X-Git-Url: http://git.bitcoin.ninja/index.cgi?a=blobdiff_plain;ds=inline;f=youtube_dl%2Futils.py;h=edeee1853e30c2b409fe53809ae7912e7966446c;hb=fec2d97ca2f1aa8b64c24b28d8c63cab052e9db4;hp=f339fcb3134c3635cc265e03793fe7f544161c2a;hpb=61ee5aeb7363d16c16cdec0914ade78af1afd8fd;p=youtube-dl
diff --git a/youtube_dl/utils.py b/youtube_dl/utils.py
index f339fcb31..edeee1853 100644
--- a/youtube_dl/utils.py
+++ b/youtube_dl/utils.py
@@ -10,12 +10,14 @@ import ctypes
import datetime
import email.utils
import errno
+import functools
import gzip
import itertools
import io
import json
import locale
import math
+import operator
import os
import pipes
import platform
@@ -31,16 +33,19 @@ import xml.etree.ElementTree
import zlib
from .compat import (
+ compat_basestring,
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,
compat_urllib_parse_urlparse,
compat_urllib_request,
compat_urlparse,
+ shlex_quote,
)
@@ -48,13 +53,19 @@ from .compat import (
compiled_regex_type = type(re.compile(''))
std_headers = {
- 'User-Agent': 'Mozilla/5.0 (X11; Linux x86_64; rv:10.0) Gecko/20100101 Firefox/10.0 (Chrome)',
+ 'User-Agent': 'Mozilla/5.0 (X11; Linux x86_64; rv:10.0) Gecko/20150101 Firefox/20.0 (Chrome)',
'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-Language': 'en-us,en;q=0.5',
}
+
+ENGLISH_MONTH_NAMES = [
+ 'January', 'February', 'March', 'April', 'May', 'June',
+ 'July', 'August', 'September', 'October', 'November', 'December']
+
+
def preferredencoding():
"""Get preferred encoding.
@@ -64,14 +75,14 @@ def preferredencoding():
try:
pref = locale.getpreferredencoding()
'TEST'.encode(pref)
- except:
+ except Exception:
pref = 'UTF-8'
return pref
def write_json_file(obj, fn):
- """ Encode obj as JSON and write it to fn, atomically """
+ """ Encode obj as JSON and write it to fn, atomically if possible """
fn = encodeFilename(fn)
if sys.version_info < (3, 0) and sys.platform != 'win32':
@@ -108,8 +119,15 @@ def write_json_file(obj, fn):
try:
with tf:
json.dump(obj, tf)
+ if sys.platform == 'win32':
+ # Need to remove existing file on Windows, else os.rename raises
+ # WindowsError or FileExistsError.
+ try:
+ os.unlink(fn)
+ except OSError:
+ pass
os.rename(tf.name, fn)
- except:
+ except Exception:
try:
os.remove(tf.name)
except OSError:
@@ -122,13 +140,13 @@ if sys.version_info >= (2, 7):
""" 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 + u"[@%s='%s']" % (key, val)
+ expr = xpath + "[@%s='%s']" % (key, val)
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, unicode):
+ if isinstance(xpath, compat_str):
xpath = xpath.encode('ascii')
for f in node.findall(xpath):
@@ -138,6 +156,8 @@ else:
# On python2.6 the xml.etree.ElementTree.Element methods don't support
# the namespace parameter
+
+
def xpath_with_ns(path, ns_map):
components = [c.split(':') for c in path.split('/')]
replaced = []
@@ -155,7 +175,7 @@ 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)
@@ -194,6 +214,10 @@ 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)
@@ -228,15 +252,12 @@ def sanitize_open(filename, open_mode):
raise
# In case of error, try to remove win32 forbidden chars
- alt_filename = os.path.join(
- re.sub('[/<>:"\\|\\\\?\\*]', '#', path_part)
- for path_part in os.path.split(filename)
- )
+ alt_filename = sanitize_path(filename)
if alt_filename == filename:
raise
else:
# An exception here should be caught in the caller
- stream = open(encodeFilename(filename), open_mode)
+ stream = open(encodeFilename(alt_filename), open_mode)
return (stream, alt_filename)
@@ -248,6 +269,7 @@ def timeconvert(timestr):
timestamp = email.utils.mktime_tz(timetuple)
return timestamp
+
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.
