}
+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 = [
'%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 %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',
def get_element_by_class(class_name, html):
- return get_element_by_attribute(
+ """Return the content of the first tag with the specified class in the passed HTML document"""
+ retval = get_elements_by_class(class_name, html)
+ return retval[0] if retval else None
+
+
+def get_element_by_attribute(attribute, value, html, escape_value=True):
+ retval = get_elements_by_attribute(attribute, value, html, escape_value)
+ return retval[0] if retval else None
+
+
+def get_elements_by_class(class_name, html):
+ """Return the content of all tags with the specified class in the passed HTML document as a list"""
+ return get_elements_by_attribute(
'class', r'[^\'"]*\b%s\b[^\'"]*' % re.escape(class_name),
html, escape_value=False)
-def get_element_by_attribute(attribute, value, html, escape_value=True):
+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"""
value = re.escape(value) if escape_value else value
- m = re.search(r'''(?xs)
+ retlist = []
+ for m in re.finditer(r'''(?xs)
<([a-zA-Z0-9:._-]+)
(?:\s+[a-zA-Z0-9:._-]+(?:=[a-zA-Z0-9:._-]*|="[^"]*"|='[^']*'))*?
\s+%s=['"]?%s['"]?
\s*>
(?P<content>.*?)
</\1>
- ''' % (re.escape(attribute), value), html)
+ ''' % (re.escape(attribute), value), html):
+ res = m.group('content')
- if not m:
- return None
- res = m.group('content')
+ if res.startswith('"') or res.startswith("'"):
+ res = res[1:-1]
- if res.startswith('"') or res.startswith("'"):
- res = res[1:-1]
+ retlist.append(unescapeHTML(res))
- return unescapeHTML(res)
+ return retlist
class HTMLAttributeParser(compat_HTMLParser):
if drive_or_unc:
norm_path.pop(0)
sanitized_path = [
- path_part if path_part in ['.', '..'] else re.sub('(?:[/<>:"\\|\\\\?\\*]|[\s.]$)', '#', path_part)
+ path_part if path_part in ['.', '..'] else re.sub(r'(?:[/<>:"\|\\?\*]|[\s.]$)', '#', path_part)
for path_part in norm_path]
if drive_or_unc:
sanitized_path.insert(0, drive_or_unc + os.path.sep)
return 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'))
libc = ctypes.cdll.LoadLibrary('libc.so.6')
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
return path.strip('/').split('/')[-1]
+def base_url(url):
+ return re.match(r'https?://[^?#&]+/', url).group()
+
+
+def urljoin(base, path):
+ if not isinstance(path, compat_str) or not path:
+ return None
+ if re.match(r'^(?:https?:)?//', path):
+ return path
+ 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):
return 'HEAD'
s = s.strip()
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]+)?$', s)
+ m = re.match(r'(?:(?:(?:(?P<days>[0-9]+):)?(?P<hours>[0-9]+):)?(?P<mins>[0-9]+):)?(?P<secs>[0-9]+)(?P<ms>\.[0-9]+)?Z?$', s)
if m:
days, hours, mins, secs, ms = m.groups()
else:
)?
(?:
(?P<secs>[0-9]+)(?P<ms>\.[0-9]+)?\s*s(?:ec(?:ond)?s?)?\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*)$', s)
+ m = re.match(r'(?i)(?:(?P<hours>[0-9.]+)\s*(?:hours?)|(?P<mins>[0-9.]+)\s*(?:mins?\.?|minutes?)\s*)Z?$', s)
if m:
hours, mins = m.groups()
else:
""" 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/rg3/youtube-dl/issues/955#issuecomment-209789656
out, _ = subprocess.Popen(
[encodeArgument(exe)] + args,
+ stdin=subprocess.PIPE,
stdout=subprocess.PIPE, stderr=subprocess.STDOUT).communicate()
except OSError:
return False
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
- elif v.startswith('/*') or v == ',':
+ elif v.startswith('/*') or v.startswith('//') or v == ',':
return ""
if v[0] in ("'", '"'):
'\\x': '\\u00',
}.get(m.group(0), m.group(0)), v[1:-1])
- INTEGER_TABLE = (
- (r'^(0[xX][0-9a-fA-F]+)\s*:?$', 16),
- (r'^(0+[0-7]+)\s*:?$', 8),
- )
-
for regex, base in INTEGER_TABLE:
im = re.match(regex, v)
if im:
return re.sub(r'''(?sx)
"(?:[^"\\]*(?:\\\\|\\['"nurtbfx/\n]))*[^"\\]*"|
'(?:[^'\\]*(?:\\\\|\\['"nurtbfx/\n]))*[^'\\]*'|
- /\*.*?\*/|,(?=\s*[\]}])|
+ {comment}|,(?={skip}[\]}}])|
[a-zA-Z_][.a-zA-Z_0-9]*|
- \b(?:0[xX][0-9a-fA-F]+|0+[0-7]+)(?:\s*:)?|
- [0-9]+(?=\s*:)
- ''', fix_kv, code)
+ \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):
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('strval') is not None or
+ # 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/rg3/youtube-dl/issues/11082).
+ 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'))
- comparison_value = m.group('strval')
+ comparison_value = m.group('strval') or m.group('intval')
else:
try:
comparison_value = int(m.group('intval'))
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)