New upstream version 2019.06.08

[youtubedl] / youtube_dl / utils.py

diff --git a/youtube_dl/utils.py b/youtube_dl/utils.py
index 52d198fa3c2eb36a1a3d41620cd645b90d52f854..ead9bd8628c88183e5ca0aaa7a0eb9b8fe03d836 100644 (file)
--- a/youtube_dl/utils.py
+++ b/youtube_dl/utils.py
@@ -1,30 +1,32 @@
 #!/usr/bin/env python
 #!/usr/bin/env python

-# -*- coding: utf-8 -*-
+# coding: utf-8

 
 from __future__ import unicode_literals
 
 
 from __future__ import unicode_literals
 

+import base64
+import binascii

 import calendar
 import codecs
 import contextlib
 import ctypes
 import datetime
 import email.utils
 import calendar
 import codecs
 import contextlib
 import ctypes
 import datetime
 import email.utils

+import email.header

 import errno
 import functools
 import gzip
 import errno
 import functools
 import gzip

-import itertools

 import io
 import io

+import itertools

 import json
 import locale
 import math
 import operator
 import os
 import json
 import locale
 import math
 import operator
 import os

-import pipes

 import platform
 import platform

+import random

 import re
 import re

-import ssl

 import socket
 import socket

-import struct
+import ssl

 import subprocess
 import sys
 import tempfile
 import subprocess
 import sys
 import tempfile


@@ -33,28 +35,54 @@ import xml.etree.ElementTree
 import zlib
 
 from .compat import (
 import zlib
 
 from .compat import (

+    compat_HTMLParseError,
+    compat_HTMLParser,

     compat_basestring,
     compat_chr,
     compat_basestring,
     compat_chr,

+    compat_cookiejar,
+    compat_ctypes_WINFUNCTYPE,
+    compat_etree_fromstring,
+    compat_expanduser,

     compat_html_entities,
     compat_html_entities,

+    compat_html_entities_html5,

     compat_http_client,
     compat_kwargs,
     compat_http_client,
     compat_kwargs,

+    compat_os_name,

     compat_parse_qs,
     compat_parse_qs,

-    compat_socket_create_connection,
+    compat_shlex_quote,

     compat_str,
     compat_str,

+    compat_struct_pack,
+    compat_struct_unpack,

     compat_urllib_error,
     compat_urllib_parse,
     compat_urllib_error,
     compat_urllib_parse,

+    compat_urllib_parse_urlencode,

     compat_urllib_parse_urlparse,
     compat_urllib_parse_urlparse,

+    compat_urllib_parse_unquote_plus,

     compat_urllib_request,
     compat_urlparse,
     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 = {
 
 # 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',
     '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',


@@ -62,10 +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']
 
 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('ÂÃÄÀÁÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖŐØŒÙÚÛÜŰÝÞßàáâãäåæçèéêëìíîïðñòóôõöőøœùúûüűýþÿ',
+                        itertools.chain('AAAAAA', ['AE'], 'CEEEEIIIIDNOOOOOOO', ['OE'], 'UUUUUY', ['TH', 'ss'],
+                                        'aaaaaa', ['ae'], 'ceeeeiiiionooooooo', ['oe'], 'uuuuuy', ['th'], 'y')))
+
+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)<script[^>]+type=(["\']?)application/ld\+json\1[^>]*>(?P<json_ld>.+?)</script>'
+

 
 def preferredencoding():
     """Get preferred encoding.
 
 def preferredencoding():
     """Get preferred encoding.


@@ -137,21 +257,17 @@ def write_json_file(obj, fn):
 
 
 if sys.version_info >= (2, 7):
 
 
 if sys.version_info >= (2, 7):

-    def find_xpath_attr(node, xpath, key, val):
+    def find_xpath_attr(node, xpath, key, val=None):

         """ Find the xpath xpath[@key=val] """
         """ Find the xpath xpath[@key=val] """

-        assert re.match(r'^[a-zA-Z-]+$', key)
-        assert re.match(r'^[a-zA-Z0-9@\s:._-]*$', val)
-        expr = xpath + "[@%s='%s']" % (key, val)
+        assert re.match(r'^[a-zA-Z_-]+$', key)
+        expr = xpath + ('[@%s]' % key if val is None else "[@%s='%s']" % (key, val))

         return node.find(expr)
 else:
         return node.find(expr)
 else:

-    def find_xpath_attr(node, xpath, key, val):
-        # 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):
-            if f.attrib.get(key) == val:
+    def find_xpath_attr(node, xpath, key, val=None):
+        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:

                 return f
         return None
 
                 return f
         return None
 


@@ -171,46 +287,139 @@ def xpath_with_ns(path, ns_map):
     return '/'.join(replaced)
 
 
     return '/'.join(replaced)
 
 

-def xpath_text(node, xpath, name=None, fatal=False):
-    if sys.version_info < (2, 7):  # Crazy 2.6
-        xpath = xpath.encode('ascii')
+def xpath_element(node, xpath, name=None, fatal=False, default=NO_DEFAULT):
+    def _find_xpath(xpath):
+        return node.find(compat_xpath(xpath))
+
+    if isinstance(xpath, (str, compat_str)):
+        n = _find_xpath(xpath)
+    else:
+        for xp in xpath:
+            n = _find_xpath(xp)
+            if n is not None:
+                break

 
 

-    n = node.find(xpath)
-    if n is None or n.text is None:
-        if fatal:
+    if n is None:
+        if default is not NO_DEFAULT:
+            return default
+        elif fatal:

             name = xpath if name is None else name
             raise ExtractorError('Could not find XML element %s' % name)
         else:
             return None
             name = xpath if name is None else name
             raise ExtractorError('Could not find XML element %s' % name)
         else:
             return None

+    return n
+
+
+def xpath_text(node, xpath, name=None, fatal=False, default=NO_DEFAULT):
+    n = xpath_element(node, xpath, name, fatal=fatal, default=default)
+    if n is None or n == default:
+        return n
+    if n.text is None:
+        if default is not NO_DEFAULT:
+            return default
+        elif fatal:
+            name = xpath if name is None else name
+            raise ExtractorError('Could not find XML element\'s text %s' % name)
+        else:
+            return None

     return n.text
 
 
     return n.text
 
 

+def xpath_attr(node, xpath, key, name=None, fatal=False, default=NO_DEFAULT):
+    n = find_xpath_attr(node, xpath, key)
+    if n is None:
+        if default is not NO_DEFAULT:
+            return default
+        elif fatal:
+            name = '%s[@%s]' % (xpath, key) if name is None else name
+            raise ExtractorError('Could not find XML attribute %s' % name)
+        else:
+            return None
+    return n.attrib[key]
+
+

 def get_element_by_id(id, html):
     """Return the content of the tag with the specified ID in the passed HTML document"""
 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):
+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_elements_by_attribute(attribute, value, html, escape_value=True):

     """Return the content of the tag with the specified attribute in the passed HTML document"""
 
     """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:._-]+)
         <([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+%s=['"]?%s['"]?

-         (?:\s+[a-zA-Z0-9:._-]+(?:=[a-zA-Z0-9:._-]+|="[^"]+"|='[^']+'))*?
+         (?:\s+[a-zA-Z0-9:._-]+(?:=[a-zA-Z0-9:._-]*|="[^"]*"|='[^']*'|))*?

         \s*>
         (?P<content>.*?)
         </\1>
         \s*>
         (?P<content>.*?)
         </\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
+    <el
+         a="foo" B="bar" c="&98;az" d=boz
+         empty= noval entity="&amp;"
+         sq='"' dq="'"
+    >
+    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):
 
 
 def clean_html(html):


@@ -221,8 +430,8 @@ def clean_html(html):
 
     # Newline vs <br />
     html = html.replace('\n', ' ')
 
     # Newline vs <br />
     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
     # Strip html tags
     html = re.sub('<.*?>', '', html)
     # Replace html entities


@@ -274,9 +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.
 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):
     """
     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 == '"':
         if char == '?' or ord(char) < 32 or ord(char) == 127:
             return ''
         elif char == '"':


@@ -320,13 +532,40 @@ def sanitize_path(s):
     if drive_or_unc:
         norm_path.pop(0)
     sanitized_path = [
     if drive_or_unc:
         norm_path.pop(0)
     sanitized_path = [

-        path_part if path_part in ['.', '..'] else re.sub('(?:[/<>:"\\|\\\\?\\*]|\.$)', '#', path_part)
+        path_part if path_part in ['.', '..'] else re.sub(r'(?:[/<>:"\|\\?\*]|[\s.]$)', '#', 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)
 
 
         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/ytdl-org/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 = []
 def orderedSet(iterable):
     """ Remove all duplicates from the input iterable """
     res = []


@@ -336,12 +575,19 @@ def orderedSet(iterable):
     return res
 
 
     return res
 
 

-def _htmlentity_transform(entity):
+def _htmlentity_transform(entity_with_semicolon):

     """Transforms an HTML entity to a character."""
     """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])
 
     # 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,
+    # '&Eacuteric' should be decoded as 'Éric'.
+    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)
     mobj = re.match(r'#(x[0-9a-fA-F]+|[0-9]+)', entity)
     if mobj is not None:
         numstr = mobj.group(1)


@@ -350,10 +596,14 @@ def _htmlentity_transform(entity):
             numstr = '0%s' % numstr
         else:
             base = 10
             numstr = '0%s' % numstr
         else:
             base = 10

-        return compat_chr(int(numstr, base))
+        # See https://github.com/ytdl-org/youtube-dl/issues/7518
+        try:
+            return compat_chr(int(numstr, base))
+        except ValueError:
+            pass

 
     # Unknown entity in name, return its literal representation
 
     # Unknown entity in name, return its literal representation

-    return ('&%s;' % entity)
+    return '&%s;' % entity

 
 
 def unescapeHTML(s):
 
 
 def unescapeHTML(s):


@@ -362,7 +612,7 @@ def unescapeHTML(s):
     assert type(s) == compat_str
 
     return re.sub(
     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():
 
 
 def get_subprocess_encoding():


@@ -394,6 +644,10 @@ def encodeFilename(s, for_subprocess=False):
     if not for_subprocess and sys.platform == 'win32' and sys.getwindowsversion()[0] >= 5:
         return s
 
     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')
 
 
     return s.encode(get_subprocess_encoding(), 'ignore')
 
 


@@ -476,7 +730,12 @@ def bug_reports_message():
     return msg
 
 
     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):
     """Error during info extraction."""
 
     def __init__(self, msg, tb=None, expected=False, cause=None, video_id=None):


@@ -517,7 +776,19 @@ class RegexNotFoundError(ExtractorError):
     pass
 
 
     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
     """Download Error exception.
 
     This exception may be thrown by FileDownloader objects if they are not


@@ -531,7 +802,7 @@ class DownloadError(Exception):
         self.exc_info = exc_info
 
 
         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
     """Same File exception.
 
     This exception will be thrown by FileDownloader objects if they detect


@@ -540,7 +811,7 @@ class SameFileError(Exception):
     pass
 
 
     pass
 
 

-class PostProcessingError(Exception):
+class PostProcessingError(YoutubeDLError):

     """Post Processing exception.
 
     This exception may be raised by PostProcessor's .run() method to
     """Post Processing exception.
 
     This exception may be raised by PostProcessor's .run() method to


@@ -548,15 +819,16 @@ class PostProcessingError(Exception):
     """
 
     def __init__(self, msg):
     """
 
     def __init__(self, msg):

+        super(PostProcessingError, self).__init__(msg)

         self.msg = msg
 
 
         self.msg = msg
 
 

-class MaxDownloadsReached(Exception):
+class MaxDownloadsReached(YoutubeDLError):

     """ --max-downloads limit has been reached. """
     pass
 
 
     """ --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
     """Unavailable Format exception.
 
     This exception will be thrown when a video is requested


@@ -565,32 +837,96 @@ class UnavailableVideoError(Exception):
     pass
 
 
     pass
 
 

-class ContentTooShortError(Exception):
+class ContentTooShortError(YoutubeDLError):

     """Content Too Short exception.
 
     This exception may be raised by FileDownloader objects when a file they
     download is too small for what the server announced first, indicating
     the connection was probably interrupted.
     """
     """Content Too Short exception.
 
     This exception may be raised by FileDownloader objects when a file they
     download is too small for what the server announced first, indicating
     the connection was probably interrupted.
     """

-    # Both in bytes
-    downloaded = None
-    expected = None

 
     def __init__(self, downloaded, expected):
 
     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
 
 
         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):
 def _create_http_connection(ydl_handler, http_class, is_https, *args, **kwargs):

-    hc = http_class(*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/ytdl-org/youtube-dl/issues/6727)
+    if sys.version_info < (3, 0):
+        kwargs['strict'] = True
+    hc = http_class(*args, **compat_kwargs(kwargs))

     source_address = ydl_handler._params.get('source_address')
     source_address = ydl_handler._params.get('source_address')

+

     if source_address is not None:
     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):
         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(
                     (self.host, self.port), self.timeout, sa)
                 if is_https:
                     self.sock = ssl.wrap_socket(


@@ -603,6 +939,16 @@ def _create_http_connection(ydl_handler, http_class, is_https, *args, **kwargs):
     return hc
 
 
     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.
 
 class YoutubeDLHandler(compat_urllib_request.HTTPHandler):
     """Handler for HTTP requests and responses.
 


@@ -610,7 +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
     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:
     removed before making the real request.
 
