X-Git-Url: https://git.rapsys.eu/youtubedl/blobdiff_plain/233624c1db781ee7dabbaf88453cf18e248dd20d..684fbbb940adb6ea4043bc437e527888687a53da:/youtube_dl/utils.py diff --git a/youtube_dl/utils.py b/youtube_dl/utils.py index e84d35d..ead9bd8 100644 --- a/youtube_dl/utils.py +++ b/youtube_dl/utils.py @@ -39,6 +39,7 @@ from .compat import ( compat_HTMLParser, compat_basestring, compat_chr, + compat_cookiejar, compat_ctypes_WINFUNCTYPE, compat_etree_fromstring, compat_expanduser, @@ -124,8 +125,8 @@ KNOWN_EXTENSIONS = ( # needed for sanitizing filenames in restricted mode ACCENT_CHARS = dict(zip('ÂÃÄÀÁÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖŐØŒÙÚÛÜŰÝÞßàáâãäåæçèéêëìíîïðñòóôõöőøœùúûüűýþÿ', - itertools.chain('AAAAAA', ['AE'], 'CEEEEIIIIDNOOOOOOO', ['OE'], 'UUUUUYP', ['ss'], - 'aaaaaa', ['ae'], 'ceeeeiiiionooooooo', ['oe'], 'uuuuuypy'))) + itertools.chain('AAAAAA', ['AE'], 'CEEEEIIIIDNOOOOOOO', ['OE'], 'UUUUUY', ['TH', 'ss'], + 'aaaaaa', ['ae'], 'ceeeeiiiionooooooo', ['oe'], 'uuuuuy', ['th'], 'y'))) DATE_FORMATS = ( '%d %B %Y', @@ -183,7 +184,7 @@ DATE_FORMATS_MONTH_FIRST.extend([ ]) PACKED_CODES_RE = r"}\('(.+)',(\d+),(\d+),'([^']+)'\.split\('\|'\)" -JSON_LD_RE = r'(?is)]+type=(["\'])application/ld\+json\1[^>]*>(?P.+?)' +JSON_LD_RE = r'(?is)]+type=(["\']?)application/ld\+json\1[^>]*>(?P.+?)' def preferredencoding(): @@ -545,7 +546,7 @@ def sanitize_url(url): return 'http:%s' % url # Fix some common typos seen so far COMMON_TYPOS = ( - # https://github.com/rg3/youtube-dl/issues/15649 + # https://github.com/ytdl-org/youtube-dl/issues/15649 (r'^httpss://', r'https://'), # https://bx1.be/lives/direct-tv/ (r'^rmtp([es]?)://', r'rtmp\1://'), @@ -595,7 +596,7 @@ def _htmlentity_transform(entity_with_semicolon): numstr = '0%s' % numstr else: base = 10 - # See https://github.com/rg3/youtube-dl/issues/7518 + # See https://github.com/ytdl-org/youtube-dl/issues/7518 try: return compat_chr(int(numstr, base)) except ValueError: @@ -860,8 +861,8 @@ class XAttrMetadataError(YoutubeDLError): self.msg = msg # Parsing code and msg - if (self.code in (errno.ENOSPC, errno.EDQUOT) or - 'No space left' in self.msg or 'Disk quota excedded' in self.msg): + if (self.code in (errno.ENOSPC, errno.EDQUOT) + or 'No space left' in self.msg or 'Disk quota excedded' in self.msg): self.reason = 'NO_SPACE' elif self.code == errno.E2BIG or 'Argument list too long' in self.msg: self.reason = 'VALUE_TOO_LONG' @@ -876,7 +877,7 @@ class XAttrUnavailableError(YoutubeDLError): def _create_http_connection(ydl_handler, http_class, is_https, *args, **kwargs): # Working around python 2 bug (see http://bugs.python.org/issue17849) by limiting # expected HTTP responses to meet HTTP/1.0 or later (see also - # https://github.com/rg3/youtube-dl/issues/6727) + # https://github.com/ytdl-org/youtube-dl/issues/6727) if sys.version_info < (3, 0): kwargs['strict'] = True hc = http_class(*args, **compat_kwargs(kwargs)) @@ -1050,7 +1051,7 @@ class YoutubeDLHandler(compat_urllib_request.HTTPHandler): resp.msg = old_resp.msg del resp.headers['Content-encoding'] # Percent-encode redirect URL of Location HTTP header to satisfy RFC 3986 (see - # https://github.com/rg3/youtube-dl/issues/6457). + # https://github.com/ytdl-org/youtube-dl/issues/6457). if 300 <= resp.code < 400: location = resp.headers.get('Location') if location: @@ -1139,6 +1140,49 @@ class YoutubeDLHTTPSHandler(compat_urllib_request.HTTPSHandler): req, **kwargs) +class YoutubeDLCookieJar(compat_cookiejar.MozillaCookieJar): + _HTTPONLY_PREFIX = '#HttpOnly_' + + def save(self, filename=None, ignore_discard=False, ignore_expires=False): + # Store session cookies with `expires` set to 0 instead of an empty + # string + for cookie in self: + if cookie.expires is None: + cookie.expires = 0 + compat_cookiejar.MozillaCookieJar.save(self, filename, ignore_discard, ignore_expires) + + def load(self, filename=None, ignore_discard=False, ignore_expires=False): + """Load cookies from a file.""" + if filename is None: + if self.filename