X-Git-Url: https://git.rapsys.eu/youtubedl/blobdiff_plain/3477c644417600d9ec8f8d2a44f82da0a4b15eb5..7b550e005da7fd498cfbf1b7c04b05e5540eb6fc:/youtube_dl/utils.py
diff --git a/youtube_dl/utils.py b/youtube_dl/utils.py
index d39f313..d2d3c1a 100644
--- a/youtube_dl/utils.py
+++ b/youtube_dl/utils.py
@@ -1,31 +1,32 @@
#!/usr/bin/env python
-# -*- coding: utf-8 -*-
+# coding: utf-8
from __future__ import unicode_literals
import base64
+import binascii
import calendar
import codecs
import contextlib
import ctypes
import datetime
import email.utils
+import email.header
import errno
import functools
import gzip
-import itertools
import io
+import itertools
import json
import locale
import math
import operator
import os
-import pipes
import platform
+import random
import re
-import ssl
import socket
-import struct
+import ssl
import subprocess
import sys
import tempfile
@@ -34,29 +35,54 @@ import xml.etree.ElementTree
import zlib
from .compat import (
+ compat_HTMLParseError,
+ compat_HTMLParser,
compat_basestring,
compat_chr,
+ compat_cookiejar,
+ compat_ctypes_WINFUNCTYPE,
compat_etree_fromstring,
+ compat_expanduser,
compat_html_entities,
+ compat_html_entities_html5,
compat_http_client,
compat_kwargs,
+ compat_os_name,
compat_parse_qs,
- compat_socket_create_connection,
+ compat_shlex_quote,
compat_str,
+ compat_struct_pack,
+ compat_struct_unpack,
compat_urllib_error,
compat_urllib_parse,
+ compat_urllib_parse_urlencode,
compat_urllib_parse_urlparse,
+ compat_urllib_parse_unquote_plus,
compat_urllib_request,
compat_urlparse,
- shlex_quote,
+ compat_xpath,
+)
+
+from .socks import (
+ ProxyType,
+ sockssocket,
)
+def register_socks_protocols():
+ # "Register" SOCKS protocols
+ # In Python < 2.6.5, urlsplit() suffers from bug https://bugs.python.org/issue7904
+ # URLs with protocols not in urlparse.uses_netloc are not handled correctly
+ for scheme in ('socks', 'socks4', 'socks4a', 'socks5'):
+ if scheme not in compat_urlparse.uses_netloc:
+ compat_urlparse.uses_netloc.append(scheme)
+
+
# This is not clearly defined otherwise
compiled_regex_type = type(re.compile(''))
std_headers = {
- 'User-Agent': 'Mozilla/5.0 (X11; Linux x86_64; rv:10.0) Gecko/20150101 Firefox/20.0 (Chrome)',
+ 'User-Agent': 'Mozilla/5.0 (X11; Linux x86_64; rv:59.0) Gecko/20100101 Firefox/59.0',
'Accept-Charset': 'ISO-8859-1,utf-8;q=0.7,*;q=0.7',
'Accept': 'text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8',
'Accept-Encoding': 'gzip, deflate',
@@ -64,12 +90,102 @@ std_headers = {
}
+USER_AGENTS = {
+ 'Safari': 'Mozilla/5.0 (X11; Linux x86_64; rv:10.0) AppleWebKit/533.20.25 (KHTML, like Gecko) Version/5.0.4 Safari/533.20.27',
+}
+
+
NO_DEFAULT = object()
ENGLISH_MONTH_NAMES = [
'January', 'February', 'March', 'April', 'May', 'June',
'July', 'August', 'September', 'October', 'November', 'December']
+MONTH_NAMES = {
+ 'en': ENGLISH_MONTH_NAMES,
+ 'fr': [
+ 'janvier', 'février', 'mars', 'avril', 'mai', 'juin',
+ 'juillet', 'août', 'septembre', 'octobre', 'novembre', 'décembre'],
+}
+
+KNOWN_EXTENSIONS = (
+ 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'aac',
+ 'flv', 'f4v', 'f4a', 'f4b',
+ 'webm', 'ogg', 'ogv', 'oga', 'ogx', 'spx', 'opus',
+ 'mkv', 'mka', 'mk3d',
+ 'avi', 'divx',
+ 'mov',
+ 'asf', 'wmv', 'wma',
+ '3gp', '3g2',
+ 'mp3',
+ 'flac',
+ 'ape',
+ 'wav',
+ 'f4f', 'f4m', 'm3u8', 'smil')
+
+# needed for sanitizing filenames in restricted mode
+ACCENT_CHARS = dict(zip('ÃÃÃÃÃÃ
ÃÃÃÃÃÃÃÃÃÃÃÃÃÃÃÃÃÅÃÅÃÃÃÃÅ°ÃÃÃà áâãäåæçèéêëìÃîïðñòóôõöÅøÅùúûüűýþÿ',
+ itertools.chain('AAAAAA', ['AE'], 'CEEEEIIIIDNOOOOOOO', ['OE'], 'UUUUUYP', ['ss'],
+ 'aaaaaa', ['ae'], 'ceeeeiiiionooooooo', ['oe'], 'uuuuuypy')))
+
+DATE_FORMATS = (
+ '%d %B %Y',
+ '%d %b %Y',
