1 from __future__ import unicode_literals
17 import xml.etree.ElementTree
21 import urllib.request as compat_urllib_request
22 except ImportError: # Python 2
23 import urllib2 as compat_urllib_request
26 import urllib.error as compat_urllib_error
27 except ImportError: # Python 2
28 import urllib2 as compat_urllib_error
31 import urllib.parse as compat_urllib_parse
32 except ImportError: # Python 2
33 import urllib as compat_urllib_parse
36 from urllib.parse import urlparse as compat_urllib_parse_urlparse
37 except ImportError: # Python 2
38 from urlparse import urlparse as compat_urllib_parse_urlparse
41 import urllib.parse as compat_urlparse
42 except ImportError: # Python 2
43 import urlparse as compat_urlparse
46 import urllib.response as compat_urllib_response
47 except ImportError: # Python 2
48 import urllib as compat_urllib_response
51 import http.cookiejar as compat_cookiejar
52 except ImportError: # Python 2
53 import cookielib as compat_cookiejar
56 import http.cookies as compat_cookies
57 except ImportError: # Python 2
58 import Cookie as compat_cookies
61 import html.entities as compat_html_entities
62 except ImportError: # Python 2
63 import htmlentitydefs as compat_html_entities
66 import http.client as compat_http_client
67 except ImportError: # Python 2
68 import httplib as compat_http_client
71 from urllib.error import HTTPError as compat_HTTPError
72 except ImportError: # Python 2
73 from urllib2 import HTTPError as compat_HTTPError
76 from urllib.request import urlretrieve as compat_urlretrieve
77 except ImportError: # Python 2
78 from urllib import urlretrieve as compat_urlretrieve
82 from subprocess import DEVNULL
83 compat_subprocess_get_DEVNULL = lambda: DEVNULL
85 compat_subprocess_get_DEVNULL = lambda: open(os.path.devnull, 'w')
88 import http.server as compat_http_server
90 import BaseHTTPServer as compat_http_server
93 compat_str = unicode # Python 2
98 from urllib.parse import unquote_to_bytes as compat_urllib_parse_unquote_to_bytes
99 from urllib.parse import unquote as compat_urllib_parse_unquote
100 from urllib.parse import unquote_plus as compat_urllib_parse_unquote_plus
101 except ImportError: # Python 2
102 _asciire = (compat_urllib_parse._asciire if hasattr(compat_urllib_parse, '_asciire')
103 else re.compile('([\x00-\x7f]+)'))
105 # HACK: The following are the correct unquote_to_bytes, unquote and unquote_plus
106 # implementations from cpython 3.4.3's stdlib. Python 2's version
107 # is apparently broken (see https://github.com/rg3/youtube-dl/pull/6244)
109 def compat_urllib_parse_unquote_to_bytes(string):
110 """unquote_to_bytes('abc%20def') -> b'abc def'."""
111 # Note: strings are encoded as UTF-8. This is only an issue if it contains
112 # unescaped non-ASCII characters, which URIs should not.
114 # Is it a string-like object?
117 if isinstance(string, compat_str):
118 string = string.encode('utf-8')
119 bits = string.split(b'%')
124 for item in bits[1:]:
126 append(compat_urllib_parse._hextochr[item[:2]])
133 def compat_urllib_parse_unquote(string, encoding='utf-8', errors='replace'):
134 """Replace %xx escapes by their single-character equivalent. The optional
135 encoding and errors parameters specify how to decode percent-encoded
136 sequences into Unicode characters, as accepted by the bytes.decode()
138 By default, percent-encoded sequences are decoded with UTF-8, and invalid
139 sequences are replaced by a placeholder character.
141 unquote('abc%20def') -> 'abc def'.
143 if '%' not in string:
150 bits = _asciire.split(string)
153 for i in range(1, len(bits), 2):
154 append(compat_urllib_parse_unquote_to_bytes(bits[i]).decode(encoding, errors))
158 def compat_urllib_parse_unquote_plus(string, encoding='utf-8', errors='replace'):
159 """Like unquote(), but also replace plus signs by spaces, as required for
160 unquoting HTML form values.
