[cleanup] Remove obsolete Python compatibility code (#17357)

Authored by: doe1080

devscripts/prepare_manpage.py

@@ -66,8 +66,7 @@ def convert_code_blocks(readme):
 
 def move_sections(readme):
     MOVE_TAG_TEMPLATE = '<!-- MANPAGE: MOVE "%s" SECTION HERE -->'
-    sections = re.findall(r'(?m)^%s$' % (
+    sections = re.findall(r'(?m)^%s$' % (re.escape(MOVE_TAG_TEMPLATE) % '(.+)'), readme)
-        re.escape(MOVE_TAG_TEMPLATE).replace(r'\%', '%') % '(.+)'), readme)
 
     for section_name in sections:
         move_tag = MOVE_TAG_TEMPLATE % section_name

test/helper.py

@@ -3,7 +3,6 @@ import hashlib
 import json
 import os.path
 import re
-import ssl
 import sys
 import types
 
@@ -318,36 +317,6 @@ def expect_info_dict(self, got_dict, expected_dict):
             'Missing keys in test definition: {}'.format(', '.join(sorted(missing_keys))))
 
 
-def assertRegexpMatches(self, text, regexp, msg=None):
-    if hasattr(self, 'assertRegexp'):
-        return self.assertRegexp(text, regexp, msg)
-    else:
-        m = re.match(regexp, text)
-        if not m:
-            note = f'Regexp didn\'t match: {regexp!r} not found'
-            if len(text) < 1000:
-                note += f' in {text!r}'
-            if msg is None:
-                msg = note
-            else:
-                msg = note + ', ' + msg
-            self.assertTrue(m, msg)
-
-
-def assertGreaterEqual(self, got, expected, msg=None):
-    if not (got >= expected):
-        if msg is None:
-            msg = f'{got!r} not greater than or equal to {expected!r}'
-        self.assertTrue(got >= expected, msg)
-
-
-def assertLessEqual(self, got, expected, msg=None):
-    if not (got <= expected):
-        if msg is None:
-            msg = f'{got!r} not less than or equal to {expected!r}'
-        self.assertTrue(got <= expected, msg)
-
-
 def assertEqual(self, got, expected, msg=None):
     if got != expected:
         if msg is None:
@@ -366,12 +335,7 @@ def expect_warnings(ydl, warnings_re):
 
 
 def http_server_port(httpd):
-    if os.name == 'java' and isinstance(httpd.socket, ssl.SSLSocket):
+    return httpd.server_address[1]
-        # In Jython SSLSocket is not a subclass of socket.socket
-        sock = httpd.socket.sock
-    else:
-        sock = httpd.socket
-    return sock.getsockname()[1]
 
 
 def verify_address_availability(address):

test/test_YoutubeDL.py

@@ -15,7 +15,7 @@ import contextlib
 import copy
 import json
 
-from test.helper import FakeYDL, assertRegexpMatches, try_rm
+from test.helper import FakeYDL, try_rm
 from yt_dlp import YoutubeDL
 from yt_dlp.extractor.common import InfoExtractor
 from yt_dlp.postprocessor.common import PostProcessor
@@ -860,10 +860,10 @@ class TestYoutubeDL(unittest.TestCase):
     def test_format_note(self):
         ydl = YoutubeDL()
         self.assertEqual(ydl._format_note({}), '')
-        assertRegexpMatches(self, ydl._format_note({
+        self.assertRegex(ydl._format_note({
             'vbr': 10,
         }), r'^\s*10k$')
-        assertRegexpMatches(self, ydl._format_note({
+        self.assertRegex(ydl._format_note({
             'fps': 30,
         }), r'^30fps$')
 