+ '%B %d %Y',
+ '%B %dst %Y',
+ '%B %dnd %Y',
+ '%B %dth %Y',
+ '%b %d %Y',
+ '%b %dst %Y',
+ '%b %dnd %Y',
+ '%b %dth %Y',
+ '%b %dst %Y %I:%M',
+ '%b %dnd %Y %I:%M',
+ '%b %dth %Y %I:%M',
+ '%Y %m %d',
+ '%Y-%m-%d',
+ '%Y/%m/%d',
+ '%Y/%m/%d %H:%M',
+ '%Y/%m/%d %H:%M:%S',
+ '%Y-%m-%d %H:%M',
+ '%Y-%m-%d %H:%M:%S',
+ '%Y-%m-%d %H:%M:%S.%f',
+ '%d.%m.%Y %H:%M',
+ '%d.%m.%Y %H.%M',
+ '%Y-%m-%dT%H:%M:%SZ',
+ '%Y-%m-%dT%H:%M:%S.%fZ',
+ '%Y-%m-%dT%H:%M:%S.%f0Z',
+ '%Y-%m-%dT%H:%M:%S',
+ '%Y-%m-%dT%H:%M:%S.%f',
+ '%Y-%m-%dT%H:%M',
+ '%b %d %Y at %H:%M',
+ '%b %d %Y at %H:%M:%S',
+ '%B %d %Y at %H:%M',
+ '%B %d %Y at %H:%M:%S',
+)
+
+DATE_FORMATS_DAY_FIRST = list(DATE_FORMATS)
+DATE_FORMATS_DAY_FIRST.extend([
+ '%d-%m-%Y',
+ '%d.%m.%Y',
+ '%d.%m.%y',
+ '%d/%m/%Y',
+ '%d/%m/%y',
+ '%d/%m/%Y %H:%M:%S',
+])
+
+DATE_FORMATS_MONTH_FIRST = list(DATE_FORMATS)
+DATE_FORMATS_MONTH_FIRST.extend([
+ '%m-%d-%Y',
+ '%m.%d.%Y',
+ '%m/%d/%Y',
+ '%m/%d/%y',
+ '%m/%d/%Y %H:%M:%S',
+])
+
+PACKED_CODES_RE = r"}\('(.+)',(\d+),(\d+),'([^']+)'\.split\('\|'\)"
+JSON_LD_RE = r'(?is)'
+
def preferredencoding():
"""Get preferred encoding.
@@ -144,18 +260,11 @@ if sys.version_info >= (2, 7):
def find_xpath_attr(node, xpath, key, val=None):
""" Find the xpath xpath[@key=val] """
assert re.match(r'^[a-zA-Z_-]+$', key)
- if val:
- assert re.match(r'^[a-zA-Z0-9@\s:._-]*$', val)
expr = xpath + ('[@%s]' % key if val is None else "[@%s='%s']" % (key, val))
return node.find(expr)
else:
def find_xpath_attr(node, xpath, key, val=None):
- # Here comes the crazy part: In 2.6, if the xpath is a unicode,
- # .//node does not match if a node is a direct child of . !
- if isinstance(xpath, compat_str):
- xpath = xpath.encode('ascii')
-
- for f in node.findall(xpath):
+ for f in node.findall(compat_xpath(xpath)):
if key not in f.attrib:
continue
if val is None or f.attrib.get(key) == val:
@@ -180,9 +289,7 @@ def xpath_with_ns(path, ns_map):
def xpath_element(node, xpath, name=None, fatal=False, default=NO_DEFAULT):
def _find_xpath(xpath):
- if sys.version_info < (2, 7): # Crazy 2.6
- xpath = xpath.encode('ascii')
- return node.find(xpath)
+ return node.find(compat_xpath(xpath))
if isinstance(xpath, (str, compat_str)):
n = _find_xpath(xpath)
@@ -233,30 +340,86 @@ def xpath_attr(node, xpath, key, name=None, fatal=False, default=NO_DEFAULT):
def get_element_by_id(id, html):
"""Return the content of the tag with the specified ID in the passed HTML document"""
- return get_element_by_attribute("id", id, html)
+ return get_element_by_attribute('id', id, html)
+
+
+def get_element_by_class(class_name, html):
+ """Return the content of the first tag with the specified class in the passed HTML document"""
+ retval = get_elements_by_class(class_name, html)
+ return retval[0] if retval else None
+
+
+def get_element_by_attribute(attribute, value, html, escape_value=True):
+ retval = get_elements_by_attribute(attribute, value, html, escape_value)
+ return retval[0] if retval else None
+
+
+def get_elements_by_class(class_name, html):
+ """Return the content of all tags with the specified class in the passed HTML document as a list"""
+ return get_elements_by_attribute(
+ 'class', r'[^\'"]*\b%s\b[^\'"]*' % re.escape(class_name),
+ html, escape_value=False)
-def get_element_by_attribute(attribute, value, html):
+def get_elements_by_attribute(attribute, value, html, escape_value=True):
"""Return the content of the tag with the specified attribute in the passed HTML document"""
- m = re.search(r'''(?xs)
+ value = re.escape(value) if escape_value else value
+
+ retlist = []
+ for m in re.finditer(r'''(?xs)
<([a-zA-Z0-9:._-]+)
- (?:\s+[a-zA-Z0-9:._-]+(?:=[a-zA-Z0-9:._-]+|="[^"]+"|='[^']+'))*?