@@ -268,6 +290,8 @@ 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:
@@ -276,10 +300,39 @@ def sanitize_filename(s, restricted=False, is_id=False):
# Common case of "Foreign band name - English song title"
if restricted and result.startswith('-_'):
result = result[2:]
+ if result.startswith('-'):
+ result = '_' + result[len('-'):]
+ result = result.lstrip('.')
if not result:
result = '_'
return result
+
+def sanitize_path(s):
+ """Sanitizes and normalizes path on Windows"""
+ if sys.platform != 'win32':
+ return s
+ drive_or_unc, _ = os.path.splitdrive(s)
+ if sys.version_info < (2, 7) and not drive_or_unc:
+ drive_or_unc, _ = os.path.splitunc(s)
+ norm_path = os.path.normpath(remove_start(s, drive_or_unc)).split(os.path.sep)
+ if drive_or_unc:
+ norm_path.pop(0)
+ sanitized_path = [
+ path_part if path_part in ['.', '..'] else re.sub('(?:[/<>:"\\|\\\\?\\*]|\.$)', '#', path_part)
+ for path_part in norm_path]
+ if drive_or_unc:
+ sanitized_path.insert(0, drive_or_unc + os.path.sep)
+ return os.path.join(*sanitized_path)
+
+
+def sanitize_url_path_consecutive_slashes(url):
+ """Collapses consecutive slashes in URLs' path"""
+ parsed_url = list(compat_urlparse.urlparse(url))
+ parsed_url[2] = re.sub(r'/{2,}', '/', parsed_url[2])
+ return compat_urlparse.urlunparse(parsed_url)
+
+
def orderedSet(iterable):
""" Remove all duplicates from the input iterable """
res = []
@@ -295,7 +348,7 @@ def _htmlentity_transform(entity):
if entity in compat_html_entities.name2codepoint:
return compat_chr(compat_html_entities.name2codepoint[entity])
- mobj = re.match(r'#(x?[0-9]+)', entity)
+ mobj = re.match(r'#(x[0-9a-fA-F]+|[0-9]+)', entity)
if mobj is not None:
numstr = mobj.group(1)
if numstr.startswith('x'):
@@ -350,7 +403,7 @@ 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)
@@ -364,6 +417,7 @@ def decodeOption(optval):
assert isinstance(optval, compat_str)
return optval
+
def formatSeconds(secs):
if secs > 3600:
return '%d:%02d:%02d' % (secs // 3600, (secs % 3600) // 60, secs % 60)
@@ -373,48 +427,45 @@ 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)
+
+
+def bug_reports_message():
+ if ytdl_is_updateable():
+ update_cmd = 'type youtube-dl -U to update'
+ else:
+ update_cmd = 'see https://yt-dl.org/update on how to update'
+ msg = '; please report this issue on https://yt-dl.org/bug .'
+ msg += ' Make sure you are using the latest version; %s.' % update_cmd
+ msg += ' Be sure to call youtube-dl with the --verbose flag and include its complete output.'
+ return msg
+
class ExtractorError(Exception):
"""Error during info extraction."""
+
def __init__(self, msg, tb=None, expected=False, cause=None, video_id=None):
""" tb, if given, is the original traceback (so that it can be printed out).
If expected is set, this is a normal error message and most likely not a bug in youtube-dl.
@@ -427,7 +478,7 @@ class ExtractorError(Exception):
if cause:
msg += ' (caused by %r)' % cause
if not expected:
- msg = msg + '; please report this issue on https://yt-dl.org/bug . Be sure to call youtube-dl with the --verbose flag and include its complete output. Make sure you are using the latest version; type youtube-dl -U to update.'
+ msg += bug_reports_message()
super(ExtractorError, self).__init__(msg)
self.traceback = tb
@@ -441,6 +492,13 @@ 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
@@ -453,6 +511,7 @@ class DownloadError(Exception):
configured to continue on errors. They will contain the appropriate
error message.