162 unquote_plus('%7e/abc+def') -> '~/abc def'
164 string = string.replace('+', ' ')
165 return compat_urllib_parse_unquote(string, encoding, errors)
168 from urllib.request import DataHandler as compat_urllib_request_DataHandler
169 except ImportError: # Python < 3.4
170 # Ported from CPython 98774:1733b3bd46db, Lib/urllib/request.py
171 class compat_urllib_request_DataHandler(compat_urllib_request.BaseHandler):
172 def data_open(self, req):
173 # data URLs as specified in RFC 2397.
175 # ignores POSTed data
178 # dataurl := "data:" [ mediatype ] [ ";base64" ] "," data
179 # mediatype := [ type "/" subtype ] *( ";" parameter )
181 # parameter := attribute "=" value
182 url = req.get_full_url()
184 scheme, data = url.split(':', 1)
185 mediatype, data = data.split(',', 1)
187 # even base64 encoded data URLs might be quoted so unquote in any case:
188 data = compat_urllib_parse_unquote_to_bytes(data)
189 if mediatype.endswith(';base64'):
190 data = binascii.a2b_base64(data)
191 mediatype = mediatype[:-7]
194 mediatype = 'text/plain;charset=US-ASCII'
196 headers = email.message_from_string(
197 'Content-type: %s\nContent-length: %d\n' % (mediatype, len(data)))
199 return compat_urllib_response.addinfourl(io.BytesIO(data), headers, url)
202 compat_basestring = basestring # Python 2
204 compat_basestring = str
207 compat_chr = unichr # Python 2
212 from xml.etree.ElementTree import ParseError as compat_xml_parse_error
213 except ImportError: # Python 2.6
214 from xml.parsers.expat import ExpatError as compat_xml_parse_error
216 if sys.version_info[0] >= 3:
217 compat_etree_fromstring = xml.etree.ElementTree.fromstring
219 # python 2.x tries to encode unicode strings with ascii (see the
220 # XMLParser._fixtext method)
221 etree = xml.etree.ElementTree
224 _etree_iter = etree.Element.iter
225 except AttributeError: # Python <=2.6
226 def _etree_iter(root):
227 for el in root.findall('*'):
229 for sub in _etree_iter(el):
232 # on 2.6 XML doesn't have a parser argument, function copied from CPython
234 def _XML(text, parser=None):
236 parser = etree.XMLParser(target=etree.TreeBuilder())
238 return parser.close()
240 def _element_factory(*args, **kwargs):
241 el = etree.Element(*args, **kwargs)
242 for k, v in el.items():
243 if isinstance(v, bytes):
244 el.set(k, v.decode('utf-8'))
247 def compat_etree_fromstring(text):
248 doc = _XML(text, parser=etree.XMLParser(target=etree.TreeBuilder(element_factory=_element_factory)))
249 for el in _etree_iter(doc):
250 if el.text is not None and isinstance(el.text, bytes):
251 el.text = el.text.decode('utf-8')
255 from urllib.parse import parse_qs as compat_parse_qs
256 except ImportError: # Python 2
257 # HACK: The following is the correct parse_qs implementation from cpython 3's stdlib.