+ (?:\s+[a-zA-Z0-9:._-]+(?:=[a-zA-Z0-9:._-]*|="[^"]*"|='[^']*'|))*?
\s+%s=['"]?%s['"]?
- (?:\s+[a-zA-Z0-9:._-]+(?:=[a-zA-Z0-9:._-]+|="[^"]+"|='[^']+'))*?
+ (?:\s+[a-zA-Z0-9:._-]+(?:=[a-zA-Z0-9:._-]*|="[^"]*"|='[^']*'|))*?
\s*>
(?P.*?)
\1>
- ''' % (re.escape(attribute), re.escape(value)), html)
-
- if not m:
- return None
- res = m.group('content')
-
- if res.startswith('"') or res.startswith("'"):
- res = res[1:-1]
-
- return unescapeHTML(res)
+ ''' % (re.escape(attribute), value), html):
+ res = m.group('content')
+
+ if res.startswith('"') or res.startswith("'"):
+ res = res[1:-1]
+
+ retlist.append(unescapeHTML(res))
+
+ return retlist
+
+
+class HTMLAttributeParser(compat_HTMLParser):
+ """Trivial HTML parser to gather the attributes for a single element"""
+ def __init__(self):
+ self.attrs = {}
+ compat_HTMLParser.__init__(self)
+
+ def handle_starttag(self, tag, attrs):
+ self.attrs = dict(attrs)
+
+
+def extract_attributes(html_element):
+ """Given a string for an HTML element such as
+
+ Decode and return a dictionary of attributes.
+ {
+ 'a': 'foo', 'b': 'bar', c: 'baz', d: 'boz',
+ 'empty': '', 'noval': None, 'entity': '&',
+ 'sq': '"', 'dq': '\''
+ }.
+ NB HTMLParser is stricter in Python 2.6 & 3.2 than in later versions,
+ but the cases in the unit test will work for all of 2.6, 2.7, 3.2-3.5.
+ """
+ parser = HTMLAttributeParser()
+ try:
+ parser.feed(html_element)
+ parser.close()
+ # Older Python may throw HTMLParseError in case of malformed HTML
+ except compat_HTMLParseError:
+ pass
+ return parser.attrs
def clean_html(html):
@@ -267,8 +430,8 @@ def clean_html(html):
# Newline vs
html = html.replace('\n', ' ')
- html = re.sub(r'\s*<\s*br\s*/?\s*>\s*', '\n', html)
- html = re.sub(r'<\s*/\s*p\s*>\s*<\s*p[^>]*>', '\n', html)
+ html = re.sub(r'(?u)\s*<\s*br\s*/?\s*>\s*', '\n', html)
+ html = re.sub(r'(?u)<\s*/\s*p\s*>\s*<\s*p[^>]*>', '\n', html)
# Strip html tags
html = re.sub('<.*?>', '', html)
# Replace html entities
@@ -320,9 +483,12 @@ def timeconvert(timestr):
def sanitize_filename(s, restricted=False, is_id=False):
"""Sanitizes a string so it could be used as part of a filename.
If restricted is set, use a stricter subset of allowed characters.
- Set is_id if this is not an arbitrary string, but an ID that should be kept if possible
+ Set is_id if this is not an arbitrary string, but an ID that should be kept
+ if possible.
