X-Git-Url: http://git.bitcoin.ninja/index.cgi?a=blobdiff_plain;f=youtube_dl%2Futils.py;h=a12b0a7de4b00e7879ea4af75d2b4f333dc8f53a;hb=121c09c7be1ac2944f3432122104c1952bfd1f04;hp=bbe554a657b6d3801ac70d2fbbce41125dca5bbd;hpb=bf94e38d3d4aeb4edd5fc10ed9f4e905ee179913;p=youtube-dl diff --git a/youtube_dl/utils.py b/youtube_dl/utils.py index bbe554a65..a12b0a7de 100644 --- a/youtube_dl/utils.py +++ b/youtube_dl/utils.py @@ -41,7 +41,6 @@ from .compat import ( compat_urllib_parse_urlparse, compat_urllib_request, compat_urlparse, - compat_WINFUNCTYPE, shlex_quote, ) @@ -206,6 +205,10 @@ def get_element_by_attribute(attribute, value, html): def clean_html(html): """Clean an HTML snippet into a readable string""" + + if html is None: # Convenience for sanitizing descriptions etc. + return html + # Newline vs
html = html.replace('\n', ' ') html = re.sub(r'\s*<\s*br\s*/?\s*>\s*', '\n', html) @@ -364,7 +367,7 @@ def encodeArgument(s): if not isinstance(s, compat_str): # Legacy code that uses byte strings # Uncomment the following line after fixing all post processors - #assert False, 'Internal error: %r should be of type %r, is %r' % (s, compat_str, type(s)) + # assert False, 'Internal error: %r should be of type %r, is %r' % (s, compat_str, type(s)) s = s.decode('ascii') return encodeFilename(s, True) @@ -389,6 +392,17 @@ def formatSeconds(secs): def make_HTTPS_handler(opts_no_check_certificate, **kwargs): + if hasattr(ssl, 'create_default_context'): # Python >= 3.4 or 2.7.9 + context = ssl.create_default_context(ssl.Purpose.CLIENT_AUTH) + if opts_no_check_certificate: + context.verify_mode = ssl.CERT_NONE + try: + return compat_urllib_request.HTTPSHandler(context=context, **kwargs) + except TypeError: + # Python 2.7.8 + # (create_default_context present but HTTPSHandler has no context=) + pass + if sys.version_info < (3, 2): import httplib @@ -410,22 +424,12 @@ def make_HTTPS_handler(opts_no_check_certificate, **kwargs): def https_open(self, req): return self.do_open(HTTPSConnectionV3, req) return HTTPSHandlerV3(**kwargs) - elif hasattr(ssl, 'create_default_context'): # Python >= 3.4 - context = ssl.create_default_context(ssl.Purpose.CLIENT_AUTH) - context.options &= ~ssl.OP_NO_SSLv3 # Allow older, not-as-secure SSLv3 - if opts_no_check_certificate: - context.verify_mode = ssl.CERT_NONE - return compat_urllib_request.HTTPSHandler(context=context, **kwargs) else: # Python < 3.4 context = ssl.SSLContext(ssl.PROTOCOL_SSLv23) context.verify_mode = (ssl.CERT_NONE if opts_no_check_certificate else ssl.CERT_REQUIRED) context.set_default_verify_paths() - try: - context.load_default_certs() - except AttributeError: - pass # Python < 3.4 return compat_urllib_request.HTTPSHandler(context=context, **kwargs) @@ -464,6 +468,13 @@ class ExtractorError(Exception): return ''.join(traceback.format_tb(self.traceback)) +class UnsupportedError(ExtractorError): + def __init__(self, url): + super(UnsupportedError, self).__init__( + 'Unsupported URL: %s' % url, expected=True) + self.url = url + + class RegexNotFoundError(ExtractorError): """Error when a regex didn't match""" pass @@ -668,9 +679,6 @@ def unified_strdate(date_str, day_first=True): '%b %dth %Y %I:%M%p', '%Y-%m-%d', '%Y/%m/%d', - '%d.%m.%Y', - '%d/%m/%Y', - '%d/%m/%y', '%Y/%m/%d %H:%M:%S', '%Y-%m-%d %H:%M:%S', '%Y-%m-%d %H:%M:%S.