     Part of this code was copied from:


@@ -626,8 +972,15 @@ class YoutubeDLHandler(compat_urllib_request.HTTPHandler):
         self._params = params
 
     def http_open(self, req):
         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(
         return self.do_open(functools.partial(

-            _create_http_connection, self, compat_http_client.HTTPConnection, False),
+            _create_http_connection, self, conn_class, False),

             req)
 
     @staticmethod
             req)
 
     @staticmethod


@@ -637,24 +990,29 @@ class YoutubeDLHandler(compat_urllib_request.HTTPHandler):
         except zlib.error:
             return zlib.decompress(data)
 
         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):
     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
+        # non-ASCII characters (see telemb.py, ard.py [#3412])
+        # urllib chokes on URLs with non-ASCII characters (see http://bugs.python.org/issue3991)
+        # To work around aforementioned issue we will replace request's original URL with
+        # percent-encoded one
+        # Since redirects are also affected (e.g. http://www.southpark.de/alle-episoden/s18e09)
+        # the code of this workaround has been moved here from YoutubeDL.urlopen()
+        url = req.get_full_url()
+        url_escaped = escape_url(url)
+
+        # Substitute URL if any change after escaping
+        if url != url_escaped:
+            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)
         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
 
         if sys.version_info < (2, 7) and '#' in req.get_full_url():
             # Python 2.6 is brain-dead when it comes to fragments


@@ -683,19 +1041,80 @@ class YoutubeDLHandler(compat_urllib_request.HTTPHandler):
                     break
                 else:
                     raise original_ioerror
                     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
             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()))
         # 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
             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/ytdl-org/youtube-dl/issues/6457).
+        if 300 <= resp.code < 400:
+            location = resp.headers.get('Location')
+            if location:
+                # 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
 
     https_request = http_request
     https_response = http_response
 
 
         return resp
 
     https_request = http_request
     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)
 class YoutubeDLHTTPSHandler(compat_urllib_request.HTTPSHandler):
     def __init__(self, params, https_conn_class=None, *args, **kwargs):
         compat_urllib_request.HTTPSHandler.__init__(self, *args, **kwargs)


@@ -704,39 +1123,129 @@ class YoutubeDLHTTPSHandler(compat_urllib_request.HTTPSHandler):
 
     def https_open(self, req):
         kwargs = {}
 
     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
         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(
         return self.do_open(functools.partial(

-            _create_http_connection, self, self._https_conn_class, True),
+            _create_http_connection, self, conn_class, True),

             req, **kwargs)
 
 
             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)
+
+    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/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:
+        #     for set_cookie_header in ('Set-Cookie', 'Set-Cookie2'):
+        #         set_cookie = response.headers.get(set_cookie_header)
+        #         if set_cookie:
+        #             set_cookie_escaped = compat_urllib_parse.quote(set_cookie, b"%/;:@&=+$,!~*'()?#[] ")
+        #             if set_cookie != set_cookie_escaped:
+        #                 del response.headers[set_cookie_header]
+        #                 response.headers[set_cookie_header] = set_cookie_escaped
+        return compat_urllib_request.HTTPCookieProcessor.http_response(self, request, response)
+
+    https_request = compat_urllib_request.HTTPCookieProcessor.http_request
+    https_response = http_response
+
+
+def extract_timezone(date_str):
+    m = re.search(
+        r'^.{8,}?(?P<tz>Z$| ?(?P<sign>\+|-)(?P<hours>[0-9]{2}):?(?P<minutes>[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 """
 
     if date_str is None:
         return None
 
 def parse_iso8601(date_str, delimiter='T', timezone=None):
     """ Return a UNIX timestamp from the given date """
 
     if date_str is None:
         return None
 

+    date_str = re.sub(r'\.[0-9]+', '', date_str)
+

     if timezone is None:
     if timezone is None:

-        m = re.search(
-            r'(\.[0-9]+)?(?:Z$| ?(?P<sign>\+|-)(?P<hours>[0-9]{2}):?(?P<minutes>[0-9]{2})$)',
-            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')))
-    date_format = '%Y-%m-%d{0}%H:%M:%S'.format(delimiter)
-    dt = datetime.datetime.strptime(date_str, date_format) - timezone
-    return calendar.timegm(dt.timetuple())
+        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
+        return calendar.timegm(dt.timetuple())
+    except ValueError:
+        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):
 
 
 def unified_strdate(date_str, day_first=True):


@@ -747,52 +1256,11 @@ def unified_strdate(date_str, day_first=True):
     upload_date = None
     # Replace commas
     date_str = date_str.replace(',', ' ')
     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)
     # 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:
         try:
             upload_date = datetime.datetime.strptime(date_str, expression).strftime('%Y%m%d')
         except ValueError:


@@ -800,16 +1268,56 @@ def unified_strdate(date_str, day_first=True):
     if upload_date is None:
         timetuple = email.utils.parsedate_tz(date_str)
         if timetuple:
     if upload_date is None:
         timetuple = email.utils.parsedate_tz(date_str)
         if timetuple:

-            upload_date = datetime.datetime(*timetuple[:6]).strftime('%Y%m%d')
-    return upload_date
+            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<tz>\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'):
 
 
 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]+$', guess):
         return guess
         return default_ext
     guess = url.partition('?')[0].rpartition('.')[2]
     if re.match(r'^[A-Za-z0-9]+$', 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
 
     else:
         return default_ext
 


@@ -827,14 +1335,14 @@ def date_from_str(date_str):
         return today
     if date_str == 'yesterday':
         return today - datetime.timedelta(days=1)
         return today
     if date_str == 'yesterday':
         return today - datetime.timedelta(days=1)

-    match = re.match('(now|today)(?P<sign>[+-])(?P<time>\d+)(?P<unit>day|week|month|year)(s)?', date_str)
+    match = re.match(r'(now|today)(?P<sign>[+-])(?P<time>\d+)(?P<unit>day|week|month|year)(s)?', date_str)

     if match is not None:
         sign = match.group('sign')
         time = int(match.group('time'))
         if sign == '-':
             time = -time
         unit = match.group('unit')
     if match is not None:
         sign = match.group('sign')
         time = int(match.group('time'))
         if sign == '-':
             time = -time
         unit = match.group('unit')

-        # A bad aproximation?
+        # A bad approximation?

         if unit == 'month':
             unit = 'day'
             time *= 30
         if unit == 'month':
             unit = 'day'
             time *= 30


@@ -844,7 +1352,7 @@ def date_from_str(date_str):
         unit += 's'
         delta = datetime.timedelta(**{unit: time})
         return today + delta
         unit += 's'
         delta = datetime.timedelta(**{unit: time})
         return today + delta

-    return datetime.datetime.strptime(date_str, "%Y%m%d").date()
+    return datetime.datetime.strptime(date_str, '%Y%m%d').date()

 
 
 def hyphenate_date(date_str):
 
 
 def hyphenate_date(date_str):


@@ -922,31 +1430,31 @@ def _windows_write_string(s, out):
     if fileno not in WIN_OUTPUT_IDS:
         return False
 
     if fileno not in WIN_OUTPUT_IDS:
         return False
 

-    GetStdHandle = ctypes.WINFUNCTYPE(
+    GetStdHandle = compat_ctypes_WINFUNCTYPE(

         ctypes.wintypes.HANDLE, ctypes.wintypes.DWORD)(
         ctypes.wintypes.HANDLE, ctypes.wintypes.DWORD)(

-        (b"GetStdHandle", ctypes.windll.kernel32))
+        ('GetStdHandle', ctypes.windll.kernel32))

     h = GetStdHandle(WIN_OUTPUT_IDS[fileno])
 
     h = GetStdHandle(WIN_OUTPUT_IDS[fileno])
 

-    WriteConsoleW = ctypes.WINFUNCTYPE(
+    WriteConsoleW = compat_ctypes_WINFUNCTYPE(

         ctypes.wintypes.BOOL, ctypes.wintypes.HANDLE, ctypes.wintypes.LPWSTR,
         ctypes.wintypes.DWORD, ctypes.POINTER(ctypes.wintypes.DWORD),
         ctypes.wintypes.BOOL, ctypes.wintypes.HANDLE, ctypes.wintypes.LPWSTR,
         ctypes.wintypes.DWORD, ctypes.POINTER(ctypes.wintypes.DWORD),

-        ctypes.wintypes.LPVOID)((b"WriteConsoleW", ctypes.windll.kernel32))
+        ctypes.wintypes.LPVOID)(('WriteConsoleW', ctypes.windll.kernel32))

     written = ctypes.wintypes.DWORD(0)
 
     written = ctypes.wintypes.DWORD(0)
 

-    GetFileType = ctypes.WINFUNCTYPE(ctypes.wintypes.DWORD, ctypes.wintypes.DWORD)((b"GetFileType", ctypes.windll.kernel32))
+    GetFileType = compat_ctypes_WINFUNCTYPE(ctypes.wintypes.DWORD, ctypes.wintypes.DWORD)(('GetFileType', ctypes.windll.kernel32))

     FILE_TYPE_CHAR = 0x0002
     FILE_TYPE_REMOTE = 0x8000
     FILE_TYPE_CHAR = 0x0002
     FILE_TYPE_REMOTE = 0x8000

-    GetConsoleMode = ctypes.WINFUNCTYPE(
+    GetConsoleMode = compat_ctypes_WINFUNCTYPE(

         ctypes.wintypes.BOOL, ctypes.wintypes.HANDLE,
         ctypes.POINTER(ctypes.wintypes.DWORD))(
         ctypes.wintypes.BOOL, ctypes.wintypes.HANDLE,
         ctypes.POINTER(ctypes.wintypes.DWORD))(

-        (b"GetConsoleMode", ctypes.windll.kernel32))
+        ('GetConsoleMode', ctypes.windll.kernel32))

     INVALID_HANDLE_VALUE = ctypes.wintypes.DWORD(-1).value
 
     def not_a_console(handle):
         if handle == INVALID_HANDLE_VALUE or handle is None:
             return True
     INVALID_HANDLE_VALUE = ctypes.wintypes.DWORD(-1).value
 
     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
 
     if not_a_console(h):
         return False


@@ -982,8 +1490,8 @@ def write_string(s, out=None, encoding=None):
         if _windows_write_string(s, out):
             return
 
         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'):
         byt = s.encode(encoding or preferredencoding(), 'ignore')
         out.write(byt)
     elif hasattr(out, 'buffer'):


@@ -1007,7 +1515,7 @@ def bytes_to_intlist(bs):
 def intlist_to_bytes(xs):
     if not xs:
         return b''
 def intlist_to_bytes(xs):
     if not xs:
         return b''

-    return struct_pack('%dB' % len(xs), *xs)
+    return compat_struct_pack('%dB' % len(xs), *xs)

 
 
 # Cross-platform file locking
 
 
 # Cross-platform file locking


@@ -1066,13 +1574,23 @@ if sys.platform == 'win32':
             raise OSError('Unlocking file failed: %r' % ctypes.FormatError())
 
 else:
             raise OSError('Unlocking file failed: %r' % ctypes.FormatError())
 
 else:

-    import fcntl
+    # Some platforms, such as Jython, is missing fcntl
+    try:
+        import fcntl

 
 

-    def _lock_file(f, exclusive):
-        fcntl.flock(f, fcntl.LOCK_EX if exclusive else fcntl.LOCK_SH)
+        def _lock_file(f, exclusive):
+            fcntl.flock(f, fcntl.LOCK_EX if exclusive else fcntl.LOCK_SH)

 
 

-    def _unlock_file(f):
-        fcntl.flock(f, fcntl.LOCK_UN)
+        def _unlock_file(f):
+            fcntl.flock(f, fcntl.LOCK_UN)
+    except ImportError:
+        UNSUPPORTED_MSG = 'file locking is not supported on this platform'
+
+        def _lock_file(f, exclusive):
+            raise IOError(UNSUPPORTED_MSG)
+
+        def _unlock_file(f):
+            raise IOError(UNSUPPORTED_MSG)

 
 
 class locked_file(object):
 
 
 class locked_file(object):


@@ -1118,14 +1636,16 @@ def shell_quote(args):
         if isinstance(a, bytes):
             # We may get a filename encoded with 'encodeFilename'
             a = a.decode(encoding)
         if isinstance(a, bytes):
             # We may get a filename encoded with 'encodeFilename'
             a = a.decode(encoding)

-        quoted_args.append(pipes.quote(a))
+        quoted_args.append(compat_shlex_quote(a))

     return ' '.join(quoted_args)
 
 
 def smuggle_url(url, data):
     """ Pass additional data in a URL for internal use. """
 
     return ' '.join(quoted_args)
 
 
 def smuggle_url(url, data):
     """ Pass additional data in a URL for internal use. """
 

-    sdata = compat_urllib_parse.urlencode(
+    url, idata = unsmuggle_url(url, {})
+    data.update(idata)
+    sdata = compat_urllib_parse_urlencode(

         {'__youtubedl_smuggle': json.dumps(data)})
     return url + '#' + sdata
 
         {'__youtubedl_smuggle': json.dumps(data)})
     return url + '#' + sdata
 


@@ -1153,65 +1673,146 @@ def format_bytes(bytes):
     return '%.2f%s' % (converted, suffix)
 
 
     return '%.2f%s' % (converted, suffix)
 
 

+def lookup_unit_table(unit_table, s):
+    units_re = '|'.join(re.escape(u) for u in unit_table)
+    m = re.match(
+        r'(?P<num>[0-9]+(?:[,.][0-9]*)?)\s*(?P<unit>%s)\b' % units_re, s)
+    if not m:
+        return None
+    num_str = m.group('num').replace(',', '.')
+    mult = unit_table[m.group('unit')]
+    return int(float(num_str) * mult)
+
+

 def parse_filesize(s):
     if s is None:
         return None
 
 def parse_filesize(s):
     if s is None:
         return None
 

-    # The lower-case forms are of course incorrect and inofficial,
+    # The lower-case forms are of course incorrect and unofficial,

     # but we support those too
     _UNIT_TABLE = {
         'B': 1,
         'b': 1,
     # but we support those too
     _UNIT_TABLE = {
         'B': 1,
         'b': 1,