"""
def replace_insane(char):
+ if restricted and char in ACCENT_CHARS:
+ return ACCENT_CHARS[char]
if char == '?' or ord(char) < 32 or ord(char) == 127:
return ''
elif char == '"':
@@ -366,13 +532,40 @@ def sanitize_path(s):
if drive_or_unc:
norm_path.pop(0)
sanitized_path = [
- path_part if path_part in ['.', '..'] else re.sub('(?:[/<>:"\\|\\\\?\\*]|[\s.]$)', '#', path_part)
+ path_part if path_part in ['.', '..'] else re.sub(r'(?:[/<>:"\|\\?\*]|[\s.]$)', '#', path_part)
for path_part in norm_path]
if drive_or_unc:
sanitized_path.insert(0, drive_or_unc + os.path.sep)
return os.path.join(*sanitized_path)
+def sanitize_url(url):
+ # Prepend protocol-less URLs with `http:` scheme in order to mitigate
+ # the number of unwanted failures due to missing protocol
+ if url.startswith('//'):
+ return 'http:%s' % url
+ # Fix some common typos seen so far
+ COMMON_TYPOS = (
+ # https://github.com/rg3/youtube-dl/issues/15649
+ (r'^httpss://', r'https://'),
+ # https://bx1.be/lives/direct-tv/
+ (r'^rmtp([es]?)://', r'rtmp\1://'),
+ )
+ for mistake, fixup in COMMON_TYPOS:
+ if re.match(mistake, url):
+ return re.sub(mistake, fixup, url)
+ return url
+
+
+def sanitized_Request(url, *args, **kwargs):
+ return compat_urllib_request.Request(sanitize_url(url), *args, **kwargs)
+
+
+def expand_path(s):
+ """Expand shell variables and ~"""
+ return os.path.expandvars(compat_expanduser(s))
+
+
def orderedSet(iterable):
""" Remove all duplicates from the input iterable """
res = []
@@ -382,12 +575,19 @@ def orderedSet(iterable):
return res
-def _htmlentity_transform(entity):
+def _htmlentity_transform(entity_with_semicolon):
"""Transforms an HTML entity to a character."""
+ entity = entity_with_semicolon[:-1]
+
# Known non-numeric HTML entity
if entity in compat_html_entities.name2codepoint:
return compat_chr(compat_html_entities.name2codepoint[entity])
+ # TODO: HTML5 allows entities without a semicolon. For example,
+ # 'Éric' should be decoded as 'Ãric'.
+ if entity_with_semicolon in compat_html_entities_html5:
+ return compat_html_entities_html5[entity_with_semicolon]
+
mobj = re.match(r'#(x[0-9a-fA-F]+|[0-9]+)', entity)
if mobj is not None:
numstr = mobj.group(1)
@@ -396,10 +596,14 @@ def _htmlentity_transform(entity):
numstr = '0%s' % numstr
else:
base = 10
- return compat_chr(int(numstr, base))
+ # See https://github.com/rg3/youtube-dl/issues/7518
+ try:
+ return compat_chr(int(numstr, base))
+ except ValueError:
+ pass
# Unknown entity in name, return its literal representation
- return ('&%s;' % entity)
+ return '&%s;' % entity
def unescapeHTML(s):
@@ -408,7 +612,7 @@ def unescapeHTML(s):
assert type(s) == compat_str
return re.sub(
- r'&([^;]+);', lambda m: _htmlentity_transform(m.group(1)), s)
+ r'&([^&;]+;)', lambda m: _htmlentity_transform(m.group(1)), s)
def get_subprocess_encoding():
@@ -440,6 +644,10 @@ def encodeFilename(s, for_subprocess=False):
if not for_subprocess and sys.platform == 'win32' and sys.getwindowsversion()[0] >= 5:
return s
+ # Jython assumes filenames are Unicode strings though reported as Python 2.x compatible
+ if sys.platform.startswith('java'):
+ return s
+
return s.encode(get_subprocess_encoding(), 'ignore')
@@ -522,7 +730,12 @@ def bug_reports_message():
return msg
-class ExtractorError(Exception):
+class YoutubeDLError(Exception):
+ """Base exception for YoutubeDL errors."""
+ pass
+
+
+class ExtractorError(YoutubeDLError):
"""Error during info extraction."""
def __init__(self, msg, tb=None, expected=False, cause=None, video_id=None):
@@ -563,7 +776,19 @@ class RegexNotFoundError(ExtractorError):
pass
-class DownloadError(Exception):
+class GeoRestrictedError(ExtractorError):
+ """Geographic restriction Error exception.
+
+ This exception may be thrown when a video is not available from your
+ geographic location due to geographic restrictions imposed by a website.
+ """
+ def __init__(self, msg, countries=None):
+ super(GeoRestrictedError, self).__init__(msg, expected=True)
+ self.msg = msg
+ self.countries = countries
+
+
+class DownloadError(YoutubeDLError):
"""Download Error exception.
This exception may be thrown by FileDownloader objects if they are not
@@ -577,7 +802,7 @@ class DownloadError(Exception):
self.exc_info = exc_info
-class SameFileError(Exception):
+class SameFileError(YoutubeDLError):
"""Same File exception.
