X-Git-Url: http://git.cielonegro.org/gitweb.cgi?a=blobdiff_plain;f=youtube_dl%2Futils.py;h=b8c52af74768053d27b4642173022af3b2c6723d;hb=796df3c631dcf28f7ebdf256e57fcfa3b6463abe;hp=5e9ae7a426277ef29aeb0cb5901eea30f85ed794;hpb=7e26084d09be5a7c0d56bc1f6faa59d222d5eac2;p=youtube-dl.git
diff --git a/youtube_dl/utils.py b/youtube_dl/utils.py
index 5e9ae7a42..b8c52af74 100644
--- a/youtube_dl/utils.py
+++ b/youtube_dl/utils.py
@@ -10,6 +10,7 @@ import ctypes
import datetime
import email.utils
import errno
+import functools
import gzip
import itertools
import io
@@ -34,7 +35,9 @@ from .compat import (
compat_chr,
compat_getenv,
compat_html_entities,
+ compat_http_client,
compat_parse_qs,
+ compat_socket_create_connection,
compat_str,
compat_urllib_error,
compat_urllib_parse,
@@ -166,7 +169,7 @@ def xpath_text(node, xpath, name=None, fatal=False):
xpath = xpath.encode('ascii')
n = node.find(xpath)
- if n is None:
+ if n is None or n.text is None:
if fatal:
name = xpath if name is None else name
raise ExtractorError('Could not find XML element %s' % name)
@@ -205,6 +208,10 @@ def get_element_by_attribute(attribute, value, html):
def clean_html(html):
"""Clean an HTML snippet into a readable string"""
+
+ if html is None: # Convenience for sanitizing descriptions etc.
+ return html
+
# Newline vs
html = html.replace('\n', ' ')
html = re.sub(r'\s*<\s*br\s*/?\s*>\s*', '\n', html)
@@ -280,6 +287,8 @@ def sanitize_filename(s, restricted=False, is_id=False):
return '_'
return char
+ # Handle timestamps
+ s = re.sub(r'[0-9]+(?::[0-9]+)+', lambda m: m.group(0).replace(':', '_'), s)
result = ''.join(map(replace_insane, s))
if not is_id:
while '__' in result:
@@ -363,7 +372,7 @@ def encodeArgument(s):
if not isinstance(s, compat_str):
# Legacy code that uses byte strings
# Uncomment the following line after fixing all post processors
- #assert False, 'Internal error: %r should be of type %r, is %r' % (s, compat_str, type(s))
+ # assert False, 'Internal error: %r should be of type %r, is %r' % (s, compat_str, type(s))
s = s.decode('ascii')
return encodeFilename(s, True)
@@ -387,45 +396,29 @@ def formatSeconds(secs):
return '%d' % secs
-def make_HTTPS_handler(opts_no_check_certificate, **kwargs):
- if sys.version_info < (3, 2):
- import httplib
-
- class HTTPSConnectionV3(httplib.HTTPSConnection):
- def __init__(self, *args, **kwargs):
- httplib.HTTPSConnection.__init__(self, *args, **kwargs)
-
- def connect(self):
- sock = socket.create_connection((self.host, self.port), self.timeout)
- if getattr(self, '_tunnel_host', False):
- self.sock = sock
- self._tunnel()
- try:
- self.sock = ssl.wrap_socket(sock, self.key_file, self.cert_file, ssl_version=ssl.PROTOCOL_TLSv1)
- except ssl.SSLError:
- self.sock = ssl.wrap_socket(sock, self.key_file, self.cert_file, ssl_version=ssl.PROTOCOL_SSLv23)
-
- class HTTPSHandlerV3(compat_urllib_request.HTTPSHandler):
- def https_open(self, req):
- return self.do_open(HTTPSConnectionV3, req)
- return HTTPSHandlerV3(**kwargs)
- elif hasattr(ssl, 'create_default_context'): # Python >= 3.4
- context = ssl.create_default_context(ssl.Purpose.CLIENT_AUTH)
- context.options &= ~ssl.OP_NO_SSLv3 # Allow older, not-as-secure SSLv3
+def make_HTTPS_handler(params, **kwargs):
+ opts_no_check_certificate = params.get('nocheckcertificate', False)
+ if hasattr(ssl, 'create_default_context'): # Python >= 3.4 or 2.7.9
+ context = ssl.create_default_context(ssl.Purpose.SERVER_AUTH)
if opts_no_check_certificate:
+ context.check_hostname = False
context.verify_mode = ssl.CERT_NONE
- return compat_urllib_request.HTTPSHandler(context=context, **kwargs)
+ try:
+ return YoutubeDLHTTPSHandler(params, context=context, **kwargs)
+ except TypeError:
+ # Python 2.7.8
+ # (create_default_context present but HTTPSHandler has no context=)
+ pass
+
+ if sys.version_info < (3, 2):
+ return YoutubeDLHTTPSHandler(params, **kwargs)
else: # Python < 3.4
- context = ssl.SSLContext(ssl.PROTOCOL_SSLv23)
+ context = ssl.SSLContext(ssl.PROTOCOL_TLSv1)
context.verify_mode = (ssl.CERT_NONE
if opts_no_check_certificate
else ssl.CERT_REQUIRED)
context.set_default_verify_paths()
- try:
- context.load_default_certs()
- except AttributeError:
- pass # Python < 3.4
- return compat_urllib_request.HTTPSHandler(context=context, **kwargs)
+ return YoutubeDLHTTPSHandler(params, context=context, **kwargs)
class ExtractorError(Exception):
@@ -463,6 +456,13 @@ class ExtractorError(Exception):
return ''.join(traceback.format_tb(self.traceback))
+class UnsupportedError(ExtractorError):
+ def __init__(self, url):
+ super(UnsupportedError, self).__init__(
+ 'Unsupported URL: %s' % url, expected=True)
+ self.url = url
+
+
class RegexNotFoundError(ExtractorError):
"""Error when a regex didn't match"""
pass
@@ -532,6 +532,28 @@ class ContentTooShortError(Exception):
self.expected = expected
+def _create_http_connection(ydl_handler, http_class, is_https, *args, **kwargs):
+ hc = http_class(*args, **kwargs)
+ source_address = ydl_handler._params.get('source_address')
+ if source_address is not None:
+ 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(
+ (self.host, self.port), self.timeout, sa)
+ if is_https:
+ self.sock = ssl.wrap_socket(
+ sock, self.key_file, self.cert_file,
+ ssl_version=ssl.PROTOCOL_TLSv1)
+ else:
+ self.sock = sock
+ hc.connect = functools.partial(_hc_connect, hc)
+
+ return hc
+
+
class YoutubeDLHandler(compat_urllib_request.HTTPHandler):
"""Handler for HTTP requests and responses.