+        'bytes': 1,

         'KiB': 1024,
         'KB': 1000,
         'kB': 1024,
         'Kb': 1000,
         'KiB': 1024,
         'KB': 1000,
         'kB': 1024,
         'Kb': 1000,

+        'kb': 1000,
+        'kilobytes': 1000,
+        'kibibytes': 1024,

         'MiB': 1024 ** 2,
         'MB': 1000 ** 2,
         'mB': 1024 ** 2,
         'Mb': 1000 ** 2,
         'MiB': 1024 ** 2,
         'MB': 1000 ** 2,
         'mB': 1024 ** 2,
         'Mb': 1000 ** 2,

+        'mb': 1000 ** 2,
+        'megabytes': 1000 ** 2,
+        'mebibytes': 1024 ** 2,

         'GiB': 1024 ** 3,
         'GB': 1000 ** 3,
         'gB': 1024 ** 3,
         'Gb': 1000 ** 3,
         'GiB': 1024 ** 3,
         'GB': 1000 ** 3,
         'gB': 1024 ** 3,
         'Gb': 1000 ** 3,

+        'gb': 1000 ** 3,
+        'gigabytes': 1000 ** 3,
+        'gibibytes': 1024 ** 3,

         'TiB': 1024 ** 4,
         'TB': 1000 ** 4,
         'tB': 1024 ** 4,
         'Tb': 1000 ** 4,
         'TiB': 1024 ** 4,
         'TB': 1000 ** 4,
         'tB': 1024 ** 4,
         'Tb': 1000 ** 4,

+        'tb': 1000 ** 4,
+        'terabytes': 1000 ** 4,
+        'tebibytes': 1024 ** 4,

         'PiB': 1024 ** 5,
         'PB': 1000 ** 5,
         'pB': 1024 ** 5,
         'Pb': 1000 ** 5,
         'PiB': 1024 ** 5,
         'PB': 1000 ** 5,
         'pB': 1024 ** 5,
         'Pb': 1000 ** 5,

+        'pb': 1000 ** 5,
+        'petabytes': 1000 ** 5,
+        'pebibytes': 1024 ** 5,

         'EiB': 1024 ** 6,
         'EB': 1000 ** 6,
         'eB': 1024 ** 6,
         'Eb': 1000 ** 6,
         'EiB': 1024 ** 6,
         'EB': 1000 ** 6,
         'eB': 1024 ** 6,
         'Eb': 1000 ** 6,

+        'eb': 1000 ** 6,
+        'exabytes': 1000 ** 6,
+        'exbibytes': 1024 ** 6,

         'ZiB': 1024 ** 7,
         'ZB': 1000 ** 7,
         'zB': 1024 ** 7,
         'Zb': 1000 ** 7,
         'ZiB': 1024 ** 7,
         'ZB': 1000 ** 7,
         'zB': 1024 ** 7,
         'Zb': 1000 ** 7,

+        'zb': 1000 ** 7,
+        'zettabytes': 1000 ** 7,
+        'zebibytes': 1024 ** 7,

         'YiB': 1024 ** 8,
         'YB': 1000 ** 8,
         'yB': 1024 ** 8,
         'Yb': 1000 ** 8,
         'YiB': 1024 ** 8,
         'YB': 1000 ** 8,
         'yB': 1024 ** 8,
         'Yb': 1000 ** 8,

+        'yb': 1000 ** 8,
+        'yottabytes': 1000 ** 8,
+        'yobibytes': 1024 ** 8,

     }
 
     }
 

-    units_re = '|'.join(re.escape(u) for u in _UNIT_TABLE)
-    m = re.match(
-        r'(?P<num>[0-9]+(?:[,.][0-9]*)?)\s*(?P<unit>%s)' % units_re, s)
-    if not m:
+    return lookup_unit_table(_UNIT_TABLE, s)
+
+
+def parse_count(s):
+    if s is None:

         return None
 
         return None
 

-    num_str = m.group('num').replace(',', '.')
-    mult = _UNIT_TABLE[m.group('unit')]
-    return int(float(num_str) * mult)
+    s = s.strip()
+
+    if re.match(r'^[\d,.]+$', s):
+        return str_to_int(s)
+
+    _UNIT_TABLE = {
+        'k': 1000,
+        'K': 1000,
+        'm': 1000 ** 2,
+        'M': 1000 ** 2,
+        'kk': 1000 ** 2,
+        'KK': 1000 ** 2,
+    }
+
+    return lookup_unit_table(_UNIT_TABLE, s)
+
+
+def parse_resolution(s):
+    if s is None:
+        return {}
+
+    mobj = re.search(r'\b(?P<w>\d+)\s*[xX×]\s*(?P<h>\d+)\b', s)
+    if mobj:
+        return {
+            'width': int(mobj.group('w')),
+            'height': int(mobj.group('h')),
+        }
+
+    mobj = re.search(r'\b(\d+)[pPiI]\b', s)
+    if mobj:
+        return {'height': int(mobj.group(1))}
+
+    mobj = re.search(r'\b([48])[kK]\b', s)
+    if mobj:
+        return {'height': int(mobj.group(1)) * 540}
+
+    return {}

 
 
 
 

-def month_by_name(name):
+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 """
 
     """ Return the number of a month by (locale-independently) English name """
 

+    month_names = MONTH_NAMES.get(lang, MONTH_NAMES['en'])
+

     try:
     try:

-        return ENGLISH_MONTH_NAMES.index(name) + 1
+        return month_names.index(name) + 1

     except ValueError:
         return None
 
     except ValueError:
         return None
 


@@ -1236,10 +1837,21 @@ def fix_xml_ampersands(xml_str):
 
 def setproctitle(title):
     assert isinstance(title, compat_str)
 
 def setproctitle(title):
     assert isinstance(title, compat_str)

+
+    # ctypes in Jython is not complete
+    # http://bugs.jython.org/issue2148
+    if sys.platform.startswith('java'):
+        return
+

     try:
     try:

-        libc = ctypes.cdll.LoadLibrary("libc.so.6")
+        libc = ctypes.cdll.LoadLibrary('libc.so.6')

     except OSError:
         return
     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('utf-8')
     buf = ctypes.create_string_buffer(len(title_bytes))
     buf.value = title_bytes
     title_bytes = title.encode('utf-8')
     buf = ctypes.create_string_buffer(len(title_bytes))
     buf.value = title_bytes


@@ -1250,14 +1862,19 @@ def setproctitle(title):
 
 
 def remove_start(s, start):
 
 
 def remove_start(s, start):

-    if s.startswith(start):
-        return s[len(start):]
-    return s
+    return s[len(start):] if s is not None and s.startswith(start) else s

 
 
 def remove_end(s, end):
 
 
 def remove_end(s, end):

-    if s.endswith(end):
-        return s[:-len(end)]
+    return s[:-len(end)] if s is not None and s.endswith(end) else s
+
+
+def remove_quotes(s):
+    if s is None or len(s) < 2:
+        return s
+    for quote in ('"', "'", ):
+        if s[0] == quote and s[-1] == quote:
+            return s[1:-1]

     return s
 
 
     return s
 
 


@@ -1266,9 +1883,33 @@ def url_basename(url):
     return path.strip('/').split('/')[-1]
 
 
     return path.strip('/').split('/')[-1]
 
 

+def base_url(url):
+    return re.match(r'https?://[^?#&]+/', url).group()
+
+
+def urljoin(base, path):
+    if isinstance(path, bytes):
+        path = path.decode('utf-8')
+    if not isinstance(path, compat_str) or not path:
+        return None
+    if re.match(r'^(?:[a-zA-Z][a-zA-Z0-9+-.]*:)?//', path):
+        return path
+    if isinstance(base, bytes):
+        base = base.decode('utf-8')
+    if not isinstance(base, compat_str) or not re.match(
+            r'^(?:https?:)?//', base):
+        return None
+    return compat_urlparse.urljoin(base, path)
+
+

 class HEADRequest(compat_urllib_request.Request):
     def get_method(self):
 class HEADRequest(compat_urllib_request.Request):
     def get_method(self):

-        return "HEAD"
+        return 'HEAD'
+
+
+class PUTRequest(compat_urllib_request.Request):
+    def get_method(self):
+        return 'PUT'

 
 
 def int_or_none(v, scale=1, default=None, get_attr=None, invscale=1):
 
 
 def int_or_none(v, scale=1, default=None, get_attr=None, invscale=1):


@@ -1277,7 +1918,12 @@ def int_or_none(v, scale=1, default=None, get_attr=None, invscale=1):
             v = getattr(v, get_attr, None)
     if v == '':
         v = None
             v = getattr(v, get_attr, None)
     if v == '':
         v = None

-    return default if v is None else (int(v) * invscale // scale)
+    if v is None:
+        return default
+    try:
+        return int(v) * invscale // scale
+    except (ValueError, TypeError):
+        return default

 
 
 def str_or_none(v, default=None):
 
 
 def str_or_none(v, default=None):


@@ -1293,7 +1939,27 @@ def str_to_int(int_str):
 
 
 def float_or_none(v, scale=1, invscale=1, default=None):
 
 
 def float_or_none(v, scale=1, invscale=1, default=None):

-    return default if v is None else (float(v) * invscale / scale)
+    if v is None:
+        return default
+    try:
+        return float(v) * invscale / scale
+    except (ValueError, TypeError):
+        return default
+
+
+def bool_or_none(v, default=None):
+    return v if isinstance(v, bool) else default
+
+
+def strip_or_none(v, default=None):
+    return v.strip() if isinstance(v, compat_str) else default
+
+
+def url_or_none(url):
+    if not url or not isinstance(url, compat_str):
+        return None
+    url = url.strip()
+    return url if re.match(r'^(?:[a-zA-Z][\da-zA-Z.+-]*:)?//', url) else None

 
 
 def parse_duration(s):
 
 
 def parse_duration(s):


@@ -1302,44 +1968,56 @@ def parse_duration(s):
 
     s = s.strip()
 
 
     s = s.strip()
 

-    m = re.match(
-        r'''(?ix)(?:P?T)?
-        (?:
-            (?P<only_mins>[0-9.]+)\s*(?:mins?|minutes?)\s*|
-            (?P<only_hours>[0-9.]+)\s*(?:hours?)|
-
-            \s*(?P<hours_reversed>[0-9]+)\s*(?:[:h]|hours?)\s*(?P<mins_reversed>[0-9]+)\s*(?:[:m]|mins?|minutes?)\s*|
-            (?:
+    days, hours, mins, secs, ms = [None] * 5
+    m = re.match(r'(?:(?:(?:(?P<days>[0-9]+):)?(?P<hours>[0-9]+):)?(?P<mins>[0-9]+):)?(?P<secs>[0-9]+)(?P<ms>\.[0-9]+)?Z?$', s)
+    if m:
+        days, hours, mins, secs, ms = m.groups()
+    else:
+        m = re.match(
+            r'''(?ix)(?:P?

                 (?:
                 (?:

-                    (?:(?P<days>[0-9]+)\s*(?:[:d]|days?)\s*)?
-                    (?P<hours>[0-9]+)\s*(?:[:h]|hours?)\s*
+                    [0-9]+\s*y(?:ears?)?\s*

                 )?
                 )?

-                (?P<mins>[0-9]+)\s*(?:[:m]|mins?|minutes?)\s*
-            )?
-            (?P<secs>[0-9]+)(?P<ms>\.[0-9]+)?\s*(?:s|secs?|seconds?)?
-        )$''', s)
-    if not m:
-        return None
-    res = 0
-    if m.group('only_mins'):
-        return float_or_none(m.group('only_mins'), invscale=60)
-    if m.group('only_hours'):
-        return float_or_none(m.group('only_hours'), invscale=60 * 60)
-    if m.group('secs'):
-        res += int(m.group('secs'))
-    if m.group('mins_reversed'):
-        res += int(m.group('mins_reversed')) * 60
-    if m.group('mins'):
-        res += int(m.group('mins')) * 60
-    if m.group('hours'):
-        res += int(m.group('hours')) * 60 * 60
-    if m.group('hours_reversed'):
-        res += int(m.group('hours_reversed')) * 60 * 60
-    if m.group('days'):
-        res += int(m.group('days')) * 24 * 60 * 60
-    if m.group('ms'):
-        res += float(m.group('ms'))
-    return res
+                (?:
+                    [0-9]+\s*m(?:onths?)?\s*
+                )?
+                (?:
+                    [0-9]+\s*w(?:eeks?)?\s*
+                )?
+                (?:
+                    (?P<days>[0-9]+)\s*d(?:ays?)?\s*
+                )?
+                T)?
+                (?:
+                    (?P<hours>[0-9]+)\s*h(?:ours?)?\s*
+                )?
+                (?:
+                    (?P<mins>[0-9]+)\s*m(?:in(?:ute)?s?)?\s*
+                )?
+                (?:
+                    (?P<secs>[0-9]+)(?P<ms>\.[0-9]+)?\s*s(?:ec(?:ond)?s?)?\s*
+                )?Z?$''', s)
+        if m:
+            days, hours, mins, secs, ms = m.groups()
+        else:
+            m = re.match(r'(?i)(?:(?P<hours>[0-9.]+)\s*(?:hours?)|(?P<mins>[0-9.]+)\s*(?:mins?\.?|minutes?)\s*)Z?$', s)
+            if m:
+                hours, mins = m.groups()
+            else:
+                return None
+
+    duration = 0
+    if secs:
+        duration += float(secs)
+    if mins:
+        duration += float(mins) * 60
+    if hours:
+        duration += float(hours) * 60 * 60
+    if days:
+        duration += float(days) * 24 * 60 * 60
+    if ms:
+        duration += float(ms)
+    return duration

 
 
 def prepend_extension(filename, ext, expected_real_ext=None):
 
 
 def prepend_extension(filename, ext, expected_real_ext=None):