This exception will be thrown by FileDownloader objects if they detect
@@ -586,7 +811,7 @@ class SameFileError(Exception):
pass
-class PostProcessingError(Exception):
+class PostProcessingError(YoutubeDLError):
"""Post Processing exception.
This exception may be raised by PostProcessor's .run() method to
@@ -594,15 +819,16 @@ class PostProcessingError(Exception):
"""
def __init__(self, msg):
+ super(PostProcessingError, self).__init__(msg)
self.msg = msg
-class MaxDownloadsReached(Exception):
+class MaxDownloadsReached(YoutubeDLError):
""" --max-downloads limit has been reached. """
pass
-class UnavailableVideoError(Exception):
+class UnavailableVideoError(YoutubeDLError):
"""Unavailable Format exception.
This exception will be thrown when a video is requested
@@ -611,7 +837,7 @@ class UnavailableVideoError(Exception):
pass
-class ContentTooShortError(Exception):
+class ContentTooShortError(YoutubeDLError):
"""Content Too Short exception.
This exception may be raised by FileDownloader objects when a file they
@@ -620,26 +846,87 @@ class ContentTooShortError(Exception):
"""
def __init__(self, downloaded, expected):
+ super(ContentTooShortError, self).__init__(
+ 'Downloaded {0} bytes, expected {1} bytes'.format(downloaded, expected)
+ )
# Both in bytes
self.downloaded = downloaded
self.expected = expected
+class XAttrMetadataError(YoutubeDLError):
+ def __init__(self, code=None, msg='Unknown error'):
+ super(XAttrMetadataError, self).__init__(msg)
+ self.code = code
+ 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):
+ self.reason = 'NO_SPACE'
+ elif self.code == errno.E2BIG or 'Argument list too long' in self.msg:
+ self.reason = 'VALUE_TOO_LONG'
+ else:
+ self.reason = 'NOT_SUPPORTED'
+
+
+class XAttrUnavailableError(YoutubeDLError):
+ pass
+
+
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)
if sys.version_info < (3, 0):
- kwargs[b'strict'] = True
- hc = http_class(*args, **kwargs)
+ kwargs['strict'] = True
+ hc = http_class(*args, **compat_kwargs(kwargs))
source_address = ydl_handler._params.get('source_address')
+
if source_address is not None:
+ # This is to workaround _create_connection() from socket where it will try all
+ # address data from getaddrinfo() including IPv6. This filters the result from
+ # getaddrinfo() based on the source_address value.
+ # This is based on the cpython socket.create_connection() function.
+ # https://github.com/python/cpython/blob/master/Lib/socket.py#L691
+ def _create_connection(address, timeout=socket._GLOBAL_DEFAULT_TIMEOUT, source_address=None):
+ host, port = address
+ err = None
+ addrs = socket.getaddrinfo(host, port, 0, socket.SOCK_STREAM)
+ af = socket.AF_INET if '.' in source_address[0] else socket.AF_INET6
+ ip_addrs = [addr for addr in addrs if addr[0] == af]
+ if addrs and not ip_addrs:
+ ip_version = 'v4' if af == socket.AF_INET else 'v6'
+ raise socket.error(
+ "No remote IP%s addresses available for connect, can't use '%s' as source address"
+ % (ip_version, source_address[0]))
+ for res in ip_addrs:
+ af, socktype, proto, canonname, sa = res
+ sock = None
+ try:
+ sock = socket.socket(af, socktype, proto)
+ if timeout is not socket._GLOBAL_DEFAULT_TIMEOUT:
+ sock.settimeout(timeout)
+ sock.bind(source_address)
+ sock.connect(sa)
+ err = None # Explicitly break reference cycle
+ return sock
+ except socket.error as _:
+ err = _
+ if sock is not None:
+ sock.close()
+ if err is not None:
+ raise err
+ else:
+ raise socket.error('getaddrinfo returns an empty list')
+ if hasattr(hc, '_create_connection'):
+ hc._create_connection = _create_connection
sa = (source_address, 0)
if hasattr(hc, 'source_address'): # Python 2.7+
hc.source_address = sa
else: # Python 2.6
def _hc_connect(self, *args, **kwargs):
- sock = compat_socket_create_connection(
+ sock = _create_connection(
(self.host, self.port), self.timeout, sa)
if is_https:
self.sock = ssl.wrap_socket(
@@ -652,6 +939,16 @@ def _create_http_connection(ydl_handler, http_class, is_https, *args, **kwargs):
return hc
+def handle_youtubedl_headers(headers):
+ filtered_headers = headers
+
+ if 'Youtubedl-no-compression' in filtered_headers:
+ filtered_headers = dict((k, v) for k, v in filtered_headers.items() if k.lower() != 'accept-encoding')
+ del filtered_headers['Youtubedl-no-compression']
+
+ return filtered_headers
+
+
class YoutubeDLHandler(compat_urllib_request.HTTPHandler):
"""Handler for HTTP requests and responses.
