Merge 51f781c75ba2152642ef17d140cd1dec7ea763a9 into 3eb8d22ddb8982ca4fb56bb7a8d6517538bf14c6

* thx seproDev, yt-dlp/yt-dlp/pull/12694/commits/ae69e3c
* also add documenting comment in `InfoExtractor`

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_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:

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/bokecc.py

@@ -32,7 +32,7 @@ class BokeCCBaseIE(InfoExtractor):
 
 
 class BokeCCIE(BokeCCBaseIE):
-    _IE_DESC = 'CC视频'
+    IE_DESC = 'CC视频'
     _VALID_URL = r'https?://union\.bokecc\.com/playvideo\.bo\?(?P<query>.*)'
 
     _TESTS = [{

youtube_dl/extractor/cloudy.py

@@ -9,7 +9,7 @@ from ..utils import (
 
 
 class CloudyIE(InfoExtractor):
-    _IE_DESC = 'cloudy.ec'
+    IE_DESC = 'cloudy.ec'
     _VALID_URL = r'https?://(?:www\.)?cloudy\.ec/(?:v/|embed\.php\?.*?\bid=)(?P<id>[A-Za-z0-9]+)'
     _TESTS = [{
         'url': 'https://www.cloudy.ec/v/af511e2527aac',

youtube_dl/extractor/common.py

@@ -422,6 +422,8 @@ class InfoExtractor(object):
     _GEO_COUNTRIES = None
     _GEO_IP_BLOCKS = None
     _WORKING = True
+    # supply this in public subclasses: used in supported sites list, etc
+    # IE_DESC = 'short description of IE'
 
     def __init__(self, downloader=None):
         """Constructor. Receives an optional downloader."""

youtube_dl/extractor/itv.py

@@ -35,15 +35,6 @@ from ..utils import (
 
 class ITVBaseIE(InfoExtractor):
 
-    def _search_nextjs_data(self, webpage, video_id, **kw):
-        transform_source = kw.pop('transform_source', None)
-        fatal = kw.pop('fatal', True)
-        return self._parse_json(
-            self._search_regex(
-                r'''<script\b[^>]+\bid=('|")__NEXT_DATA__\1[^>]*>(?P<js>[^<]+)</script>''',
-                webpage, 'next.js data', group='js', fatal=fatal, **kw),
-            video_id, transform_source=transform_source, fatal=fatal)
-
     def __handle_request_webpage_error(self, err, video_id=None, errnote=None, fatal=True):
         if errnote is False:
             return False
@@ -109,7 +100,9 @@ class ITVBaseIE(InfoExtractor):
 
 class ITVIE(ITVBaseIE):
     _VALID_URL = r'https?://(?:www\.)?itv\.com/(?:(?P<w>watch)|hub)/[^/]+/(?(w)[\w-]+/)(?P<id>\w+)'
-    _IE_DESC = 'ITVX'
+    IE_DESC = 'ITVX'
+    _WORKING = False
+
     _TESTS = [{
         'note': 'Hub URLs redirect to ITVX',
         'url': 'https://www.itv.com/hub/liar/2a4547a0012',
@@ -270,7 +263,7 @@ class ITVIE(ITVBaseIE):
                 'ext': determine_ext(href, 'vtt'),
             })
 
-        next_data = self._search_nextjs_data(webpage, video_id, fatal=False, default='{}')
+        next_data = self._search_nextjs_data(webpage, video_id, fatal=False, default={})
         video_data.update(traverse_obj(next_data, ('props', 'pageProps', ('title', 'episode')), expected_type=dict)[0] or {})
         title = traverse_obj(video_data, 'headerTitle', 'episodeTitle')
         info = self._og_extract(webpage, require_title=not title)
@@ -323,7 +316,7 @@ class ITVIE(ITVBaseIE):
 
 class ITVBTCCIE(ITVBaseIE):
     _VALID_URL = r'https?://(?:www\.)?itv\.com/(?!(?:watch|hub)/)(?:[^/]+/)+(?P<id>[^/?#&]+)'
-    _IE_DESC = 'ITV articles: News, British Touring Car Championship'
+    IE_DESC = 'ITV articles: News, British Touring Car Championship'
     _TESTS = [{
         'note': 'British Touring Car Championship',
         'url': 'https://www.itv.com/btcc/articles/btcc-2018-all-the-action-from-brands-hatch',

youtube_dl/extractor/senateisvp.py

@@ -47,7 +47,7 @@ class SenateISVPIE(InfoExtractor):
         ['vetaff', '76462', 'http://vetaff-f.akamaihd.net'],
         ['arch', '', 'http://ussenate-f.akamaihd.net/']
     ]
-    _IE_NAME = 'senate.gov'
+    IE_NAME = 'senate.gov'
     _VALID_URL = r'https?://(?:www\.)?senate\.gov/isvp/?\?(?P<qs>.+)'
     _TESTS = [{
         'url': 'http://www.senate.gov/isvp/?comm=judiciary&type=live&stt=&filename=judiciary031715&auto_play=false&wmode=transparent&poster=http%3A%2F%2Fwww.judiciary.senate.gov%2Fthemes%2Fjudiciary%2Fimages%2Fvideo-poster-flash-fit.png',

youtube_dl/jsinterp.py

@@ -686,6 +686,8 @@ class JSInterpreter(object):
             raise self.Exception('Cannot get index {idx!r:.100}'.format(**locals()), expr=repr(obj), cause=e)
 
     def _dump(self, obj, namespace):
+        if obj is JS_Undefined:
+            return 'undefined'
         try:
             return json.dumps(obj)
         except TypeError:

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