@@ -1372,8 +2050,12 @@ def get_exe_version(exe, args=['--version'],
     """ Returns the version of the specified executable,
     or False if the executable is not present """
     try:
     """ Returns the version of the specified executable,
     or False if the executable is not present """
     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/ytdl-org/youtube-dl/issues/955#issuecomment-209789656

         out, _ = subprocess.Popen(
             [encodeArgument(exe)] + args,
         out, _ = subprocess.Popen(
             [encodeArgument(exe)] + args,

+            stdin=subprocess.PIPE,

             stdout=subprocess.PIPE, stderr=subprocess.STDOUT).communicate()
     except OSError:
         return False
             stdout=subprocess.PIPE, stderr=subprocess.STDOUT).communicate()
     except OSError:
         return False


@@ -1400,9 +2082,12 @@ class PagedList(object):
 
 
 class OnDemandPagedList(PagedList):
 
 
 class OnDemandPagedList(PagedList):

-    def __init__(self, pagefunc, pagesize):
+    def __init__(self, pagefunc, pagesize, use_cache=True):

         self._pagefunc = pagefunc
         self._pagesize = pagesize
         self._pagefunc = pagefunc
         self._pagesize = pagesize

+        self._use_cache = use_cache
+        if use_cache:
+            self._cache = {}

 
     def getslice(self, start=0, end=None):
         res = []
 
     def getslice(self, start=0, end=None):
         res = []


@@ -1412,7 +2097,13 @@ class OnDemandPagedList(PagedList):
             if start >= nextfirstid:
                 continue
 
             if start >= nextfirstid:
                 continue
 

-            page_results = list(self._pagefunc(pagenum))
+            page_results = None
+            if self._use_cache:
+                page_results = self._cache.get(pagenum)
+            if page_results is None:
+                page_results = list(self._pagefunc(pagenum))
+            if self._use_cache:
+                self._cache[pagenum] = page_results

 
             startv = (
                 start % self._pagesize
 
             startv = (
                 start % self._pagesize


@@ -1498,29 +2189,13 @@ def escape_url(url):
     """Escape URL as suggested by RFC 3986"""
     url_parsed = compat_urllib_parse_urlparse(url)
     return url_parsed._replace(
     """Escape URL as suggested by RFC 3986"""
     url_parsed = compat_urllib_parse_urlparse(url)
     return url_parsed._replace(

+        netloc=url_parsed.netloc.encode('idna').decode('ascii'),

         path=escape_rfc3986(url_parsed.path),
         params=escape_rfc3986(url_parsed.params),
         query=escape_rfc3986(url_parsed.query),
         fragment=escape_rfc3986(url_parsed.fragment)
     ).geturl()
 
         path=escape_rfc3986(url_parsed.path),
         params=escape_rfc3986(url_parsed.params),
         query=escape_rfc3986(url_parsed.query),
         fragment=escape_rfc3986(url_parsed.fragment)
     ).geturl()
 

-try:
-    struct.pack('!I', 0)
-except TypeError:
-    # In Python 2.6 (and some 2.7 versions), struct requires a bytes argument
-    def struct_pack(spec, *args):
-        if isinstance(spec, compat_str):
-            spec = spec.encode('ascii')
-        return struct.pack(spec, *args)
-
-    def struct_unpack(spec, *args):
-        if isinstance(spec, compat_str):
-            spec = spec.encode('ascii')
-        return struct.unpack(spec, *args)
-else:
-    struct_pack = struct.pack
-    struct_unpack = struct.unpack
-

 
 def read_batch_urls(batch_fd):
     def fixup(url):
 
 def read_batch_urls(batch_fd):
     def fixup(url):


@@ -1539,30 +2214,130 @@ def read_batch_urls(batch_fd):
 
 
 def urlencode_postdata(*args, **kargs):
 
 
 def urlencode_postdata(*args, **kargs):

-    return compat_urllib_parse.urlencode(*args, **kargs).encode('ascii')
+    return compat_urllib_parse_urlencode(*args, **kargs).encode('ascii')
+
+
+def update_url_query(url, query):
+    if not query:
+        return url
+    parsed_url = compat_urlparse.urlparse(url)
+    qs = compat_parse_qs(parsed_url.query)
+    qs.update(query)
+    return compat_urlparse.urlunparse(parsed_url._replace(
+        query=compat_urllib_parse_urlencode(qs, True)))
+
+
+def update_Request(req, url=None, data=None, headers={}, query={}):
+    req_headers = req.headers.copy()
+    req_headers.update(headers)
+    req_data = data or req.data
+    req_url = update_url_query(url or req.get_full_url(), query)
+    req_get_method = req.get_method()
+    if req_get_method == 'HEAD':
+        req_type = HEADRequest
+    elif req_get_method == 'PUT':
+        req_type = PUTRequest
+    else:
+        req_type = compat_urllib_request.Request
+    new_req = req_type(
+        req_url, data=req_data, headers=req_headers,
+        origin_req_host=req.origin_req_host, unverifiable=req.unverifiable)
+    if hasattr(req, 'timeout'):
+        new_req.timeout = req.timeout
+    return new_req
+
+
+def _multipart_encode_impl(data, boundary):
+    content_type = 'multipart/form-data; boundary=%s' % boundary
+
+    out = b''
+    for k, v in data.items():
+        out += b'--' + boundary.encode('ascii') + b'\r\n'
+        if isinstance(k, compat_str):
+            k = k.encode('utf-8')
+        if isinstance(v, compat_str):
+            v = v.encode('utf-8')
+        # RFC 2047 requires non-ASCII field names to be encoded, while RFC 7578
+        # suggests sending UTF-8 directly. Firefox sends UTF-8, too
+        content = b'Content-Disposition: form-data; name="' + k + b'"\r\n\r\n' + v + b'\r\n'
+        if boundary.encode('ascii') in content:
+            raise ValueError('Boundary overlaps with data')
+        out += content
+
+    out += b'--' + boundary.encode('ascii') + b'--\r\n'
+
+    return out, content_type
+
+
+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
+        objects.
+    boundary:
+        If specified a Unicode object, it's used as the boundary. Otherwise
+        a random boundary is generated.
+
+    Reference: https://tools.ietf.org/html/rfc7578
+    '''
+    has_specified_boundary = boundary is not None
+
+    while True:
+        if boundary is None:
+            boundary = '---------------' + str(random.randrange(0x0fffffff, 0xffffffff))
+
+        try:
+            out, content_type = _multipart_encode_impl(data, boundary)
+            break
+        except ValueError:
+            if has_specified_boundary:
+                raise
+            boundary = None

 
 

+    return out, content_type

 
 

-try:
-    etree_iter = xml.etree.ElementTree.Element.iter
-except AttributeError:  # Python <=2.6
-    etree_iter = lambda n: n.findall('.//*')

 
 

+def dict_get(d, key_or_keys, default=None, skip_false_values=True):
+    if isinstance(key_or_keys, (list, tuple)):
+        for key in key_or_keys:
+            if key not in d or d[key] is None or skip_false_values and not d[key]:
+                continue
+            return d[key]
+        return default
+    return d.get(key_or_keys, default)

 
 

-def parse_xml(s):
-    class TreeBuilder(xml.etree.ElementTree.TreeBuilder):
-        def doctype(self, name, pubid, system):
-            pass  # Ignore doctypes

 
 

-    parser = xml.etree.ElementTree.XMLParser(target=TreeBuilder())
-    kwargs = {'parser': parser} if sys.version_info >= (2, 7) else {}
-    tree = xml.etree.ElementTree.XML(s.encode('utf-8'), **kwargs)
-    # Fix up XML parser in Python 2.x
-    if sys.version_info < (3, 0):
-        for n in etree_iter(tree):
-            if n.text is not None:
-                if not isinstance(n.text, compat_str):
-                    n.text = n.text.decode('utf-8')
-    return tree
+def try_get(src, getter, expected_type=None):
+    if not isinstance(getter, (list, tuple)):
+        getter = [getter]
+    for get in getter:
+        try:
+            v = get(src)
+        except (AttributeError, KeyError, TypeError, IndexError):
+            pass
+        else:
+            if expected_type is None or isinstance(v, expected_type):
+                return v
+
+
+def merge_dicts(*dicts):
+    merged = {}
+    for a_dict in 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])):
+                merged[k] = v
+    return merged
+
+
+def encode_compat_str(string, encoding=preferredencoding(), errors='strict'):
+    return string if isinstance(string, compat_str) else compat_str(string, encoding, errors)

 
 
 US_RATINGS = {
 
 
 US_RATINGS = {


@@ -1574,41 +2349,81 @@ US_RATINGS = {
 }
 
 
 }
 
 

+TV_PARENTAL_GUIDELINES = {
+    'TV-Y': 0,
+    'TV-Y7': 7,
+    'TV-G': 0,
+    'TV-PG': 0,
+    'TV-14': 14,
+    'TV-MA': 17,
+}
+
+

 def parse_age_limit(s):
 def parse_age_limit(s):

-    if s is None:
+    if type(s) == int:
+        return s if 0 <= s <= 21 else None
+    if not isinstance(s, compat_basestring):

         return None
     m = re.match(r'^(?P<age>\d{1,2})\+?$', s)
         return None
     m = re.match(r'^(?P<age>\d{1,2})\+?$', s)

-    return int(m.group('age')) if m else US_RATINGS.get(s, None)
+    if m:
+        return int(m.group('age'))
+    if s in US_RATINGS:
+        return US_RATINGS[s]
+    m = re.match(r'^TV[_-]?(%s)$' % '|'.join(k[3:] for k in TV_PARENTAL_GUIDELINES), s)
+    if m:
+        return TV_PARENTAL_GUIDELINES['TV-' + m.group(1)]
+    return None

 
 
 def strip_jsonp(code):
     return re.sub(
 
 
 def strip_jsonp(code):
     return re.sub(

-        r'(?s)^[a-zA-Z0-9_]+\s*\(\s*(.*)\);?\s*?(?://[^\n]*)*$', r'\1', code)
+        r'''(?sx)^
+            (?:window\.)?(?P<func_name>[a-zA-Z0-9_.$]*)
+            (?:\s*&&\s*(?P=func_name))?
+            \s*\(\s*(?P<callback_data>.*)\);?
+            \s*?(?://[^\n]*)*$''',
+        r'\g<callback_data>', code)

 
 
 def js_to_json(code):
 
 
 def js_to_json(code):

+    COMMENT_RE = r'/\*(?:(?!\*/).)*?\*/|//[^\n]*'
+    SKIP_RE = r'\s*(?:{comment})?\s*'.format(comment=COMMENT_RE)
+    INTEGER_TABLE = (
+        (r'(?s)^(0[xX][0-9a-fA-F]+){skip}:?$'.format(skip=SKIP_RE), 16),
+        (r'(?s)^(0+[0-7]+){skip}:?$'.format(skip=SKIP_RE), 8),
+    )
+

     def fix_kv(m):
         v = m.group(0)
         if v in ('true', 'false', 'null'):
             return v
     def fix_kv(m):
         v = m.group(0)
         if v in ('true', 'false', 'null'):
             return v

-        if v.startswith('"'):
-            return v
-        if v.startswith("'"):
-            v = v[1:-1]
-            v = re.sub(r"\\\\|\\'|\"", lambda m: {
-                '\\\\': '\\\\',
-                "\\'": "'",
+        elif v.startswith('/*') or v.startswith('//') or v == ',':
+            return ""
+
+        if v[0] in ("'", '"'):
+            v = re.sub(r'(?s)\\.|"', lambda m: {

                 '"': '\\"',
                 '"': '\\"',

-            }[m.group(0)], v)
+                "\\'": "'",
+                '\\\n': '',
+                '\\x': '\\u00',
+            }.get(m.group(0), m.group(0)), v[1:-1])
+
+        for regex, base in INTEGER_TABLE:
+            im = re.match(regex, v)
+            if im:
+                i = int(im.group(1), base)
+                return '"%d":' % i if v.endswith(':') else '%d' % i
+

         return '"%s"' % v
 
         return '"%s"' % v
 

-    res = re.sub(r'''(?x)
-        "(?:[^"\\]*(?:\\\\|\\['"nu]))*[^"\\]*"|
-        '(?:[^'\\]*(?:\\\\|\\['"nu]))*[^'\\]*'|
-        [a-zA-Z_][.a-zA-Z_0-9]*
-        ''', fix_kv, code)
-    res = re.sub(r',(\s*[\]}])', lambda m: m.group(1), res)
-    return res
+    return re.sub(r'''(?sx)
+        "(?:[^"\\]*(?:\\\\|\\['"nurtbfx/\n]))*[^"\\]*"|
+        '(?:[^'\\]*(?:\\\\|\\['"nurtbfx/\n]))*[^'\\]*'|
+        {comment}|,(?={skip}[\]}}])|
+        (?:(?<![0-9])[eE]|[a-df-zA-DF-Z_])[.a-zA-Z_0-9]*|
+        \b(?:0[xX][0-9a-fA-F]+|0+[0-7]+)(?:{skip}:)?|
+        [0-9]+(?={skip}:)
+        '''.format(comment=COMMENT_RE, skip=SKIP_RE), fix_kv, code)

 
 
 def qualities(quality_ids):
 
 
 def qualities(quality_ids):


@@ -1656,25 +2471,93 @@ def ytdl_is_updateable():
 
 def args_to_str(args):
     # Get a short string representation for a subprocess command
 
 def args_to_str(args):
     # Get a short string representation for a subprocess command

-    return ' '.join(shlex_quote(a) for a in args)
+    return ' '.join(compat_shlex_quote(a) for a in args)
+
+
+def error_to_compat_str(err):
+    err_str = str(err)
+    # On python 2 error byte string must be decoded with proper
+    # encoding rather than ascii
+    if sys.version_info[0] < 3:
+        err_str = err_str.decode(preferredencoding())
+    return err_str