@@ -659,7 +956,7 @@ class YoutubeDLHandler(compat_urllib_request.HTTPHandler):
the standard headers to every HTTP request and handles gzipped and
deflated responses from web servers. If compression is to be avoided in
a particular request, the original request in the program code only has
- to include the HTTP header "Youtubedl-No-Compression", which will be
+ to include the HTTP header "Youtubedl-no-compression", which will be
removed before making the real request.
Part of this code was copied from:
@@ -675,8 +972,15 @@ class YoutubeDLHandler(compat_urllib_request.HTTPHandler):
self._params = params
def http_open(self, req):
+ conn_class = compat_http_client.HTTPConnection
+
+ socks_proxy = req.headers.get('Ytdl-socks-proxy')
+ if socks_proxy:
+ conn_class = make_socks_conn_class(conn_class, socks_proxy)
+ del req.headers['Ytdl-socks-proxy']
+
return self.do_open(functools.partial(
- _create_http_connection, self, compat_http_client.HTTPConnection, False),
+ _create_http_connection, self, conn_class, False),
req)
@staticmethod
@@ -686,14 +990,6 @@ class YoutubeDLHandler(compat_urllib_request.HTTPHandler):
except zlib.error:
return zlib.decompress(data)
- @staticmethod
- def addinfourl_wrapper(stream, headers, url, code):
- if hasattr(compat_urllib_request.addinfourl, 'getcode'):
- return compat_urllib_request.addinfourl(stream, headers, url, code)
- ret = compat_urllib_request.addinfourl(stream, headers, url)
- ret.code = code
- return ret
-
def http_request(self, req):
# According to RFC 3986, URLs can not contain non-ASCII characters, however this is not
# always respected by websites, some tend to give out URLs with non percent-encoded
@@ -708,22 +1004,15 @@ class YoutubeDLHandler(compat_urllib_request.HTTPHandler):
# Substitute URL if any change after escaping
if url != url_escaped:
- req_type = HEADRequest if req.get_method() == 'HEAD' else compat_urllib_request.Request
- new_req = req_type(
- url_escaped, data=req.data, headers=req.headers,
- origin_req_host=req.origin_req_host, unverifiable=req.unverifiable)
- new_req.timeout = req.timeout
- req = new_req
+ req = update_Request(req, url=url_escaped)
for h, v in std_headers.items():
# Capitalize is needed because of Python bug 2275: http://bugs.python.org/issue2275
# The dict keys are capitalized because of this bug by urllib
if h.capitalize() not in req.headers:
req.add_header(h, v)
- if 'Youtubedl-no-compression' in req.headers:
- if 'Accept-encoding' in req.headers:
- del req.headers['Accept-encoding']
- del req.headers['Youtubedl-no-compression']
+
+ req.headers = handle_youtubedl_headers(req.headers)
if sys.version_info < (2, 7) and '#' in req.get_full_url():
# Python 2.6 is brain-dead when it comes to fragments
@@ -752,13 +1041,15 @@ class YoutubeDLHandler(compat_urllib_request.HTTPHandler):
break
else:
raise original_ioerror
- resp = self.addinfourl_wrapper(uncompressed, old_resp.headers, old_resp.url, old_resp.code)
+ resp = compat_urllib_request.addinfourl(uncompressed, old_resp.headers, old_resp.url, old_resp.code)
resp.msg = old_resp.msg
+ del resp.headers['Content-encoding']
# deflate
if resp.headers.get('Content-encoding', '') == 'deflate':
gz = io.BytesIO(self.deflate(resp.read()))
- resp = self.addinfourl_wrapper(gz, old_resp.headers, old_resp.url, old_resp.code)
+ resp = compat_urllib_request.addinfourl(gz, old_resp.headers, old_resp.url, old_resp.code)
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).