258 # Python 2's version is apparently totally broken
260 def _parse_qsl(qs, keep_blank_values=False, strict_parsing=False,
261 encoding='utf-8', errors='replace'):
262 qs, _coerce_result = qs, compat_str
263 pairs = [s2 for s1 in qs.split('&') for s2 in s1.split(';')]
265 for name_value in pairs:
266 if not name_value and not strict_parsing:
268 nv = name_value.split('=', 1)
271 raise ValueError('bad query field: %r' % (name_value,))
272 # Handle case of a control-name with no equal sign
273 if keep_blank_values:
277 if len(nv[1]) or keep_blank_values:
278 name = nv[0].replace('+', ' ')
279 name = compat_urllib_parse_unquote(
280 name, encoding=encoding, errors=errors)
281 name = _coerce_result(name)
282 value = nv[1].replace('+', ' ')
283 value = compat_urllib_parse_unquote(
284 value, encoding=encoding, errors=errors)
285 value = _coerce_result(value)
286 r.append((name, value))
289 def compat_parse_qs(qs, keep_blank_values=False, strict_parsing=False,
290 encoding='utf-8', errors='replace'):
292 pairs = _parse_qsl(qs, keep_blank_values, strict_parsing,
293 encoding=encoding, errors=errors)
294 for name, value in pairs:
295 if name in parsed_result:
296 parsed_result[name].append(value)
298 parsed_result[name] = [value]
302 from shlex import quote as shlex_quote
303 except ImportError: # Python < 3.3
305 if re.match(r'^[-_\w./]+$', s):
308 return "'" + s.replace("'", "'\"'\"'") + "'"
311 if sys.version_info >= (2, 7, 3):
312 compat_shlex_split = shlex.split
314 # Working around shlex issue with unicode strings on some python 2
315 # versions (see http://bugs.python.org/issue1548891)
316 def compat_shlex_split(s, comments=False, posix=True):
317 if isinstance(s, compat_str):
318 s = s.encode('utf-8')
319 return shlex.split(s, comments, posix)
329 compat_os_name = os._name if os.name == 'java' else os.name
332 if sys.version_info >= (3, 0):
333 compat_getenv = os.getenv
334 compat_expanduser = os.path.expanduser
336 # Environment variables should be decoded with filesystem encoding.
337 # Otherwise it will fail if any non-ASCII characters present (see #3854 #3217 #2918)
339 def compat_getenv(key, default=None):
340 from .utils import get_filesystem_encoding
341 env = os.getenv(key, default)
343 env = env.decode(get_filesystem_encoding())
346 # HACK: The default implementations of os.path.expanduser from cpython do not decode
347 # environment variables with filesystem encoding. We will work around this by
348 # providing adjusted implementations.
349 # The following are os.path.expanduser implementations from cpython 2.7.8 stdlib
350 # for different platforms with correct environment variables decoding.
352 if compat_os_name == 'posix':
353 def compat_expanduser(path):
354 """Expand ~ and ~user constructions. If user or $HOME is unknown,
356 if not path.startswith('~'):
358 i = path.find('/', 1)
362 if 'HOME' not in os.environ:
364 userhome = pwd.getpwuid(os.getuid()).pw_dir
366 userhome = compat_getenv('HOME')
370 pwent = pwd.getpwnam(path[1:i])
373 userhome = pwent.pw_dir
374 userhome = userhome.rstrip('/')
375 return (userhome + path[i:]) or '/'
376 elif compat_os_name == 'nt' or compat_os_name == 'ce':
377 def compat_expanduser(path):
378 """Expand ~ and ~user constructs.