 
 
 def mimetype2ext(mt):
 
 
 def mimetype2ext(mt):

+    if mt is None:
+        return None
+
+    ext = {
+        'audio/mp4': 'm4a',
+        # Per RFC 3003, audio/mpeg can be .mp1, .mp2 or .mp3. Here use .mp3 as
+        # it's the most popular one
+        'audio/mpeg': 'mp3',
+    }.get(mt)
+    if ext is not None:
+        return ext
+

     _, _, res = mt.rpartition('/')
     _, _, res = mt.rpartition('/')

+    res = res.split(';')[0].strip().lower()

 
     return {
 
     return {

-        'x-ms-wmv': 'wmv',
-        'x-mp4-fragmented': 'mp4',
+        '3gpp': '3gp',
+        'smptett+xml': 'tt',
+        'ttaf+xml': 'dfxp',

         'ttml+xml': 'ttml',
         'ttml+xml': 'ttml',

+        'x-flv': 'flv',
+        'x-mp4-fragmented': 'mp4',
+        'x-ms-sami': 'sami',
+        'x-ms-wmv': 'wmv',
+        'mpegurl': 'm3u8',
+        'x-mpegurl': 'm3u8',
+        'vnd.apple.mpegurl': 'm3u8',
+        'dash+xml': 'mpd',
+        'f4m+xml': 'f4m',
+        'hds+xml': 'f4m',
+        'vnd.ms-sstr+xml': 'ism',
+        'quicktime': 'mov',
+        'mp2t': 'ts',

     }.get(res, res)
 
 
     }.get(res, res)
 
 

+def parse_codecs(codecs_str):
+    # http://tools.ietf.org/html/rfc6381
+    if not codecs_str:
+        return {}
+    splited_codecs = list(filter(None, map(
+        lambda str: str.strip(), codecs_str.strip().strip(',').split(','))))
+    vcodec, acodec = None, None
+    for full_codec in splited_codecs:
+        codec = full_codec.split('.')[0]
+        if codec in ('avc1', 'avc2', 'avc3', 'avc4', 'vp9', 'vp8', 'hev1', 'hev2', 'h263', 'h264', 'mp4v', 'hvc1', 'av01'):
+            if not vcodec:
+                vcodec = full_codec
+        elif codec in ('mp4a', 'opus', 'vorbis', 'mp3', 'aac', 'ac-3', 'ec-3', 'eac3', 'dtsc', 'dtse', 'dtsh', 'dtsl'):
+            if not acodec:
+                acodec = full_codec
+        else:
+            write_string('WARNING: Unknown codec %s\n' % full_codec, sys.stderr)
+    if not vcodec and not acodec:
+        if len(splited_codecs) == 2:
+            return {
+                'vcodec': vcodec,
+                'acodec': acodec,
+            }
+        elif len(splited_codecs) == 1:
+            return {
+                'vcodec': 'none',
+                'acodec': vcodec,
+            }
+    else:
+        return {
+            'vcodec': vcodec or 'none',
+            'acodec': acodec or 'none',
+        }
+    return {}
+
+

 def urlhandle_detect_ext(url_handle):
 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
+    getheader = url_handle.headers.get

 
     cd = getheader('Content-Disposition')
     if cd:
 
     cd = getheader('Content-Disposition')
     if cd:


@@ -1687,6 +2570,10 @@ def urlhandle_detect_ext(url_handle):
     return mimetype2ext(getheader('Content-Type'))
 
 
     return mimetype2ext(getheader('Content-Type'))
 
 

+def encode_data_uri(data, mime_type):
+    return 'data:%s;base64,%s' % (mime_type, base64.b64encode(data).decode('ascii'))
+
+

 def age_restricted(content_limit, age_limit):
     """ Returns True iff the content should be blocked """
 
 def age_restricted(content_limit, age_limit):
     """ Returns True iff the content should be blocked """
 


@@ -1761,6 +2648,7 @@ def _match_one(filter_part, dct):
         \s*(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
         (?:
             (?P<intval>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)|
         \s*(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
         (?:
             (?P<intval>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?)|

+            (?P<quote>["\'])(?P<quotedstrval>(?:\\.|(?!(?P=quote)|\\).)+?)(?P=quote)|

             (?P<strval>(?![0-9.])[a-z0-9A-Z]*)
         )
         \s*$
             (?P<strval>(?![0-9.])[a-z0-9A-Z]*)
         )
         \s*$


@@ -1768,11 +2656,22 @@ def _match_one(filter_part, dct):
     m = operator_rex.search(filter_part)
     if m:
         op = COMPARISON_OPERATORS[m.group('op')]
     m = operator_rex.search(filter_part)
     if m:
         op = COMPARISON_OPERATORS[m.group('op')]

-        if m.group('strval') is not None:
+        actual_value = dct.get(m.group('key'))
+        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/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'))
             if m.group('op') not in ('=', '!='):
                 raise ValueError(
                     'Operator %s does not support string values!' % m.group('op'))

-            comparison_value = m.group('strval')
+            comparison_value = m.group('quotedstrval') or m.group('strval') or m.group('intval')
+            quote = m.group('quote')
+            if quote is not None:
+                comparison_value = comparison_value.replace(r'\%s' % quote, quote)

         else:
             try:
                 comparison_value = int(m.group('intval'))
         else:
             try:
                 comparison_value = int(m.group('intval'))


@@ -1784,14 +2683,13 @@ def _match_one(filter_part, dct):
                     raise ValueError(
                         'Invalid integer value %r in filter part %r' % (
                             m.group('intval'), filter_part))
                     raise ValueError(
                         'Invalid integer value %r in filter part %r' % (
                             m.group('intval'), filter_part))

-        actual_value = dct.get(m.group('key'))

         if actual_value is None:
             return m.group('none_inclusive')
         return op(actual_value, comparison_value)
 
     UNARY_OPERATORS = {
         if actual_value is None:
             return m.group('none_inclusive')
         return op(actual_value, comparison_value)
 
     UNARY_OPERATORS = {

-        '': lambda v: v is not None,
-        '!': lambda v: v is None,
+        '': lambda v: (v is True) if isinstance(v, bool) else (v is not None),
+        '!': lambda v: (v is False) if isinstance(v, bool) else (v is None),

     }
     operator_rex = re.compile(r'''(?x)\s*
         (?P<op>%s)\s*(?P<key>[a-z_]+)
     }
     operator_rex = re.compile(r'''(?x)\s*
         (?P<op>%s)\s*(?P<key>[a-z_]+)


@@ -1825,15 +2723,15 @@ def match_filter_func(filter_str):
 
 def parse_dfxp_time_expr(time_expr):
     if not time_expr:
 
 def parse_dfxp_time_expr(time_expr):
     if not time_expr:

-        return 0.0
+        return

 
     mobj = re.match(r'^(?P<time_offset>\d+(?:\.\d+)?)s?$', time_expr)
     if mobj:
         return float(mobj.group('time_offset'))
 
 
     mobj = re.match(r'^(?P<time_offset>\d+(?:\.\d+)?)s?$', time_expr)
     if mobj:
         return float(mobj.group('time_offset'))
 

-    mobj = re.match(r'^(\d+):(\d\d):(\d\d(?:\.\d+)?)$', time_expr)
+    mobj = re.match(r'^(\d+):(\d\d):(\d\d(?:(?:\.|:)\d+)?)$', time_expr)

     if mobj:
     if mobj:

-        return 3600 * int(mobj.group(1)) + 60 * int(mobj.group(2)) + float(mobj.group(3))
+        return 3600 * int(mobj.group(1)) + 60 * int(mobj.group(2)) + float(mobj.group(3).replace(':', '.'))

 
 
 def srt_subtitles_timecode(seconds):
 
 
 def srt_subtitles_timecode(seconds):


@@ -1841,35 +2739,160 @@ def srt_subtitles_timecode(seconds):
 
 
 def dfxp2srt(dfxp_data):
 
 
 def dfxp2srt(dfxp_data):

-    _x = functools.partial(xpath_with_ns, ns_map={'ttml': 'http://www.w3.org/ns/ttml'})
+    '''
+    @param dfxp_data A bytes-like object containing DFXP data
+    @returns A unicode object containing converted SRT data
+    '''
+    LEGACY_NAMESPACES = (
+        (b'http://www.w3.org/ns/ttml', [
+            b'http://www.w3.org/2004/11/ttaf1',
+            b'http://www.w3.org/2006/04/ttaf1',
+            b'http://www.w3.org/2006/10/ttaf1',
+        ]),
+        (b'http://www.w3.org/ns/ttml#styling', [
+            b'http://www.w3.org/ns/ttml#style',
+        ]),
+    )
+
+    SUPPORTED_STYLING = [
+        'color',
+        'fontFamily',
+        'fontSize',
+        'fontStyle',
+        'fontWeight',
+        'textDecoration'
+    ]

 
 

-    def parse_node(node):
-        str_or_empty = functools.partial(str_or_none, default='')
+    _x = functools.partial(xpath_with_ns, ns_map={
+        'xml': 'http://www.w3.org/XML/1998/namespace',
+        'ttml': 'http://www.w3.org/ns/ttml',
+        'tts': 'http://www.w3.org/ns/ttml#styling',
+    })

 
 

-        out = str_or_empty(node.text)
+    styles = {}
+    default_style = {}

 
 

-        for child in node:
-            if child.tag in (_x('ttml:br'), 'br'):
-                out += '\n' + str_or_empty(child.tail)
-            elif child.tag in (_x('ttml:span'), 'span'):
-                out += str_or_empty(parse_node(child))
+    class TTMLPElementParser(object):
+        _out = ''
+        _unclosed_elements = []
+        _applied_styles = []
+
+        def start(self, tag, attrib):
+            if tag in (_x('ttml:br'), 'br'):
+                self._out += '\n'

             else:
             else:

-                out += str_or_empty(xml.etree.ElementTree.tostring(child))
+                unclosed_elements = []
+                style = {}
+                element_style_id = attrib.get('style')
+                if default_style:
+                    style.update(default_style)
+                if element_style_id:
+                    style.update(styles.get(element_style_id, {}))
+                for prop in SUPPORTED_STYLING:
+                    prop_val = attrib.get(_x('tts:' + prop))
+                    if prop_val:
+                        style[prop] = prop_val
+                if style:
+                    font = ''
+                    for k, v in sorted(style.items()):
+                        if self._applied_styles and self._applied_styles[-1].get(k) == v:
+                            continue
+                        if k == 'color':
+                            font += ' color="%s"' % v
+                        elif k == 'fontSize':
+                            font += ' size="%s"' % v
+                        elif k == 'fontFamily':
+                            font += ' face="%s"' % v
+                        elif k == 'fontWeight' and v == 'bold':
+                            self._out += '<b>'
+                            unclosed_elements.append('b')
+                        elif k == 'fontStyle' and v == 'italic':
+                            self._out += '<i>'
+                            unclosed_elements.append('i')
+                        elif k == 'textDecoration' and v == 'underline':
+                            self._out += '<u>'
+                            unclosed_elements.append('u')
+                    if font:
+                        self._out += '<font' + font + '>'
+                        unclosed_elements.append('font')
+                    applied_style = {}
+                    if self._applied_styles:
+                        applied_style.update(self._applied_styles[-1])
+                    applied_style.update(style)
+                    self._applied_styles.append(applied_style)
+                self._unclosed_elements.append(unclosed_elements)
+
+        def end(self, tag):
+            if tag not in (_x('ttml:br'), 'br'):
+                unclosed_elements = self._unclosed_elements.pop()
+                for element in reversed(unclosed_elements):
+                    self._out += '</%s>' % element
+                if unclosed_elements and self._applied_styles:
+                    self._applied_styles.pop()
+
+        def data(self, data):
+            self._out += data
+
+        def close(self):
+            return self._out.strip()
+
+    def parse_node(node):
+        target = TTMLPElementParser()
+        parser = xml.etree.ElementTree.XMLParser(target=target)
+        parser.feed(xml.etree.ElementTree.tostring(node))
+        return parser.close()

 
 

-        return out
+    for k, v in LEGACY_NAMESPACES:
+        for ns in v:
+            dfxp_data = dfxp_data.replace(ns, k)

 
 

-    dfxp = xml.etree.ElementTree.fromstring(dfxp_data.encode('utf-8'))
+    dfxp = compat_etree_fromstring(dfxp_data)

     out = []
     paras = dfxp.findall(_x('.//ttml:p')) or dfxp.findall('.//p')
 
     if not paras:
         raise ValueError('Invalid dfxp/TTML subtitle')
 
     out = []
     paras = dfxp.findall(_x('.//ttml:p')) or dfxp.findall('.//p')
 
     if not paras:
         raise ValueError('Invalid dfxp/TTML subtitle')
 

+    repeat = False
+    while True:
+        for style in dfxp.findall(_x('.//ttml:style')):
+            style_id = style.get('id') or style.get(_x('xml:id'))
+            if not style_id:
+                continue
+            parent_style_id = style.get('style')
+            if parent_style_id:
+                if parent_style_id not in styles:
+                    repeat = True
+                    continue
+                styles[style_id] = styles[parent_style_id].copy()
+            for prop in SUPPORTED_STYLING:
+                prop_val = style.get(_x('tts:' + prop))
+                if prop_val:
+                    styles.setdefault(style_id, {})[prop] = prop_val
+        if repeat:
+            repeat = False
+        else:
+            break
+
+    for p in ('body', 'div'):
+        ele = xpath_element(dfxp, [_x('.//ttml:' + p), './/' + p])
+        if ele is None:
+            continue
+        style = styles.get(ele.get('style'))
+        if not style:
+            continue
+        default_style.update(style)
+

     for para, index in zip(paras, itertools.count(1)):
     for para, index in zip(paras, itertools.count(1)):

