(map(replace_insane, s))
     if not is_id:
-        while '__' in result:
-            result = result.replace('__', '_')
+        result = re.sub(r'_{2,}', '_', result)
         result = result.strip('_')
         # Common case of "Foreign band name - English song title"
         if restricted and result.startswith('-_'):
             result = result[2:]
-        if result.startswith('-'):
-            result = '_' + result[len('-'):]
-        result = result.lstrip('.')
         if not result:
             result = '_'
     return result