%f', @@ -685,10 +693,16 @@ def unified_strdate(date_str, day_first=True): ] if day_first: format_expressions.extend([ + '%d.%m.%Y', + '%d/%m/%Y', + '%d/%m/%y', '%d/%m/%Y %H:%M:%S', ]) else: format_expressions.extend([ + '%m.%d.%Y', + '%m/%d/%Y', + '%m/%d/%y', '%m/%d/%Y %H:%M:%S', ]) for expression in format_expressions: @@ -818,24 +832,24 @@ def _windows_write_string(s, out): if fileno not in WIN_OUTPUT_IDS: return False - GetStdHandle = compat_WINFUNCTYPE( + GetStdHandle = ctypes.WINFUNCTYPE( ctypes.wintypes.HANDLE, ctypes.wintypes.DWORD)( - ("GetStdHandle", ctypes.windll.kernel32)) + (b"GetStdHandle", ctypes.windll.kernel32)) h = GetStdHandle(WIN_OUTPUT_IDS[fileno]) - WriteConsoleW = compat_WINFUNCTYPE( + WriteConsoleW = ctypes.WINFUNCTYPE( ctypes.wintypes.BOOL, ctypes.wintypes.HANDLE, ctypes.wintypes.LPWSTR, ctypes.wintypes.DWORD, ctypes.POINTER(ctypes.wintypes.DWORD), - ctypes.wintypes.LPVOID)(("WriteConsoleW", ctypes.windll.kernel32)) + ctypes.wintypes.LPVOID)((b"WriteConsoleW", ctypes.windll.kernel32)) written = ctypes.wintypes.DWORD(0) - GetFileType = compat_WINFUNCTYPE(ctypes.wintypes.DWORD, ctypes.wintypes.DWORD)(("GetFileType", ctypes.windll.kernel32)) + GetFileType = ctypes.WINFUNCTYPE(ctypes.wintypes.DWORD, ctypes.wintypes.DWORD)((b"GetFileType", ctypes.windll.kernel32)) FILE_TYPE_CHAR = 0x0002 FILE_TYPE_REMOTE = 0x8000 - GetConsoleMode = compat_WINFUNCTYPE( + GetConsoleMode = ctypes.WINFUNCTYPE( ctypes.wintypes.BOOL, ctypes.wintypes.HANDLE, ctypes.POINTER(ctypes.wintypes.DWORD))( - ("GetConsoleMode", ctypes.windll.kernel32)) + (b"GetConsoleMode", ctypes.windll.kernel32)) INVALID_HANDLE_VALUE = ctypes.wintypes.DWORD(-1).value def not_a_console(handle): @@ -1262,18 +1276,25 @@ def check_executable(exe, args=[]): def get_exe_version(exe, args=['--version'], - version_re=r'version\s+([0-9._-a-zA-Z]+)', - unrecognized='present'): + version_re=None, unrecognized='present'): """ Returns the version of the specified executable, or False if the executable is not present """ try: - out, err = subprocess.Popen( + out, _ = subprocess.Popen( [exe] + args, stdout=subprocess.PIPE, stderr=subprocess.STDOUT).communicate() except OSError: return False - firstline = out.partition(b'\n')[0].decode('ascii', 'ignore') - m = re.search(version_re, firstline) + if isinstance(out, bytes): # Python 2.x + out = out.decode('ascii', 'ignore') + return detect_exe_version(out, version_re, unrecognized) + + +def detect_exe_version(output, version_re=None, unrecognized='present'): + assert isinstance(output, compat_str) + if version_re is None: + version_re = r'version\s+([-0-9._a-zA-Z]+)' + m = re.search(version_re, output) if m: return m.group(1) else: @@ -1536,3 +1557,23 @@ def ytdl_is_updateable(): def args_to_str(args): # Get a short string representation for a subprocess command return ' '.join(shlex_quote(a) for a in args) + + +def urlhandle_detect_ext(url_handle): + try: + url_handle.headers + getheader = lambda h: url_handle.headers[h] + except AttributeError: # Python < 3 + getheader = url_handle.info().getheader + + return getheader('Content-Type').split("/")[1] + + +def age_restricted(content_limit, age_limit): + """ Returns True iff the content should be blocked """ + + if age_limit is None: # No limit set + return False + if content_limit is None: + return False # Content available for everyone + return age_limit < content_limit