is not None: + filename = self.filename + else: + raise ValueError(compat_cookiejar.MISSING_FILENAME_TEXT) + + cf = io.StringIO() + with open(filename) as f: + for line in f: + if line.startswith(self._HTTPONLY_PREFIX): + line = line[len(self._HTTPONLY_PREFIX):] + cf.write(compat_str(line)) + cf.seek(0) + self._really_load(cf, filename, ignore_discard, ignore_expires) + # Session cookies are denoted by either `expires` field set to + # an empty string or 0. MozillaCookieJar only recognizes the former + # (see [1]). So we need force the latter to be recognized as session + # cookies on our own. + # Session cookies may be important for cookies-based authentication, + # e.g. usually, when user does not check 'Remember me' check box while + # logging in on a site, some important cookies are stored as session + # cookies so that not recognizing them will result in failed login. + # 1. https://bugs.python.org/issue17164 + for cookie in self: + # Treat `expires=0` cookies as session cookies + if cookie.expires == 0: + cookie.expires = None + cookie.discard = True + + class YoutubeDLCookieProcessor(compat_urllib_request.HTTPCookieProcessor): def __init__(self, cookiejar=None): compat_urllib_request.HTTPCookieProcessor.__init__(self, cookiejar) @@ -1146,7 +1190,7 @@ class YoutubeDLCookieProcessor(compat_urllib_request.HTTPCookieProcessor): def http_response(self, request, response): # Python 2 will choke on next HTTP request in row if there are non-ASCII # characters in Set-Cookie HTTP header of last response (see - # https://github.com/rg3/youtube-dl/issues/6769). + # https://github.com/ytdl-org/youtube-dl/issues/6769). # In order to at least prevent crashing we will percent encode Set-Cookie # header before HTTPCookieProcessor starts processing it. # if sys.version_info < (3, 0) and response.headers: @@ -1409,8 +1453,8 @@ def _windows_write_string(s, out): def not_a_console(handle): if handle == INVALID_HANDLE_VALUE or handle is None: return True - return ((GetFileType(handle) & ~FILE_TYPE_REMOTE) != FILE_TYPE_CHAR or - GetConsoleMode(handle, ctypes.byref(ctypes.wintypes.DWORD())) == 0) + return ((GetFileType(handle) & ~FILE_TYPE_REMOTE) != FILE_TYPE_CHAR + or GetConsoleMode(handle, ctypes.byref(ctypes.wintypes.DWORD())) == 0) if not_a_console(h): return False @@ -1446,8 +1490,8 @@ def write_string(s, out=None, encoding=None): if _windows_write_string(s, out): return - if ('b' in getattr(out, 'mode', '') or - sys.version_info[0] < 3): # Python 2 lies about mode of sys.stderr + if ('b' in getattr(out, 'mode', '') + or sys.version_info[0] < 3): # Python 2 lies about mode of sys.stderr byt = s.encode(encoding or preferredencoding(), 'ignore') out.write(byt) elif hasattr(out, 'buffer'): @@ -1754,6 +1798,14 @@ def parse_resolution(s): return {} +def parse_bitrate(s): + if not isinstance(s, compat_str): + return + mobj = re.search(r'\b(\d+)\s*kbps', s) + if mobj: + return int(mobj.group(1)) + + def month_by_name(name, lang='en'): """ Return the number of a month by (locale-independently) English name """ @@ -1840,7 +1892,7 @@ def urljoin(base, path): path = path.decode('utf-8') if not isinstance(path, compat_str) or not path: return None - if re.match(r'^(?:https?:)?//', path): + if re.match(r'^(?:[a-zA-Z][a-zA-Z0-9+-.]*:)?//', path): return path if isinstance(base, bytes): base = base.decode('utf-8') @@ -1870,7 +1922,7 @@ def int_or_none(v, scale=1, default=None, get_attr=None, invscale=1): return default try: return int(v) * invscale // scale - except ValueError: + except (ValueError, TypeError): return default @@ -1891,7 +1943,7 @@ def float_or_none(v, scale=1, invscale=1, default=None): return default try: return float(v) * invscale / scale - except ValueError: + except (ValueError, TypeError): return default @@ -1899,8 +1951,8 @@ def bool_or_none(v, default=None): return v if isinstance(v, bool) else default -def strip_or_none(v): - return None if v is None else v.strip() +def strip_or_none(v, default=None): + return v.strip() if