-        begin_time = parse_dfxp_time_expr(para.attrib['begin'])
+        begin_time = parse_dfxp_time_expr(para.attrib.get('begin'))

         end_time = parse_dfxp_time_expr(para.attrib.get('end'))
         end_time = parse_dfxp_time_expr(para.attrib.get('end'))

+        dur = parse_dfxp_time_expr(para.attrib.get('dur'))
+        if begin_time is None:
+            continue

         if not end_time:
         if not end_time:

-            end_time = begin_time + parse_dfxp_time_expr(para.attrib['dur'])
+            if not dur:
+                continue
+            end_time = begin_time + dur

         out.append('%d\n%s --> %s\n%s\n\n' % (
             index,
             srt_subtitles_timecode(begin_time),
         out.append('%d\n%s --> %s\n%s\n\n' % (
             index,
             srt_subtitles_timecode(begin_time),


@@ -1879,6 +2902,758 @@ def dfxp2srt(dfxp_data):
     return ''.join(out)
 
 
     return ''.join(out)
 
 

+def cli_option(params, command_option, param):
+    param = params.get(param)
+    if param:
+        param = compat_str(param)
+    return [command_option, param] if param is not None else []
+
+
+def cli_bool_option(params, command_option, param, true_value='true', false_value='false', separator=None):
+    param = params.get(param)
+    if param is None:
+        return []
+    assert isinstance(param, bool)
+    if separator:
+        return [command_option + separator + (true_value if param else false_value)]
+    return [command_option, true_value if param else false_value]
+
+
+def cli_valueless_option(params, command_option, param, expected_value=True):
+    param = params.get(param)
+    return [command_option] if param == expected_value else []
+
+
+def cli_configuration_args(params, param, default=[]):
+    ex_args = params.get(param)
+    if ex_args is None:
+        return default
+    assert isinstance(ex_args, list)
+    return ex_args
+
+
+class ISO639Utils(object):
+    # See http://www.loc.gov/standards/iso639-2/ISO-639-2_utf-8.txt
+    _lang_map = {
+        'aa': 'aar',
+        'ab': 'abk',
+        'ae': 'ave',
+        'af': 'afr',
+        'ak': 'aka',
+        'am': 'amh',
+        'an': 'arg',
+        'ar': 'ara',
+        'as': 'asm',
+        'av': 'ava',
+        'ay': 'aym',
+        'az': 'aze',
+        'ba': 'bak',
+        'be': 'bel',
+        'bg': 'bul',
+        'bh': 'bih',
+        'bi': 'bis',
+        'bm': 'bam',
+        'bn': 'ben',
+        'bo': 'bod',
+        'br': 'bre',
+        'bs': 'bos',
+        'ca': 'cat',
+        'ce': 'che',
+        'ch': 'cha',
+        'co': 'cos',
+        'cr': 'cre',
+        'cs': 'ces',
+        'cu': 'chu',
+        'cv': 'chv',
+        'cy': 'cym',
+        'da': 'dan',
+        'de': 'deu',
+        'dv': 'div',
+        'dz': 'dzo',
+        'ee': 'ewe',
+        'el': 'ell',
+        'en': 'eng',
+        'eo': 'epo',
+        'es': 'spa',
+        'et': 'est',
+        'eu': 'eus',
+        'fa': 'fas',
+        'ff': 'ful',
+        'fi': 'fin',
+        'fj': 'fij',
+        'fo': 'fao',
+        'fr': 'fra',
+        'fy': 'fry',
+        'ga': 'gle',
+        'gd': 'gla',
+        'gl': 'glg',
+        'gn': 'grn',
+        'gu': 'guj',
+        'gv': 'glv',
+        'ha': 'hau',
+        'he': 'heb',
+        'iw': 'heb',  # Replaced by he in 1989 revision
+        'hi': 'hin',
+        'ho': 'hmo',
+        'hr': 'hrv',
+        'ht': 'hat',
+        'hu': 'hun',
+        'hy': 'hye',
+        'hz': 'her',
+        'ia': 'ina',
+        'id': 'ind',
+        'in': 'ind',  # Replaced by id in 1989 revision
+        'ie': 'ile',
+        'ig': 'ibo',
+        'ii': 'iii',
+        'ik': 'ipk',
+        'io': 'ido',
+        'is': 'isl',
+        'it': 'ita',
+        'iu': 'iku',
+        'ja': 'jpn',
+        'jv': 'jav',
+        'ka': 'kat',
+        'kg': 'kon',
+        'ki': 'kik',
+        'kj': 'kua',
+        'kk': 'kaz',
+        'kl': 'kal',
+        'km': 'khm',
+        'kn': 'kan',
+        'ko': 'kor',
+        'kr': 'kau',
+        'ks': 'kas',
+        'ku': 'kur',
+        'kv': 'kom',
+        'kw': 'cor',
+        'ky': 'kir',
+        'la': 'lat',
+        'lb': 'ltz',
+        'lg': 'lug',
+        'li': 'lim',
+        'ln': 'lin',
+        'lo': 'lao',
+        'lt': 'lit',
+        'lu': 'lub',
+        'lv': 'lav',
+        'mg': 'mlg',
+        'mh': 'mah',
+        'mi': 'mri',
+        'mk': 'mkd',
+        'ml': 'mal',
+        'mn': 'mon',
+        'mr': 'mar',
+        'ms': 'msa',
+        'mt': 'mlt',
+        'my': 'mya',
+        'na': 'nau',
+        'nb': 'nob',
+        'nd': 'nde',
+        'ne': 'nep',
+        'ng': 'ndo',
+        'nl': 'nld',
+        'nn': 'nno',
+        'no': 'nor',
+        'nr': 'nbl',
+        'nv': 'nav',
+        'ny': 'nya',
+        'oc': 'oci',
+        'oj': 'oji',
+        'om': 'orm',
+        'or': 'ori',
+        'os': 'oss',
+        'pa': 'pan',
+        'pi': 'pli',
+        'pl': 'pol',
+        'ps': 'pus',
+        'pt': 'por',
+        'qu': 'que',
+        'rm': 'roh',
+        'rn': 'run',
+        'ro': 'ron',
+        'ru': 'rus',
+        'rw': 'kin',
+        'sa': 'san',
+        'sc': 'srd',
+        'sd': 'snd',
+        'se': 'sme',
+        'sg': 'sag',
+        'si': 'sin',
+        'sk': 'slk',
+        'sl': 'slv',
+        'sm': 'smo',
+        'sn': 'sna',
+        'so': 'som',
+        'sq': 'sqi',
+        'sr': 'srp',
+        'ss': 'ssw',
+        'st': 'sot',
+        'su': 'sun',
+        'sv': 'swe',
+        'sw': 'swa',
+        'ta': 'tam',
+        'te': 'tel',
+        'tg': 'tgk',
+        'th': 'tha',
+        'ti': 'tir',
+        'tk': 'tuk',
+        'tl': 'tgl',
+        'tn': 'tsn',
+        'to': 'ton',
+        'tr': 'tur',
+        'ts': 'tso',
+        'tt': 'tat',
+        'tw': 'twi',
+        'ty': 'tah',
+        'ug': 'uig',
+        'uk': 'ukr',
+        'ur': 'urd',
+        'uz': 'uzb',
+        've': 'ven',
+        'vi': 'vie',
+        'vo': 'vol',
+        'wa': 'wln',
+        'wo': 'wol',
+        'xh': 'xho',
+        'yi': 'yid',
+        'ji': 'yid',  # Replaced by yi in 1989 revision
+        'yo': 'yor',
+        'za': 'zha',
+        'zh': 'zho',
+        'zu': 'zul',
+    }
+
+    @classmethod
+    def short2long(cls, code):
+        """Convert language code from ISO 639-1 to ISO 639-2/T"""
+        return cls._lang_map.get(code[:2])
+
+    @classmethod
+    def long2short(cls, code):
+        """Convert language code from ISO 639-2/T to ISO 639-1"""
+        for short_name, long_name in cls._lang_map.items():
+            if long_name == code:
+                return short_name
+
+
+class ISO3166Utils(object):
+    # From http://data.okfn.org/data/core/country-list
+    _country_map = {
+        'AF': 'Afghanistan',
+        'AX': 'Åland Islands',
+        'AL': 'Albania',
+        'DZ': 'Algeria',
+        'AS': 'American Samoa',
+        'AD': 'Andorra',
+        'AO': 'Angola',
+        'AI': 'Anguilla',
+        'AQ': 'Antarctica',
+        'AG': 'Antigua and Barbuda',
+        'AR': 'Argentina',
+        'AM': 'Armenia',
+        'AW': 'Aruba',
+        'AU': 'Australia',
+        'AT': 'Austria',
+        'AZ': 'Azerbaijan',
+        'BS': 'Bahamas',
+        'BH': 'Bahrain',
+        'BD': 'Bangladesh',
+        'BB': 'Barbados',
+        'BY': 'Belarus',
+        'BE': 'Belgium',
+        'BZ': 'Belize',
+        'BJ': 'Benin',
+        'BM': 'Bermuda',
+        'BT': 'Bhutan',
+        'BO': 'Bolivia, Plurinational State of',
+        'BQ': 'Bonaire, Sint Eustatius and Saba',
+        'BA': 'Bosnia and Herzegovina',
+        'BW': 'Botswana',
+        'BV': 'Bouvet Island',
+        'BR': 'Brazil',
+        'IO': 'British Indian Ocean Territory',
+        'BN': 'Brunei Darussalam',
+        'BG': 'Bulgaria',
+        'BF': 'Burkina Faso',
+        'BI': 'Burundi',
+        'KH': 'Cambodia',
+        'CM': 'Cameroon',
+        'CA': 'Canada',
+        'CV': 'Cape Verde',
+        'KY': 'Cayman Islands',
+        'CF': 'Central African Republic',
+        'TD': 'Chad',
+        'CL': 'Chile',
+        'CN': 'China',
+        'CX': 'Christmas Island',
+        'CC': 'Cocos (Keeling) Islands',
+        'CO': 'Colombia',
+        'KM': 'Comoros',
+        'CG': 'Congo',
+        'CD': 'Congo, the Democratic Republic of the',
+        'CK': 'Cook Islands',
+        'CR': 'Costa Rica',
+        'CI': 'Côte d\'Ivoire',
+        'HR': 'Croatia',
+        'CU': 'Cuba',
+        'CW': 'Curaçao',
+        'CY': 'Cyprus',
+        'CZ': 'Czech Republic',
+        'DK': 'Denmark',
+        'DJ': 'Djibouti',
+        'DM': 'Dominica',
+        'DO': 'Dominican Republic',
+        'EC': 'Ecuador',
+        'EG': 'Egypt',
+        'SV': 'El Salvador',
+        'GQ': 'Equatorial Guinea',
+        'ER': 'Eritrea',
+        'EE': 'Estonia',
+        'ET': 'Ethiopia',
+        'FK': 'Falkland Islands (Malvinas)',
+        'FO': 'Faroe Islands',
+        'FJ': 'Fiji',
+        'FI': 'Finland',
+        'FR': 'France',
+        'GF': 'French Guiana',
+        'PF': 'French Polynesia',
+        'TF': 'French Southern Territories',
+        'GA': 'Gabon',
+        'GM': 'Gambia',
+        'GE': 'Georgia',
+        'DE': 'Germany',
+        'GH': 'Ghana',
+        'GI': 'Gibraltar',
+        'GR': 'Greece',
+        'GL': 'Greenland',
+        'GD': 'Grenada',
+        'GP': 'Guadeloupe',
+        'GU': 'Guam',
+        'GT': 'Guatemala',
+        'GG': 'Guernsey',
+        'GN': 'Guinea',
+        'GW': 'Guinea-Bissau',
+        'GY': 'Guyana',
+        'HT': 'Haiti',
+        'HM': 'Heard Island and McDonald Islands',
+        'VA': 'Holy See (Vatican City State)',
+        'HN': 'Honduras',
+        'HK': 'Hong Kong',
+        'HU': 'Hungary',
+        'IS': 'Iceland',
+        'IN': 'India',
+        'ID': 'Indonesia',
+        'IR': 'Iran, Islamic Republic of',
+        'IQ': 'Iraq',
+        'IE': 'Ireland',
+        'IM': 'Isle of Man',
+        'IL': 'Israel',
+        'IT': 'Italy',
+        'JM': 'Jamaica',
+        'JP': 'Japan',
+        'JE': 'Jersey',
+        'JO': 'Jordan',
+        'KZ': 'Kazakhstan',
+        'KE': 'Kenya',
+        'KI': 'Kiribati',
+        'KP': 'Korea, Democratic People\'s Republic of',
+        'KR': 'Korea, Republic of',
+        'KW': 'Kuwait',
+        'KG': 'Kyrgyzstan',
+        'LA': 'Lao People\'s Democratic Republic',
+        'LV': 'Latvia',
+        'LB': 'Lebanon',
+        'LS': 'Lesotho',
+        'LR': 'Liberia',
+        'LY': 'Libya',
+        'LI': 'Liechtenstein',
+        'LT': 'Lithuania',
+        'LU': 'Luxembourg',
+        'MO': 'Macao',
+        'MK': 'Macedonia, the Former Yugoslav Republic of',
+        'MG': 'Madagascar',
+        'MW': 'Malawi',
+        'MY': 'Malaysia',
+        'MV': 'Maldives',
+        'ML': 'Mali',
+        'MT': 'Malta',
+        'MH': 'Marshall Islands',
+        'MQ': 'Martinique',
+        'MR': 'Mauritania',
+        'MU': 'Mauritius',
+        'YT': 'Mayotte',
+        'MX': 'Mexico',
+        'FM': 'Micronesia, Federated States of',
+        'MD': 'Moldova, Republic of',
+        'MC': 'Monaco',
+        'MN': 'Mongolia',
+        'ME': 'Montenegro',
+        'MS': 'Montserrat',
+        'MA': 'Morocco',
+        'MZ': 'Mozambique',
+        'MM': 'Myanmar',
+        'NA': 'Namibia',
+        'NR': 'Nauru',
+        'NP': 'Nepal',
+        'NL': 'Netherlands',
+        'NC': 'New Caledonia',