test/test_download.py

@@ -13,8 +13,6 @@ import hashlib
 import json
 
 from test.helper import (
-    assertGreaterEqual,
-    assertLessEqual,
     expect_info_dict,
     expect_warnings,
     get_params,
@@ -201,8 +199,8 @@ def generator(test_case, tname):
             num_entries = len(res_dict.get('entries', []))
             if 'playlist_mincount' in test_case:
                 mincount = test_case['playlist_mincount']
-                assertGreaterEqual(
+                self.assertGreaterEqual(
-                    self, num_entries, mincount,
+                    num_entries, mincount,
                     f'Expected at least {mincount} entries in playlist {test_url}, but got only {num_entries}')
             if 'playlist_count' in test_case:
                 count = test_case['playlist_count']
@@ -212,8 +210,8 @@ def generator(test_case, tname):
                     f'Expected exactly {count} entries in playlist {test_url}, but got {got}')
             if 'playlist_maxcount' in test_case:
                 maxcount = test_case['playlist_maxcount']
-                assertLessEqual(
+                self.assertLessEqual(
-                    self, num_entries, maxcount,
+                    num_entries, maxcount,
                     f'Expected at most {maxcount} entries in playlist {test_url}, but got more')
             if 'playlist_duration_sum' in test_case:
                 got_duration = sum(e['duration'] for e in res_dict['entries'])
@@ -241,8 +239,8 @@ def generator(test_case, tname):
                         if params.get('test'):
                             expected_minsize = max(expected_minsize, 10000)
                         got_fsize = os.path.getsize(tc_filename)
-                        assertGreaterEqual(
+                        self.assertGreaterEqual(
-                            self, got_fsize, expected_minsize,
+                            got_fsize, expected_minsize,
                             f'Expected {tc_filename} to be at least {format_bytes(expected_minsize)}, '
                             f'but it\'s only {format_bytes(got_fsize)} ')
                     if 'md5' in tc:

test/test_networking.py

@@ -984,28 +984,15 @@ class TestUrllibRequestHandler(TestRequestHandlerBase):
             ):
                 validate_and_send(rh, Request(f'https://127.0.0.1:{self.https_port}/headers'))
 
-    @pytest.mark.parametrize('req,match,version_check', [
+    @pytest.mark.parametrize('req,match', [
         # https://github.com/python/cpython/blob/987b712b4aeeece336eed24fcc87a950a756c3e2/Lib/http/client.py#L1256
-        # bpo-39603: Check implemented in 3.7.9+, 3.8.5+
+        (Request('http://127.0.0.1', method='GET\n'), 'method can\'t contain control characters'),
-        (
-            Request('http://127.0.0.1', method='GET\n'),
-            'method can\'t contain control characters',
-            lambda v: v < (3, 7, 9) or (3, 8, 0) <= v < (3, 8, 5),
-        ),
         # https://github.com/python/cpython/blob/987b712b4aeeece336eed24fcc87a950a756c3e2/Lib/http/client.py#L1265
-        # bpo-38576: Check implemented in 3.7.8+, 3.8.3+
+        (Request('http://127.0.0. 1', method='GET'), 'URL can\'t contain control characters'),
-        (
-            Request('http://127.0.0. 1', method='GET'),
-            'URL can\'t contain control characters',
-            lambda v: v < (3, 7, 8) or (3, 8, 0) <= v < (3, 8, 3),
-        ),
         # https://github.com/python/cpython/blob/987b712b4aeeece336eed24fcc87a950a756c3e2/Lib/http/client.py#L1288C31-L1288C50
-        (Request('http://127.0.0.1', headers={'foo\n': 'bar'}), 'Invalid header name', None),
+        (Request('http://127.0.0.1', headers={'foo\n': 'bar'}), 'Invalid header name'),
     ])
-    def test_httplib_validation_errors(self, handler, req, match, version_check):
+    def test_httplib_validation_errors(self, handler, req, match):
-        if version_check and version_check(sys.version_info):
-            pytest.skip(f'Python {sys.version} version does not have the required validation for this test.')
-
         with handler() as rh:
             with pytest.raises(RequestError, match=match) as exc_info:
                 validate_and_send(rh, req)