380 If user or $HOME is unknown, do nothing."""
384 while i < n and path[i] not in '/\\':
387 if 'HOME' in os.environ:
388 userhome = compat_getenv('HOME')
389 elif 'USERPROFILE' in os.environ:
390 userhome = compat_getenv('USERPROFILE')
391 elif 'HOMEPATH' not in os.environ:
395 drive = compat_getenv('HOMEDRIVE')
398 userhome = os.path.join(drive, compat_getenv('HOMEPATH'))
401 userhome = os.path.join(os.path.dirname(userhome), path[1:i])
403 return userhome + path[i:]
405 compat_expanduser = os.path.expanduser
408 if sys.version_info < (3, 0):
410 from .utils import preferredencoding
411 print(s.encode(preferredencoding(), 'xmlcharrefreplace'))
414 assert isinstance(s, compat_str)
419 subprocess_check_output = subprocess.check_output
420 except AttributeError:
421 def subprocess_check_output(*args, **kwargs):
422 assert 'input' not in kwargs
423 p = subprocess.Popen(*args, stdout=subprocess.PIPE, **kwargs)
424 output, _ = p.communicate()
427 raise subprocess.CalledProcessError(ret, p.args, output=output)
430 if sys.version_info < (3, 0) and sys.platform == 'win32':
431 def compat_getpass(prompt, *args, **kwargs):
432 if isinstance(prompt, compat_str):
433 from .utils import preferredencoding
434 prompt = prompt.encode(preferredencoding())
435 return getpass.getpass(prompt, *args, **kwargs)
437 compat_getpass = getpass.getpass
439 # Python < 2.6.5 require kwargs to be bytes
443 _testfunc(**{'x': 0})
445 def compat_kwargs(kwargs):
446 return dict((bytes(k), v) for k, v in kwargs.items())
448 compat_kwargs = lambda kwargs: kwargs
451 if sys.version_info < (2, 7):
452 def compat_socket_create_connection(address, timeout, source_address=None):
455 for res in socket.getaddrinfo(host, port, 0, socket.SOCK_STREAM):
456 af, socktype, proto, canonname, sa = res
459 sock = socket.socket(af, socktype, proto)
460 sock.settimeout(timeout)
462 sock.bind(source_address)
465 except socket.error as _:
472 raise socket.error('getaddrinfo returns an empty list')
474 compat_socket_create_connection = socket.create_connection
477 # Fix https://github.com/rg3/youtube-dl/issues/4223
478 # See http://bugs.python.org/issue9161 for what is broken
479 def workaround_optparse_bug9161():
480 op = optparse.OptionParser()
481 og = optparse.OptionGroup(op, 'foo')
485 real_add_option = optparse.OptionGroup.add_option
487 def _compat_add_option(self, *args, **kwargs):
489 v.encode('ascii', 'replace') if isinstance(v, compat_str)
491 bargs = [enc(a) for a in args]
493 (k, enc(v)) for k, v in kwargs.items())
494 return real_add_option(self, *bargs, **bkwargs)
495 optparse.OptionGroup.add_option = _compat_add_option
497 if hasattr(shutil, 'get_terminal_size'): # Python >= 3.3
498 compat_get_terminal_size = shutil.get_terminal_size
500 _terminal_size = collections.namedtuple('terminal_size', ['columns', 'lines'])
502 def compat_get_terminal_size(fallback=(80, 24)):
503 columns = compat_getenv('COLUMNS')
505 columns = int(columns)
508 lines = compat_getenv('LINES')
514 if columns is None or lines is None or columns <= 0 or lines <= 0:
516 sp = subprocess.Popen(
518 stdout=subprocess.PIPE, stderr=subprocess.PIPE)
519 out, err = sp.communicate()
520 _lines, _columns = map(int, out.split())
522 _columns, _lines = _terminal_size(*fallback)
524 if columns is None or columns <= 0:
526 if lines is None or lines <= 0:
528 return _terminal_size(columns, lines)
531 itertools.count(start=0, step=1)
532 compat_itertools_count = itertools.count
533 except TypeError: # Python 2.6
534 def compat_itertools_count(start=0, step=1):
540 if sys.version_info >= (3, 0):
541 from tokenize import tokenize as compat_tokenize_tokenize
543 from tokenize import generate_tokens as compat_tokenize_tokenize
551 'compat_etree_fromstring',
553 'compat_get_terminal_size',
556 'compat_html_entities',
557 'compat_http_client',
558 'compat_http_server',
559 'compat_itertools_count',
565 'compat_shlex_split',
566 'compat_socket_create_connection',
568 'compat_subprocess_get_DEVNULL',
569 'compat_tokenize_tokenize',
570 'compat_urllib_error',
571 'compat_urllib_parse',
572 'compat_urllib_parse_unquote',
573 'compat_urllib_parse_unquote_plus',
574 'compat_urllib_parse_unquote_to_bytes',
575 'compat_urllib_parse_urlparse',
576 'compat_urllib_request',
577 'compat_urllib_request_DataHandler',
578 'compat_urllib_response',
580 'compat_urlretrieve',
581 'compat_xml_parse_error',
583 'subprocess_check_output',
584 'workaround_optparse_bug9161',