+        'NZ': 'New Zealand',
+        'NI': 'Nicaragua',
+        'NE': 'Niger',
+        'NG': 'Nigeria',
+        'NU': 'Niue',
+        'NF': 'Norfolk Island',
+        'MP': 'Northern Mariana Islands',
+        'NO': 'Norway',
+        'OM': 'Oman',
+        'PK': 'Pakistan',
+        'PW': 'Palau',
+        'PS': 'Palestine, State of',
+        'PA': 'Panama',
+        'PG': 'Papua New Guinea',
+        'PY': 'Paraguay',
+        'PE': 'Peru',
+        'PH': 'Philippines',
+        'PN': 'Pitcairn',
+        'PL': 'Poland',
+        'PT': 'Portugal',
+        'PR': 'Puerto Rico',
+        'QA': 'Qatar',
+        'RE': 'Réunion',
+        'RO': 'Romania',
+        'RU': 'Russian Federation',
+        'RW': 'Rwanda',
+        'BL': 'Saint Barthélemy',
+        'SH': 'Saint Helena, Ascension and Tristan da Cunha',
+        'KN': 'Saint Kitts and Nevis',
+        'LC': 'Saint Lucia',
+        'MF': 'Saint Martin (French part)',
+        'PM': 'Saint Pierre and Miquelon',
+        'VC': 'Saint Vincent and the Grenadines',
+        'WS': 'Samoa',
+        'SM': 'San Marino',
+        'ST': 'Sao Tome and Principe',
+        'SA': 'Saudi Arabia',
+        'SN': 'Senegal',
+        'RS': 'Serbia',
+        'SC': 'Seychelles',
+        'SL': 'Sierra Leone',
+        'SG': 'Singapore',
+        'SX': 'Sint Maarten (Dutch part)',
+        'SK': 'Slovakia',
+        'SI': 'Slovenia',
+        'SB': 'Solomon Islands',
+        'SO': 'Somalia',
+        'ZA': 'South Africa',
+        'GS': 'South Georgia and the South Sandwich Islands',
+        'SS': 'South Sudan',
+        'ES': 'Spain',
+        'LK': 'Sri Lanka',
+        'SD': 'Sudan',
+        'SR': 'Suriname',
+        'SJ': 'Svalbard and Jan Mayen',
+        'SZ': 'Swaziland',
+        'SE': 'Sweden',
+        'CH': 'Switzerland',
+        'SY': 'Syrian Arab Republic',
+        'TW': 'Taiwan, Province of China',
+        'TJ': 'Tajikistan',
+        'TZ': 'Tanzania, United Republic of',
+        'TH': 'Thailand',
+        'TL': 'Timor-Leste',
+        'TG': 'Togo',
+        'TK': 'Tokelau',
+        'TO': 'Tonga',
+        'TT': 'Trinidad and Tobago',
+        'TN': 'Tunisia',
+        'TR': 'Turkey',
+        'TM': 'Turkmenistan',
+        'TC': 'Turks and Caicos Islands',
+        'TV': 'Tuvalu',
+        'UG': 'Uganda',
+        'UA': 'Ukraine',
+        'AE': 'United Arab Emirates',
+        'GB': 'United Kingdom',
+        'US': 'United States',
+        'UM': 'United States Minor Outlying Islands',
+        'UY': 'Uruguay',
+        'UZ': 'Uzbekistan',
+        'VU': 'Vanuatu',
+        'VE': 'Venezuela, Bolivarian Republic of',
+        'VN': 'Viet Nam',
+        'VG': 'Virgin Islands, British',
+        'VI': 'Virgin Islands, U.S.',
+        'WF': 'Wallis and Futuna',
+        'EH': 'Western Sahara',
+        'YE': 'Yemen',
+        'ZM': 'Zambia',
+        'ZW': 'Zimbabwe',
+    }
+
+    @classmethod
+    def short2full(cls, code):
+        """Convert an ISO 3166-2 country code to the corresponding full name"""
+        return cls._country_map.get(code.upper())
+
+
+class GeoUtils(object):
+    # Major IPv4 address blocks per country
+    _country_ip_map = {
+        'AD': '85.94.160.0/19',
+        'AE': '94.200.0.0/13',
+        'AF': '149.54.0.0/17',
+        'AG': '209.59.64.0/18',
+        'AI': '204.14.248.0/21',
+        'AL': '46.99.0.0/16',
+        'AM': '46.70.0.0/15',
+        'AO': '105.168.0.0/13',
+        'AP': '159.117.192.0/21',
+        'AR': '181.0.0.0/12',
+        'AS': '202.70.112.0/20',
+        'AT': '84.112.0.0/13',
+        'AU': '1.128.0.0/11',
+        'AW': '181.41.0.0/18',
+        'AZ': '5.191.0.0/16',
+        'BA': '31.176.128.0/17',
+        'BB': '65.48.128.0/17',
+        'BD': '114.130.0.0/16',
+        'BE': '57.0.0.0/8',
+        'BF': '129.45.128.0/17',
+        'BG': '95.42.0.0/15',
+        'BH': '37.131.0.0/17',
+        'BI': '154.117.192.0/18',
+        'BJ': '137.255.0.0/16',
+        'BL': '192.131.134.0/24',
+        'BM': '196.12.64.0/18',
+        'BN': '156.31.0.0/16',
+        'BO': '161.56.0.0/16',
+        'BQ': '161.0.80.0/20',
+        'BR': '152.240.0.0/12',
+        'BS': '24.51.64.0/18',
+        'BT': '119.2.96.0/19',
+        'BW': '168.167.0.0/16',
+        'BY': '178.120.0.0/13',
+        'BZ': '179.42.192.0/18',
+        'CA': '99.224.0.0/11',
+        'CD': '41.243.0.0/16',
+        'CF': '196.32.200.0/21',
+        'CG': '197.214.128.0/17',
+        'CH': '85.0.0.0/13',
+        'CI': '154.232.0.0/14',
+        'CK': '202.65.32.0/19',
+        'CL': '152.172.0.0/14',
+        'CM': '165.210.0.0/15',
+        'CN': '36.128.0.0/10',
+        'CO': '181.240.0.0/12',
+        'CR': '201.192.0.0/12',
+        'CU': '152.206.0.0/15',
+        'CV': '165.90.96.0/19',
+        'CW': '190.88.128.0/17',
+        'CY': '46.198.0.0/15',
+        'CZ': '88.100.0.0/14',
+        'DE': '53.0.0.0/8',
+        'DJ': '197.241.0.0/17',
+        'DK': '87.48.0.0/12',
+        'DM': '192.243.48.0/20',
+        'DO': '152.166.0.0/15',
+        'DZ': '41.96.0.0/12',
+        'EC': '186.68.0.0/15',
+        'EE': '90.190.0.0/15',
+        'EG': '156.160.0.0/11',
+        'ER': '196.200.96.0/20',
+        'ES': '88.0.0.0/11',
+        'ET': '196.188.0.0/14',
+        'EU': '2.16.0.0/13',
+        'FI': '91.152.0.0/13',
+        'FJ': '144.120.0.0/16',
+        'FM': '119.252.112.0/20',
+        'FO': '88.85.32.0/19',
+        'FR': '90.0.0.0/9',
+        'GA': '41.158.0.0/15',
+        'GB': '25.0.0.0/8',
+        'GD': '74.122.88.0/21',
+        'GE': '31.146.0.0/16',
+        'GF': '161.22.64.0/18',
+        'GG': '62.68.160.0/19',
+        'GH': '45.208.0.0/14',
+        'GI': '85.115.128.0/19',
+        'GL': '88.83.0.0/19',
+        'GM': '160.182.0.0/15',
+        'GN': '197.149.192.0/18',
+        'GP': '104.250.0.0/19',
+        'GQ': '105.235.224.0/20',
+        'GR': '94.64.0.0/13',
+        'GT': '168.234.0.0/16',
+        'GU': '168.123.0.0/16',
+        'GW': '197.214.80.0/20',
+        'GY': '181.41.64.0/18',
+        'HK': '113.252.0.0/14',
+        'HN': '181.210.0.0/16',
+        'HR': '93.136.0.0/13',
+        'HT': '148.102.128.0/17',
+        'HU': '84.0.0.0/14',
+        'ID': '39.192.0.0/10',
+        'IE': '87.32.0.0/12',
+        'IL': '79.176.0.0/13',
+        'IM': '5.62.80.0/20',
+        'IN': '117.192.0.0/10',
+        'IO': '203.83.48.0/21',
+        'IQ': '37.236.0.0/14',
+        'IR': '2.176.0.0/12',
+        'IS': '82.221.0.0/16',
+        'IT': '79.0.0.0/10',
+        'JE': '87.244.64.0/18',
+        'JM': '72.27.0.0/17',
+        'JO': '176.29.0.0/16',
+        'JP': '126.0.0.0/8',
+        'KE': '105.48.0.0/12',
+        'KG': '158.181.128.0/17',
+        'KH': '36.37.128.0/17',
+        'KI': '103.25.140.0/22',
+        'KM': '197.255.224.0/20',
+        'KN': '198.32.32.0/19',
+        'KP': '175.45.176.0/22',
+        'KR': '175.192.0.0/10',
+        'KW': '37.36.0.0/14',
+        'KY': '64.96.0.0/15',
+        'KZ': '2.72.0.0/13',
+        'LA': '115.84.64.0/18',
+        'LB': '178.135.0.0/16',
+        'LC': '192.147.231.0/24',
+        'LI': '82.117.0.0/19',
+        'LK': '112.134.0.0/15',
+        'LR': '41.86.0.0/19',
+        'LS': '129.232.0.0/17',
+        'LT': '78.56.0.0/13',
+        'LU': '188.42.0.0/16',
+        'LV': '46.109.0.0/16',
+        'LY': '41.252.0.0/14',
+        'MA': '105.128.0.0/11',
+        'MC': '88.209.64.0/18',
+        'MD': '37.246.0.0/16',
+        'ME': '178.175.0.0/17',
+        'MF': '74.112.232.0/21',
+        'MG': '154.126.0.0/17',
+        'MH': '117.103.88.0/21',
+        'MK': '77.28.0.0/15',
+        'ML': '154.118.128.0/18',
+        'MM': '37.111.0.0/17',
+        'MN': '49.0.128.0/17',
+        'MO': '60.246.0.0/16',
+        'MP': '202.88.64.0/20',
+        'MQ': '109.203.224.0/19',
+        'MR': '41.188.64.0/18',
+        'MS': '208.90.112.0/22',
+        'MT': '46.11.0.0/16',
+        'MU': '105.16.0.0/12',
+        'MV': '27.114.128.0/18',
+        'MW': '105.234.0.0/16',
+        'MX': '187.192.0.0/11',
+        'MY': '175.136.0.0/13',
+        'MZ': '197.218.0.0/15',
+        'NA': '41.182.0.0/16',
+        'NC': '101.101.0.0/18',
+        'NE': '197.214.0.0/18',
+        'NF': '203.17.240.0/22',
+        'NG': '105.112.0.0/12',
+        'NI': '186.76.0.0/15',
+        'NL': '145.96.0.0/11',
+        'NO': '84.208.0.0/13',
+        'NP': '36.252.0.0/15',
+        'NR': '203.98.224.0/19',
+        'NU': '49.156.48.0/22',
+        'NZ': '49.224.0.0/14',
+        'OM': '5.36.0.0/15',
+        'PA': '186.72.0.0/15',
+        'PE': '186.160.0.0/14',
+        'PF': '123.50.64.0/18',
+        'PG': '124.240.192.0/19',
+        'PH': '49.144.0.0/13',
+        'PK': '39.32.0.0/11',
+        'PL': '83.0.0.0/11',
+        'PM': '70.36.0.0/20',
+        'PR': '66.50.0.0/16',
+        'PS': '188.161.0.0/16',
+        'PT': '85.240.0.0/13',
+        'PW': '202.124.224.0/20',
+        'PY': '181.120.0.0/14',
+        'QA': '37.210.0.0/15',
+        'RE': '139.26.0.0/16',
+        'RO': '79.112.0.0/13',
+        'RS': '178.220.0.0/14',
+        'RU': '5.136.0.0/13',
+        'RW': '105.178.0.0/15',
+        'SA': '188.48.0.0/13',
+        'SB': '202.1.160.0/19',
+        'SC': '154.192.0.0/11',
+        'SD': '154.96.0.0/13',
+        'SE': '78.64.0.0/12',
+        'SG': '152.56.0.0/14',
+        'SI': '188.196.0.0/14',
+        'SK': '78.98.0.0/15',
+        'SL': '197.215.0.0/17',
+        'SM': '89.186.32.0/19',
+        'SN': '41.82.0.0/15',
+        'SO': '197.220.64.0/19',
+        'SR': '186.179.128.0/17',
+        'SS': '105.235.208.0/21',
+        'ST': '197.159.160.0/19',
+        'SV': '168.243.0.0/16',
+        'SX': '190.102.0.0/20',
+        'SY': '5.0.0.0/16',
+        'SZ': '41.84.224.0/19',
+        'TC': '65.255.48.0/20',
+        'TD': '154.68.128.0/19',
+        'TG': '196.168.0.0/14',
+        'TH': '171.96.0.0/13',
+        'TJ': '85.9.128.0/18',
+        'TK': '27.96.24.0/21',
+        'TL': '180.189.160.0/20',
+        'TM': '95.85.96.0/19',
+        'TN': '197.0.0.0/11',
+        'TO': '175.176.144.0/21',
+        'TR': '78.160.0.0/11',
+        'TT': '186.44.0.0/15',
+        'TV': '202.2.96.0/19',
+        'TW': '120.96.0.0/11',
+        'TZ': '156.156.0.0/14',
+        'UA': '93.72.0.0/13',
+        'UG': '154.224.0.0/13',
+        'US': '3.0.0.0/8',
+        'UY': '167.56.0.0/13',
+        'UZ': '82.215.64.0/18',
+        'VA': '212.77.0.0/19',
+        'VC': '24.92.144.0/20',
+        'VE': '186.88.0.0/13',
+        'VG': '172.103.64.0/18',
+        'VI': '146.226.0.0/16',
+        'VN': '14.160.0.0/11',
+        'VU': '202.80.32.0/20',
+        'WF': '117.20.32.0/21',
+        'WS': '202.4.32.0/19',
+        'YE': '134.35.0.0/16',
+        'YT': '41.242.116.0/22',
+        'ZA': '41.0.0.0/11',
+        'ZM': '165.56.0.0/13',
+        'ZW': '41.85.192.0/19',
+    }
+
+    @classmethod
+    def random_ipv4(cls, code_or_block):
+        if len(code_or_block) == 2:
+            block = cls._country_ip_map.get(code_or_block.upper())
+            if not block:
+                return None
+        else:
+            block = code_or_block
+        addr, preflen = block.split('/')
+        addr_min = compat_struct_unpack('!L', socket.inet_aton(addr))[0]
+        addr_max = addr_min | (0xffffffff >> int(preflen))
+        return compat_str(socket.inet_ntoa(
+            compat_struct_pack('!L', random.randint(addr_min, addr_max))))
+
+