test/test_utils.py

@@ -1347,15 +1347,8 @@ class TestUtil(unittest.TestCase):
         self.assertEqual(extract_attributes('<e _:funny-name1=1>'), {'_:funny-name1': '1'})
         self.assertEqual(extract_attributes('<e x="Fáilte 世界 \U0001f600">'), {'x': 'Fáilte 世界 \U0001f600'})
         self.assertEqual(extract_attributes('<e x="décompose&#769;">'), {'x': 'décompose\u0301'})
-        # "Narrow" Python builds don't support unicode code points outside BMP.
+        self.assertEqual(extract_attributes('<e x="Smile &#128512;!">'), {'x': 'Smile \U0001f600!'})
-        try:
+        # Malformed HTML should not break attribute extraction
-            chr(0x10000)
-            supports_outside_bmp = True
-        except ValueError:
-            supports_outside_bmp = False
-        if supports_outside_bmp:
-            self.assertEqual(extract_attributes('<e x="Smile &#128512;!">'), {'x': 'Smile \U0001f600!'})
-        # Malformed HTML should not break attributes extraction on older Python
         self.assertEqual(extract_attributes('<mal"formed/>'), {})
 
     def test_clean_html(self):

yt_dlp/YoutubeDL.py

@@ -2398,8 +2398,7 @@ class YoutubeDL:
             selectors = []
             current_selector = None
             for type_, string_, start, _, _ in tokens:
-                # ENCODING is only defined in Python 3.x
+                if type_ == tokenize.ENCODING:
-                if type_ == getattr(tokenize, 'ENCODING', None):
                     continue
                 elif type_ in [tokenize.NAME, tokenize.NUMBER]:
                     current_selector = FormatSelector(SINGLE, string_, [])

yt_dlp/aes.py

@@ -293,7 +293,7 @@ def aes_decrypt_text(data, password, key_size_bytes):
     - Mode of operation is 'counter'
 
     @param {str} data                    Base64 encoded string
-    @param {str,unicode} password        Password (will be encoded with utf-8)
+    @param {str} password                Password (will be encoded with UTF-8)
     @param {int} key_size_bytes          Possible values: 16 for 128-Bit, 24 for 192-Bit or 32 for 256-Bit
     @returns {str}                       Decrypted data
     """

yt_dlp/compat/__init__.py

@@ -8,9 +8,8 @@ passthrough_module(__name__, '._deprecated')
 del passthrough_module
 
 
-# HTMLParseError has been deprecated in Python 3.3 and removed in
+# HTMLParseError was deprecated in Python 3.3 and removed in Python 3.5.
-# Python 3.5. Introducing dummy exception for Python >3.5 for compatible
+# Keep a replacement for API compatibility and uniform exception handling.
-# and uniform cross-version exception handling
 class compat_HTMLParseError(ValueError):
     pass
 

yt_dlp/downloader/ism.py

@@ -154,7 +154,7 @@ def write_piff_header(stream, params):
         sample_entry_payload += u16.pack(0x18)  # depth
         sample_entry_payload += s16.pack(-1)  # pre defined
 
-        codec_private_data = binascii.unhexlify(params['codec_private_data'].encode())
+        codec_private_data = binascii.unhexlify(params['codec_private_data'])
         if fourcc in ('H264', 'AVC1'):
             sps, pps = codec_private_data.split(u32.pack(1))[1:]
             avcc_payload = u8.pack(1)  # configuration version

yt_dlp/extractor/common.py

@@ -432,29 +432,28 @@ class InfoExtractor:
 
     chapter:        Name or title of the chapter the video belongs to.
     chapter_number: Number of the chapter the video belongs to, as an integer.
-    chapter_id:     Id of the chapter the video belongs to, as a unicode string.
+    chapter_id:     Id of the chapter the video belongs to.
 
     The following fields should only be used when the video is an episode of some
     series, programme or podcast:
 
     series:         Title of the series or programme the video episode belongs to.
-    series_id:      Id of the series or programme the video episode belongs to, as a unicode string.
+    series_id:      Id of the series or programme the video episode belongs to.
     season:         Title of the season the video episode belongs to.
     season_number:  Number of the season the video episode belongs to, as an integer.
-    season_id:      Id of the season the video episode belongs to, as a unicode string.
+    season_id:      Id of the season the video episode belongs to.
     episode:        Title of the video episode. Unlike mandatory video title field,
                     this field should denote the exact title of the video episode
                     without any kind of decoration.
     episode_number: Number of the video episode within a season, as an integer.
-    episode_id:     Id of the video episode, as a unicode string.
+    episode_id:     Id of the video episode.
 