"""
+
def __init__(self, msg, exc_info=None):
""" exc_info, if given, is the original exception that caused the trouble (as returned by sys.exc_info()). """
super(DownloadError, self).__init__(msg)
@@ -474,9 +533,11 @@ class PostProcessingError(Exception):
This exception may be raised by PostProcessor's .run() method to
indicate an error in the postprocessing task.
"""
+
def __init__(self, msg):
self.msg = msg
+
class MaxDownloadsReached(Exception):
""" --max-downloads limit has been reached. """
pass
@@ -506,6 +567,29 @@ class ContentTooShortError(Exception):
self.downloaded = downloaded
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.
@@ -524,6 +608,15 @@ 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:
@@ -541,17 +634,14 @@ 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
@@ -593,42 +683,63 @@ class YoutubeDLHandler(compat_urllib_request.HTTPHandler):
https_response = http_response
-def parse_iso8601(date_str, delimiter='T'):
+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):
+ kwargs = {}
+ 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
+ return self.do_open(functools.partial(
+ _create_http_connection, self, self._https_conn_class, True),
+ req, **kwargs)
+
+
+def parse_iso8601(date_str, delimiter='T', timezone=None):
""" Return a UNIX timestamp from the given date """
if date_str is None:
return None
- m = re.search(
- r'(\.[0-9]+)?(?:Z$| ?(?P\+|-)(?P[0-9]{2}):?(?P[0-9]{2})$)',
- date_str)
- if not m:
- timezone = datetime.timedelta()
- else:
- date_str = date_str[:-len(m.group(0))]
- if not m.group('sign'):
+ if timezone is None:
+ m = re.search(
+ r'(\.[0-9]+)?(?:Z$| ?(?P\+|-)(?P[0-9]{2}):?(?P[0-9]{2})$)',
+ date_str)
+ if not m:
timezone = datetime.timedelta()
else:
- sign = 1 if m.group('sign') == '+' else -1
- timezone = datetime.timedelta(
- hours=sign * int(m.group('hours')),
- minutes=sign * int(m.group('minutes')))
+ date_str = date_str[:-len(m.group(0))]
+ if not m.group('sign'):
+ timezone = datetime.timedelta()
+ else:
+ sign = 1 if m.group('sign') == '+' else -1
+ timezone = datetime.timedelta(
+ hours=sign * int(m.group('hours')),
+ minutes=sign * int(m.group('minutes')))
date_format = '%Y-%m-%d{0}%H:%M:%S'.format(delimiter)
dt = datetime.datetime.strptime(date_str, date_format) - timezone
return calendar.timegm(dt.timetuple())
-def unified_strdate(date_str):
+def unified_strdate(date_str, day_first=True):
"""Return a string with the date in the format YYYYMMDD"""
if date_str is None:
return None
-
upload_date = None
- #Replace commas
+ # Replace commas
date_str = date_str.replace(',', ' ')
# %z (UTC offset) is only supported in python>=3.2
- date_str = re.sub(r' ?(\+|-)[0-9]{2}:?[0-9]{2}$', '', date_str)
+ 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)
+
format_expressions = [
'%d %B %Y',
'%d %b %Y',
@@ -637,13 +748,10 @@ 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',
@@ -655,6 +763,22 @@ 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',
+ '%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:
try:
upload_date = datetime.datetime.strptime(date_str, expression).strftime('%Y%m%d')
@@ -666,6 +790,7 @@ def unified_strdate(date_str):
upload_date = datetime.datetime(*timetuple[:6]).strftime('%Y%m%d')
return upload_date
+
def determine_ext(url, default_ext='unknown_video'):
if url is None:
return default_ext
@@ -675,16 +800,20 @@ def determine_ext(url, default_ext='unknown_video'):
else:
return default_ext
+
def subtitles_filename(filename, sub_lang, sub_format):
return filename.rsplit('.', 1)[0] + '.' + sub_lang + '.' + sub_format
+
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