@@ -550,6 +572,15 @@ class YoutubeDLHandler(compat_urllib_request.HTTPHandler):
public domain.
"""
+ def __init__(self, params, *args, **kwargs):
+ compat_urllib_request.HTTPHandler.__init__(self, *args, **kwargs)
+ self._params = params
+
+ def http_open(self, req):
+ return self.do_open(functools.partial(
+ _create_http_connection, self, compat_http_client.HTTPConnection, False),
+ req)
+
@staticmethod
def deflate(data):
try:
@@ -567,17 +598,14 @@ class YoutubeDLHandler(compat_urllib_request.HTTPHandler):
def http_request(self, req):
for h, v in std_headers.items():
- if h not in req.headers:
+ # 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']
- if 'Youtubedl-user-agent' in req.headers:
- if 'User-agent' in req.headers:
- del req.headers['User-agent']
- req.headers['User-agent'] = req.headers['Youtubedl-user-agent']
- del req.headers['Youtubedl-user-agent']
if sys.version_info < (2, 7) and '#' in req.get_full_url():
# Python 2.6 is brain-dead when it comes to fragments
@@ -619,6 +647,18 @@ class YoutubeDLHandler(compat_urllib_request.HTTPHandler):
https_response = http_response
+class YoutubeDLHTTPSHandler(compat_urllib_request.HTTPSHandler):
+ def __init__(self, params, https_conn_class=None, *args, **kwargs):
+ compat_urllib_request.HTTPSHandler.__init__(self, *args, **kwargs)
+ self._https_conn_class = https_conn_class or compat_http_client.HTTPSConnection
+ self._params = params
+
+ def https_open(self, req):
+ return self.do_open(functools.partial(
+ _create_http_connection, self, self._https_conn_class, True),
+ req)
+
+
def parse_iso8601(date_str, delimiter='T'):
""" Return a UNIX timestamp from the given date """
@@ -644,17 +684,19 @@ def parse_iso8601(date_str, delimiter='T'):
return calendar.timegm(dt.timetuple())
-def unified_strdate(date_str):
+def unified_strdate(date_str, day_first=True):
"""Return a string with the date in the format YYYYMMDD"""
if date_str is None:
return None
-
upload_date = None
# Replace commas
date_str = date_str.replace(',', ' ')
# %z (UTC offset) is only supported in python>=3.2
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)
+
format_expressions = [
'%d %B %Y',
'%d %b %Y',
@@ -663,13 +705,10 @@ def unified_strdate(date_str):
'%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',
- '%d.%m.%Y',
- '%d/%m/%Y',
- '%d/%m/%y',
'%Y/%m/%d %H:%M:%S',
- '%d/%m/%Y %H:%M:%S',
'%Y-%m-%d %H:%M:%S',
'%Y-%m-%d %H:%M:%S.%f',
'%d.%m.%Y %H:%M',
@@ -681,6 +720,20 @@ def unified_strdate(date_str):
'%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 %H:%M:%S',
+ ])
+ else:
+ format_expressions.extend([
+ '%m.%d.%Y',
+ '%m/%d/%Y',
+ '%m/%d/%y',
+ '%m/%d/%Y %H:%M:%S',
+ ])
for expression in format_expressions:
try:
upload_date = datetime.datetime.strptime(date_str, expression).strftime('%Y%m%d')
@@ -712,8 +765,10 @@ def date_from_str(date_str):
Return a datetime object from a string in the format YYYYMMDD or
(now|today)[+-][0-9](day|week|month|year)(s)?"""
today = datetime.date.today()
- if date_str == 'now'or date_str == 'today':
+ if date_str in ('now', 'today'):
return today
+ if date_str == 'yesterday':
+ return today - datetime.timedelta(days=1)
match = re.match('(now|today)(?P[+-])(?P