     The following fields should only be used when the media is a track or a part of
     a music album:
 
     track:          Title of the track.
     track_number:   Number of the track within an album or a disc, as an integer.
-    track_id:       Id of the track (useful in case of custom indexing, e.g. 6.iii),
+    track_id:       Id of the track (useful for custom indexing, e.g. 6.iii).
-                    as a unicode string.
     artists:        List of artists of the track.
     composers:      List of composers of the piece.
     genres:         List of genres of the track.
@@ -487,7 +486,7 @@ class InfoExtractor:
     creator:        Use "creators" instead.
                     The creator of the video.
 
-    Unless mentioned otherwise, the fields should be Unicode strings.
+    Unless mentioned otherwise, the fields should be strings.
 
     Unless mentioned otherwise, None is equivalent to absence of information.
 

yt_dlp/extractor/itv.py

@@ -114,7 +114,7 @@ class ITVIE(InfoExtractor):
         # See: https://github.com/yt-dlp/yt-dlp/issues/986
         platform_tag_subs, featureset_subs = next(
             ((platform_tag, featureset)
-             for platform_tag, featuresets in reversed(list(variants.items())) for featureset in featuresets
+             for platform_tag, featuresets in reversed(variants.items()) for featureset in featuresets
              if try_get(featureset, lambda x: x[2]) == 'outband-webvtt'),
             (None, None))
 
@@ -143,7 +143,7 @@ class ITVIE(InfoExtractor):
         # See: https://github.com/yt-dlp/yt-dlp/issues/986
         platform_tag_video, featureset_video = next(
             ((platform_tag, featureset)
-             for platform_tag, featuresets in reversed(list(variants.items())) for featureset in featuresets
+             for platform_tag, featuresets in reversed(variants.items()) for featureset in featuresets
              if set(try_get(featureset, lambda x: x[:2]) or []) == {'aes', 'hls'}),
             (None, None))
         if not platform_tag_video or not featureset_video:

yt_dlp/extractor/youtube/pot/_director.py

@@ -76,7 +76,7 @@ class YoutubeIEContentProviderLogger(IEContentProviderLogger):
         if self.log_level <= self.LogLevel.ERROR:
             self.__ie._downloader.report_error(
                 self._format_msg(message), is_error=False,
-                tb=''.join(traceback.format_exception(None, cause, cause.__traceback__)) if cause else None)
+                tb=''.join(traceback.format_exception(cause)) if cause else None)
 
 
 class PoTokenCache:

yt_dlp/networking/_helper.py

@@ -108,9 +108,7 @@ def make_ssl_context(
     context = ssl.SSLContext(ssl.PROTOCOL_TLS_CLIENT)
     context.check_hostname = verify
     context.verify_mode = ssl.CERT_REQUIRED if verify else ssl.CERT_NONE
-    # OpenSSL 1.1.1+ Python 3.8+ keylog file
+    context.keylog_filename = os.environ.get('SSLKEYLOGFILE') or None
-    if hasattr(context, 'keylog_filename'):
-        context.keylog_filename = os.environ.get('SSLKEYLOGFILE') or None
 
     # Some servers may reject requests if ALPN extension is not sent. See:
     # https://github.com/python/cpython/issues/85140

yt_dlp/networking/_requests.py

@@ -123,7 +123,7 @@ class RequestsResponseAdapter(Response):
         # Work around issue with `.read(amt)` then `.read()`
         # See: https://github.com/urllib3/urllib3/issues/3636
         if amt is None:
-            # compat: py3.9: Python 3.9 preallocates the whole read buffer, read in chunks
+            # Read in chunks to avoid preallocating a large buffer
             read_chunk = functools.partial(self.fp.read, 1 << 20, decode_content=True)
             return b''.join(iter(read_chunk, b''))
         # Interact with urllib3 response directly.