isinstance(v, compat_str) else default def url_or_none(url): @@ -2000,7 +2052,7 @@ def get_exe_version(exe, args=['--version'], try: # STDIN should be redirected too. On UNIX-like systems, ffmpeg triggers # SIGTTOU if youtube-dl is run in the background. - # See https://github.com/rg3/youtube-dl/issues/955#issuecomment-209789656 + # See https://github.com/ytdl-org/youtube-dl/issues/955#issuecomment-209789656 out, _ = subprocess.Popen( [encodeArgument(exe)] + args, stdin=subprocess.PIPE, @@ -2276,10 +2328,10 @@ def merge_dicts(*dicts): for k, v in a_dict.items(): if v is None: continue - if (k not in merged or - (isinstance(v, compat_str) and v and - isinstance(merged[k], compat_str) and - not merged[k])): + if (k not in merged + or (isinstance(v, compat_str) and v + and isinstance(merged[k], compat_str) + and not merged[k])): merged[k] = v return merged @@ -2605,14 +2657,14 @@ def _match_one(filter_part, dct): if m: op = COMPARISON_OPERATORS[m.group('op')] actual_value = dct.get(m.group('key')) - if (m.group('quotedstrval') is not None or - m.group('strval') is not None or + if (m.group('quotedstrval') is not None + or m.group('strval') is not None # If the original field is a string and matching comparisonvalue is # a number we should respect the origin of the original field # and process comparison value as a string (see - # https://github.com/rg3/youtube-dl/issues/11082). - actual_value is not None and m.group('intval') is not None and - isinstance(actual_value, compat_str)): + # https://github.com/ytdl-org/youtube-dl/issues/11082). + or actual_value is not None and m.group('intval') is not None + and isinstance(actual_value, compat_str)): if m.group('op') not in ('=', '!='): raise ValueError( 'Operator %s does not support string values!' % m.group('op')) @@ -2940,6 +2992,7 @@ class ISO639Utils(object): 'gv': 'glv', 'ha': 'hau', 'he': 'heb', + 'iw': 'heb', # Replaced by he in 1989 revision 'hi': 'hin', 'ho': 'hmo', 'hr': 'hrv', @@ -2949,6 +3002,7 @@ class ISO639Utils(object): 'hz': 'her', 'ia': 'ina', 'id': 'ind', + 'in': 'ind', # Replaced by id in 1989 revision 'ie': 'ile', 'ig': 'ibo', 'ii': 'iii', @@ -3063,6 +3117,7 @@ class ISO639Utils(object): 'wo': 'wol', 'xh': 'xho', 'yi': 'yid', + 'ji': 'yid', # Replaced by yi in 1989 revision 'yo': 'yor', 'za': 'zha', 'zh': 'zho', @@ -3757,7 +3812,7 @@ def urshift(val, n): # Based on png2str() written by @gdkchan and improved by @yokrysty -# Originally posted at https://github.com/rg3/youtube-dl/issues/9706 +# Originally posted at https://github.com/ytdl-org/youtube-dl/issues/9706 def decode_png(png_data): # Reference: https://www.w3.org/TR/PNG/ header = png_data[8:] @@ -3872,7 +3927,7 @@ def write_xattr(path, key, value): if hasattr(xattr, 'set'): # pyxattr # Unicode arguments are not supported in python-pyxattr until # version 0.5.0 - # See https://github.com/rg3/youtube-dl/issues/5498 + # See https://github.com/ytdl-org/youtube-dl/issues/5498 pyxattr_required_version = '0.5.0' if version_tuple(xattr.__version__) < version_tuple(pyxattr_required_version): # TODO: fallback to CLI tools @@ -3918,9 +3973,9 @@ def write_xattr(path, key, value): executable = 'xattr' opts = ['-w', key, value] - cmd = ([encodeFilename(executable, True)] + - [encodeArgument(o) for o in opts] + - [encodeFilename(path, True)]) + cmd = ([encodeFilename(executable, True)] + + [encodeArgument(o) for o in opts] + + [encodeFilename(path, True)]) try: p = subprocess.Popen( @@ -3948,8 +4003,12 @@ def write_xattr(path, key, value): def random_birthday(year_field, month_field, day_field): + start_date = datetime.date(1950, 1, 1) + end_date = datetime.date(1995, 12, 31) + offset = random.randint(0, (end_date - start_date).days) + random_date = start_date + datetime.timedelta(offset) return { - year_field: str(random.randint(1950, 1995)), - month_field: str(random.randint(1, 12)), - day_field: str(random.randint(1, 31)), + year_field: str(random_date.year), + month_field: str(random_date.month), + day_field: str(random_date.day), }