if 300 <= resp.code < 400:
@@ -767,9 +1058,13 @@ class YoutubeDLHandler(compat_urllib_request.HTTPHandler):
# As of RFC 2616 default charset is iso-8859-1 that is respected by python 3
if sys.version_info >= (3, 0):
location = location.encode('iso-8859-1').decode('utf-8')
+ else:
+ location = location.decode('utf-8')
location_escaped = escape_url(location)
if location != location_escaped:
del resp.headers['Location']
+ if sys.version_info < (3, 0):
+ location_escaped = location_escaped.encode('utf-8')
resp.headers['Location'] = location_escaped
return resp
@@ -777,6 +1072,49 @@ class YoutubeDLHandler(compat_urllib_request.HTTPHandler):
https_response = http_response
+def make_socks_conn_class(base_class, socks_proxy):
+ assert issubclass(base_class, (
+ compat_http_client.HTTPConnection, compat_http_client.HTTPSConnection))
+
+ url_components = compat_urlparse.urlparse(socks_proxy)
+ if url_components.scheme.lower() == 'socks5':
+ socks_type = ProxyType.SOCKS5
+ elif url_components.scheme.lower() in ('socks', 'socks4'):
+ socks_type = ProxyType.SOCKS4
+ elif url_components.scheme.lower() == 'socks4a':
+ socks_type = ProxyType.SOCKS4A
+
+ def unquote_if_non_empty(s):
+ if not s:
+ return s
+ return compat_urllib_parse_unquote_plus(s)
+
+ proxy_args = (
+ socks_type,
+ url_components.hostname, url_components.port or 1080,
+ True, # Remote DNS
+ unquote_if_non_empty(url_components.username),
+ unquote_if_non_empty(url_components.password),
+ )
+
+ class SocksConnection(base_class):
+ def connect(self):
+ self.sock = sockssocket()
+ self.sock.setproxy(*proxy_args)
+ if type(self.timeout) in (int, float):
+ self.sock.settimeout(self.timeout)
+ self.sock.connect((self.host, self.port))
+
+ if isinstance(self, compat_http_client.HTTPSConnection):
+ if hasattr(self, '_context'): # Python > 2.6
+ self.sock = self._context.wrap_socket(
+ self.sock, server_hostname=self.host)
+ else:
+ self.sock = ssl.wrap_socket(self.sock)
+
+ return SocksConnection
+
+
class YoutubeDLHTTPSHandler(compat_urllib_request.HTTPSHandler):
def __init__(self, params, https_conn_class=None, *args, **kwargs):
compat_urllib_request.HTTPSHandler.__init__(self, *args, **kwargs)
@@ -785,15 +1123,50 @@ class YoutubeDLHTTPSHandler(compat_urllib_request.HTTPSHandler):
def https_open(self, req):
kwargs = {}
+ conn_class = self._https_conn_class
+
if hasattr(self, '_context'): # python > 2.6
kwargs['context'] = self._context
if hasattr(self, '_check_hostname'): # python 3.x
kwargs['check_hostname'] = self._check_hostname
+
+ socks_proxy = req.headers.get('Ytdl-socks-proxy')
+ if socks_proxy:
+ conn_class = make_socks_conn_class(conn_class, socks_proxy)
+ del req.headers['Ytdl-socks-proxy']
+
return self.do_open(functools.partial(
- _create_http_connection, self, self._https_conn_class, True),
+ _create_http_connection, self, conn_class, True),
req, **kwargs)
+class YoutubeDLCookieJar(compat_cookiejar.MozillaCookieJar):
+ 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):
+ compat_cookiejar.MozillaCookieJar.load(self, 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)
@@ -818,6 +1191,24 @@ class YoutubeDLCookieProcessor(compat_urllib_request.HTTPCookieProcessor):
https_response = http_response
+def extract_timezone(date_str):
+ m = re.search(
+ r'^.{8,}?(?PZ$| ?(?P\+|-)(?P[0-9]{2}):?(?P[0-9]{2})$)',
+ date_str)
+ if not m:
+ timezone = datetime.timedelta()
+ else:
+ date_str = date_str[:-len(m.group('tz'))]
+ if not m.group('sign'):
+ timezone = datetime.timedelta()
+ else:
+ sign = 1 if m.group('sign') == '+' else -1
+ timezone = datetime.timedelta(
+ hours=sign * int(m.group('hours')),
+ minutes=sign * int(m.group('minutes')))
+ return timezone, date_str
+
+
def parse_iso8601(date_str, delimiter='T', timezone=None):
""" Return a UNIX timestamp from the given date """
@@ -827,20 +1218,8 @@ def parse_iso8601(date_str, delimiter='T', timezone=None):
date_str = re.sub(r'\.[0-9]+', '', date_str)
if timezone is None:
- m = re.search(
- r'(?:Z$| ?(?P\+|-)(?P[0-9]{2}):?(?P[0-9]{2})$)',
- date_str)
- if not m:
- timezone = datetime.timedelta()
- else:
- date_str = date_str[:-len(m.group(0))]
- if not m.group('sign'):
- timezone = datetime.timedelta()
- else:
- sign = 1 if m.group('sign') == '+' else -1
- timezone = datetime.timedelta(
- hours=sign * int(m.group('hours')),
- minutes=sign * int(m.group('minutes')))
+ timezone, date_str = extract_timezone(date_str)
+
try:
date_format = '%Y-%m-%d{0}%H:%M:%S'.format(delimiter)
dt = datetime.datetime.strptime(date_str, date_format) - timezone
@@ -849,6 +1228,10 @@ def parse_iso8601(date_str, delimiter='T', timezone=None):
pass
+def date_formats(day_first=True):
+ return DATE_FORMATS_DAY_FIRST if day_first else DATE_FORMATS_MONTH_FIRST
+
+
def unified_strdate(date_str, day_first=True):
"""Return a string with the date in the format YYYYMMDD"""
@@ -857,52 +1240,11 @@ def unified_strdate(date_str, day_first=True):
upload_date = None
# Replace commas
date_str = date_str.replace(',', ' ')
- # %z (UTC offset) is only supported in python>=3.2
- if not re.match(r'^[0-9]{1,2}-[0-9]{1,2}-[0-9]{4}$', date_str):
- date_str = re.sub(r' ?(\+|-)[0-9]{2}:?[0-9]{2}$', '', date_str)
# Remove AM/PM + timezone
date_str = re.sub(r'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str)
+ _, date_str = extract_timezone(date_str)
- format_expressions = [
- '%d %B %Y',
- '%d %b %Y',
- '%B %d %Y',
- '%b %d %Y',
- '%b %dst %Y %I:%M%p',
- '%b %dnd %Y %I:%M%p',
- '%b %dth %Y %I:%M%p',
- '%Y %m %d',
- '%Y-%m-%d',
- '%Y/%m/%d',
- '%Y/%m/%d %H:%M:%S',
- '%Y-%m-%d %H:%M:%S',
- '%Y-%m-%d %H:%M:%S.%f',
- '%d.%m.%Y %H:%M',
- '%d.%m.%Y %H.%M',
- '%Y-%m-%dT%H:%M:%SZ',
- '%Y-%m-%dT%H:%M:%S.%fZ',
- '%Y-%m-%dT%H:%M:%S.%f0Z',
- '%Y-%m-%dT%H:%M:%S',
- '%Y-%m-%dT%H:%M:%S.%f',
- '%Y-%m-%dT%H:%M',
- ]
- if day_first:
- format_expressions.extend([
- '%d-%m-%Y',
- '%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',
- '%m/%d/%Y %H:%M:%S',
- ])
- for expression in format_expressions:
+ for expression in date_formats(day_first):
try:
upload_date = datetime.datetime.strptime(date_str, expression).strftime('%Y%m%d')
except ValueError:
@@ -910,17 +1252,56 @@ def unified_strdate(date_str, day_first=True):
if upload_date is None:
timetuple = email.utils.parsedate_tz(date_str)
if timetuple:
- upload_date = datetime.datetime(*timetuple[:6]).strftime('%Y%m%d')
+ try:
+ upload_date = datetime.datetime(*timetuple[:6]).strftime('%Y%m%d')
+ except ValueError:
+ pass
if upload_date is not None:
return compat_str(upload_date)
+def unified_timestamp(date_str, day_first=True):
+ if date_str is None:
+ return None
+
+ date_str = re.sub(r'[,|]', '', date_str)
+
+ pm_delta = 12 if re.search(r'(?i)PM', date_str) else 0
+ timezone, date_str = extract_timezone(date_str)
+
+ # Remove AM/PM + timezone
+ date_str = re.sub(r'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str)
+
+ # Remove unrecognized timezones from ISO 8601 alike timestamps
+ m = re.search(r'\d{1,2}:\d{1,2}(?:\.\d+)?(?P\s*[A-Z]+)$', date_str)
+ if m:
+ date_str = date_str[:-len(m.group('tz'))]
+
+ # Python only supports microseconds, so remove nanoseconds
+ m = re.search(r'^([0-9]{4,}-[0-9]{1,2}-[0-9]{1,2}T[0-9]{1,2}:[0-9]{1,2}:[0-9]{1,2}\.[0-9]{6})[0-9]+$', date_str)
+ if m:
+ date_str = m.group(1)
+
+ for expression in date_formats(day_first):
+ try:
+ dt = datetime.datetime.strptime(date_str, expression) - timezone + datetime.timedelta(hours=pm_delta)
+ return calendar.timegm(dt.timetuple())
+ except ValueError:
+ pass
+ timetuple = email.utils.parsedate_tz(date_str)
+ if timetuple:
+ return calendar.timegm(timetuple) + pm_delta * 3600
+
+
def determine_ext(url, default_ext='unknown_video'):
- if url is None:
+ if url is None or '.' not in url:
return default_ext
guess = url.partition('?')[0].rpartition('.')[2]
if re.match(r'^[A-Za-z0-9]+$', guess):
return guess
+ # Try extract ext from URLs like http://example.com/foo/bar.mp4/?download
+ elif guess.rstrip('/') in KNOWN_EXTENSIONS:
+ return guess.rstrip('/')
else:
return default_ext
@@ -938,14 +1319,14 @@ def date_from_str(date_str):
return today
if date_str == 'yesterday':
return today - datetime.timedelta(days=1)
- match = re.match('(now|today)(?P[+-])(?P