yt_dlp/networking/_urllib.py

@@ -296,13 +296,9 @@ class UrllibResponseAdapter(Response):
     """
 
     def __init__(self, res: http.client.HTTPResponse | urllib.response.addinfourl):
-        # addinfourl: In Python 3.9+, .status was introduced and .getcode() was deprecated [1]
-        # HTTPResponse: .getcode() was deprecated, .status always existed [2]
-        # 1. https://docs.python.org/3/library/urllib.request.html#urllib.response.addinfourl.getcode
-        # 2. https://docs.python.org/3.10/library/http.client.html#http.client.HTTPResponse.status
         super().__init__(
             fp=res, headers=res.headers, url=res.url,
-            status=getattr(res, 'status', None) or res.getcode(), reason=getattr(res, 'reason', None))
+            status=res.status, reason=getattr(res, 'reason', None))
 
     def read(self, amt=None):
         if self.closed:

yt_dlp/utils/_utils.py

@@ -238,11 +238,9 @@ def find_xpath_attr(node, xpath, key, val=None):
     expr = xpath + (f'[@{key}]' if val is None else f"[@{key}='{val}']")
     return node.find(expr)
 
-# On python2.6 the xml.etree.ElementTree.Element methods don't support
-# the namespace parameter
-
 
 def xpath_with_ns(path, ns_map):
+    """Expand namespace-prefixed names to Clark notation."""
     components = [c.split(':') for c in path.split('/')]
     replaced = []
     for c in components:
@@ -876,7 +874,7 @@ class Popen(subprocess.Popen):
 
         self.__text_mode = kwargs.get('encoding') or kwargs.get('errors') or text or kwargs.get('universal_newlines')
         if text is True:
-            kwargs['universal_newlines'] = True  # For 3.6 compatibility
+            kwargs['text'] = True
             kwargs.setdefault('encoding', 'utf-8')
             kwargs.setdefault('errors', 'replace')
 
@@ -1012,7 +1010,7 @@ class ExtractorError(YoutubeDLError):
     def format_traceback(self):
         return join_nonempty(
             self.traceback and ''.join(traceback.format_tb(self.traceback)),
-            self.cause and ''.join(traceback.format_exception(None, self.cause, self.cause.__traceback__)[1:]),
+            self.cause and ''.join(traceback.format_exception(self.cause)[1:]),
             delim='\n') or None
 
     def __setattr__(self, name, value):
@@ -1960,11 +1958,7 @@ def setproctitle(title):
         libc = ctypes.cdll.LoadLibrary('libc.so.6')
     except OSError:
         return
-    except TypeError:
+
-        # LoadLibrary in Windows Python 2.7.13 only expects
-        # a bytestring, but since unicode_literals turns
-        # every string into a unicode string, it fails.
-        return
     title_bytes = title.encode()
     buf = ctypes.create_string_buffer(len(title_bytes))
     buf.value = title_bytes
@@ -2655,11 +2649,10 @@ def multipart_encode(data, boundary=None):
     Encode a dict to RFC 7578-compliant form-data
 
     data:
-        A dict where keys and values can be either Unicode or bytes-like
+        A dict where keys and values can be either str or bytes-like objects.
-        objects.
     boundary:
-        If specified a Unicode object, it's used as the boundary. Otherwise
+        An ASCII string to use as the boundary. If omitted, a random boundary
-        a random boundary is generated.
+        is generated.
 
     Reference: https://tools.ietf.org/html/rfc7578
     """
@@ -3422,7 +3415,7 @@ def ass_subtitles_timecode(seconds):
 def dfxp2srt(dfxp_data):
     """
     @param dfxp_data A bytes-like object containing DFXP data
-    @returns A unicode object containing converted SRT data
+    @returns A string containing the converted SRT data
     """
     LEGACY_NAMESPACES = (
         (b'http://www.w3.org/ns/ttml', [