 class PerRequestProxyHandler(compat_urllib_request.ProxyHandler):
     def __init__(self, proxies=None):
         # Set default handlers
 class PerRequestProxyHandler(compat_urllib_request.ProxyHandler):
     def __init__(self, proxies=None):
         # Set default handlers


@@ -1886,7 +3661,7 @@ class PerRequestProxyHandler(compat_urllib_request.ProxyHandler):
             setattr(self, '%s_open' % type,
                     lambda r, proxy='__noproxy__', type=type, meth=self.proxy_open:
                         meth(r, proxy, type))
             setattr(self, '%s_open' % type,
                     lambda r, proxy='__noproxy__', type=type, meth=self.proxy_open:
                         meth(r, proxy, type))

-        return compat_urllib_request.ProxyHandler.__init__(self, proxies)
+        compat_urllib_request.ProxyHandler.__init__(self, proxies)

 
     def proxy_open(self, req, proxy, type):
         req_proxy = req.headers.get('Ytdl-request-proxy')
 
     def proxy_open(self, req, proxy, type):
         req_proxy = req.headers.get('Ytdl-request-proxy')


@@ -1896,5 +3671,344 @@ class PerRequestProxyHandler(compat_urllib_request.ProxyHandler):
 
         if proxy == '__noproxy__':
             return None  # No Proxy
 
         if proxy == '__noproxy__':
             return None  # No Proxy

+        if compat_urlparse.urlparse(proxy).scheme.lower() in ('socks', 'socks4', 'socks4a', 'socks5'):
+            req.add_header('Ytdl-socks-proxy', proxy)
+            # youtube-dl's http/https handlers do wrapping the socket with socks
+            return None

         return compat_urllib_request.ProxyHandler.proxy_open(
             self, req, proxy, type)
         return compat_urllib_request.ProxyHandler.proxy_open(
             self, req, proxy, type)

+
+
+# Both long_to_bytes and bytes_to_long are adapted from PyCrypto, which is
+# released into Public Domain
+# https://github.com/dlitz/pycrypto/blob/master/lib/Crypto/Util/number.py#L387
+
+def long_to_bytes(n, blocksize=0):
+    """long_to_bytes(n:long, blocksize:int) : string
+    Convert a long integer to a byte string.
+
+    If optional blocksize is given and greater than zero, pad the front of the
+    byte string with binary zeros so that the length is a multiple of
+    blocksize.
+    """
+    # after much testing, this algorithm was deemed to be the fastest
+    s = b''
+    n = int(n)
+    while n > 0:
+        s = compat_struct_pack('>I', n & 0xffffffff) + s
+        n = n >> 32
+    # strip off leading zeros
+    for i in range(len(s)):
+        if s[i] != b'\000'[0]:
+            break
+    else:
+        # only happens when n == 0
+        s = b'\000'
+        i = 0
+    s = s[i:]
+    # add back some pad bytes.  this could be done more efficiently w.r.t. the
+    # de-padding being done above, but sigh...
+    if blocksize > 0 and len(s) % blocksize:
+        s = (blocksize - len(s) % blocksize) * b'\000' + s
+    return s
+
+
+def bytes_to_long(s):
+    """bytes_to_long(string) : long
+    Convert a byte string to a long integer.
+
+    This is (essentially) the inverse of long_to_bytes().
+    """
+    acc = 0
+    length = len(s)
+    if length % 4:
+        extra = (4 - length % 4)
+        s = b'\000' * extra + s
+        length = length + extra
+    for i in range(0, length, 4):
+        acc = (acc << 32) + compat_struct_unpack('>I', s[i:i + 4])[0]
+    return acc
+
+
+def ohdave_rsa_encrypt(data, exponent, modulus):
+    '''
+    Implement OHDave's RSA algorithm. See http://www.ohdave.com/rsa/
+
+    Input:
+        data: data to encrypt, bytes-like object
+        exponent, modulus: parameter e and N of RSA algorithm, both integer
+    Output: hex string of encrypted data
+
+    Limitation: supports one block encryption only
+    '''
+
+    payload = int(binascii.hexlify(data[::-1]), 16)
+    encrypted = pow(payload, exponent, modulus)
+    return '%x' % encrypted
+
+
+def pkcs1pad(data, length):
+    """
+    Padding input data with PKCS#1 scheme
+
+    @param {int[]} data        input data
+    @param {int}   length      target length
+    @returns {int[]}           padded data
+    """
+    if len(data) > length - 11:
+        raise ValueError('Input data too long for PKCS#1 padding')
+
+    pseudo_random = [random.randint(0, 254) for _ in range(length - len(data) - 3)]
+    return [0, 2] + pseudo_random + [0] + data
+
+
+def encode_base_n(num, n, table=None):
+    FULL_TABLE = '0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ'
+    if not table:
+        table = FULL_TABLE[:n]
+
+    if n > len(table):
+        raise ValueError('base %d exceeds table length %d' % (n, len(table)))
+
+    if num == 0:
+        return table[0]
+
+    ret = ''
+    while num:
+        ret = table[num % n] + ret
+        num = num // n
+    return ret
+
+
+def decode_packed_codes(code):
+    mobj = re.search(PACKED_CODES_RE, code)
+    obfucasted_code, base, count, symbols = mobj.groups()
+    base = int(base)
+    count = int(count)
+    symbols = symbols.split('|')
+    symbol_table = {}
+
+    while count:
+        count -= 1
+        base_n_count = encode_base_n(count, base)
+        symbol_table[base_n_count] = symbols[count] or base_n_count
+
+    return re.sub(
+        r'\b(\w+)\b', lambda mobj: symbol_table[mobj.group(0)],
+        obfucasted_code)
+
+
+def parse_m3u8_attributes(attrib):
+    info = {}
+    for (key, val) in re.findall(r'(?P<key>[A-Z0-9-]+)=(?P<val>"[^"]+"|[^",]+)(?:,|$)', attrib):
+        if val.startswith('"'):
+            val = val[1:-1]
+        info[key] = val
+    return info
+
+
+def urshift(val, n):
+    return val >> n if val >= 0 else (val + 0x100000000) >> n
+
+
+# Based on png2str() written by @gdkchan and improved by @yokrysty
+# 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:]
+
+    if png_data[:8] != b'\x89PNG\x0d\x0a\x1a\x0a' or header[4:8] != b'IHDR':
+        raise IOError('Not a valid PNG file.')
+
+    int_map = {1: '>B', 2: '>H', 4: '>I'}
+    unpack_integer = lambda x: compat_struct_unpack(int_map[len(x)], x)[0]
+
+    chunks = []
+
+    while header:
+        length = unpack_integer(header[:4])
+        header = header[4:]
+
+        chunk_type = header[:4]
+        header = header[4:]
+
+        chunk_data = header[:length]
+        header = header[length:]
+
+        header = header[4:]  # Skip CRC
+
+        chunks.append({
+            'type': chunk_type,
+            'length': length,
+            'data': chunk_data
+        })
+
+    ihdr = chunks[0]['data']
+
+    width = unpack_integer(ihdr[:4])
+    height = unpack_integer(ihdr[4:8])
+
+    idat = b''
+
+    for chunk in chunks:
+        if chunk['type'] == b'IDAT':
+            idat += chunk['data']
+
+    if not idat:
+        raise IOError('Unable to read PNG data.')
+
+    decompressed_data = bytearray(zlib.decompress(idat))
+
+    stride = width * 3
+    pixels = []
+
+    def _get_pixel(idx):
+        x = idx % stride
+        y = idx // stride
+        return pixels[y][x]
+
+    for y in range(height):
+        basePos = y * (1 + stride)
+        filter_type = decompressed_data[basePos]
+
+        current_row = []
+
+        pixels.append(current_row)
+
+        for x in range(stride):
+            color = decompressed_data[1 + basePos + x]
+            basex = y * stride + x
+            left = 0
+            up = 0
+
+            if x > 2:
+                left = _get_pixel(basex - 3)
+            if y > 0:
+                up = _get_pixel(basex - stride)
+
+            if filter_type == 1:  # Sub
+                color = (color + left) & 0xff
+            elif filter_type == 2:  # Up
+                color = (color + up) & 0xff
+            elif filter_type == 3:  # Average
+                color = (color + ((left + up) >> 1)) & 0xff
+            elif filter_type == 4:  # Paeth
+                a = left
+                b = up
+                c = 0
+
+                if x > 2 and y > 0:
+                    c = _get_pixel(basex - stride - 3)
+
+                p = a + b - c
+
+                pa = abs(p - a)
+                pb = abs(p - b)
+                pc = abs(p - c)
+
+                if pa <= pb and pa <= pc:
+                    color = (color + a) & 0xff
+                elif pb <= pc:
+                    color = (color + b) & 0xff
+                else:
+                    color = (color + c) & 0xff
+
+            current_row.append(color)
+
+    return width, height, pixels
+
+
+def write_xattr(path, key, value):
+    # This mess below finds the best xattr tool for the job
+    try:
+        # try the pyxattr module...
+        import xattr
+
+        if hasattr(xattr, 'set'):  # pyxattr
+            # Unicode arguments are not supported in python-pyxattr until
+            # version 0.5.0
+            # 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
+                raise XAttrUnavailableError(
+                    'python-pyxattr is detected but is too old. '
+                    'youtube-dl requires %s or above while your version is %s. '
+                    'Falling back to other xattr implementations' % (
+                        pyxattr_required_version, xattr.__version__))
+
+            setxattr = xattr.set
+        else:  # xattr
+            setxattr = xattr.setxattr
+
+        try:
+            setxattr(path, key, value)
+        except EnvironmentError as e:
+            raise XAttrMetadataError(e.errno, e.strerror)
+
+    except ImportError:
+        if compat_os_name == 'nt':
+            # Write xattrs to NTFS Alternate Data Streams:
+            # http://en.wikipedia.org/wiki/NTFS#Alternate_data_streams_.28ADS.29
+            assert ':' not in key
+            assert os.path.exists(path)
+
+            ads_fn = path + ':' + key
+            try:
+                with open(ads_fn, 'wb') as f:
+                    f.write(value)
+            except EnvironmentError as e:
+                raise XAttrMetadataError(e.errno, e.strerror)
+        else:
+            user_has_setfattr = check_executable('setfattr', ['--version'])
+            user_has_xattr = check_executable('xattr', ['-h'])
+
+            if user_has_setfattr or user_has_xattr:
+
+                value = value.decode('utf-8')
+                if user_has_setfattr:
+                    executable = 'setfattr'
+                    opts = ['-n', key, '-v', value]
+                elif user_has_xattr:
+                    executable = 'xattr'
+                    opts = ['-w', key, value]
+
+                cmd = ([encodeFilename(executable, True)]
+                       + [encodeArgument(o) for o in opts]
+                       + [encodeFilename(path, True)])
+
+                try:
+                    p = subprocess.Popen(
+                        cmd, stdout=subprocess.PIPE, stderr=subprocess.PIPE, stdin=subprocess.PIPE)
+                except EnvironmentError as e:
+                    raise XAttrMetadataError(e.errno, e.strerror)
+                stdout, stderr = p.communicate()
+                stderr = stderr.decode('utf-8', 'replace')
+                if p.returncode != 0:
+                    raise XAttrMetadataError(p.returncode, stderr)
+
+            else:
+                # On Unix, and can't find pyxattr, setfattr, or xattr.
+                if sys.platform.startswith('linux'):
+                    raise XAttrUnavailableError(
+                        "Couldn't find a tool to set the xattrs. "
+                        "Install either the python 'pyxattr' or 'xattr' "
+                        "modules, or the GNU 'attr' package "
+                        "(which contains the 'setfattr' tool).")
+                else:
+                    raise XAttrUnavailableError(
+                        "Couldn't find a tool to set the xattrs. "
+                        "Install either the python 'xattr' module, "
+                        "or the 'xattr' binary.")
+
+
+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_date.year),
+        month_field: str(random_date.month),
+        day_field: str(random_date.day),
+    }