X-Git-Url: http://git.bitcoin.ninja/index.cgi?a=blobdiff_plain;f=youtube_dl%2Futils.py;h=b84436ed64cc264659dd0a70811e2812612709cc;hb=af03000ad5a445f03fbacb63ce626f8dcfe785c7;hp=b460393bf37106a47304dfa97dd7b1efa37117cd;hpb=6cc622327ff8289f94894f3695ed31014c61cf8e;p=youtube-dl
diff --git a/youtube_dl/utils.py b/youtube_dl/utils.py
index b460393bf..b84436ed6 100644
--- a/youtube_dl/utils.py
+++ b/youtube_dl/utils.py
@@ -184,6 +184,7 @@ DATE_FORMATS_MONTH_FIRST.extend([
])
PACKED_CODES_RE = r"}\('(.+)',(\d+),(\d+),'([^']+)'\.split\('\|'\)"
+JSON_LD_RE = r'(?is)'
def preferredencoding():
@@ -1228,7 +1229,7 @@ def unified_timestamp(date_str, day_first=True):
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]+$', guess):
@@ -1865,6 +1866,13 @@ def strip_or_none(v):
return None if v is None else v.strip()
+def url_or_none(url):
+ if not url or not isinstance(url, compat_str):
+ return None
+ url = url.strip()
+ return url if re.match(r'^(?:[a-zA-Z][\da-zA-Z.+-]*:)?//', url) else None
+
+
def parse_duration(s):
if not isinstance(s, compat_basestring):
return None
@@ -2272,13 +2280,16 @@ def parse_age_limit(s):
return int(m.group('age'))
if s in US_RATINGS:
return US_RATINGS[s]
- return TV_PARENTAL_GUIDELINES.get(s)
+ m = re.match(r'^TV[_-]?(%s)$' % '|'.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(
r'''(?sx)^
- (?:window\.)?(?P[a-zA-Z0-9_.$]+)
+ (?:window\.)?(?P[a-zA-Z0-9_.$]*)
(?:\s*&&\s*(?P=func_name))?
\s*\(\s*(?P.*)\);?
\s*?(?://[^\n]*)*$''',
@@ -2664,6 +2675,7 @@ def dfxp2srt(dfxp_data):
]
_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',
})
@@ -2755,7 +2767,9 @@ def dfxp2srt(dfxp_data):
repeat = False
while True:
for style in dfxp.findall(_x('.//ttml:style')):
- style_id = style.get('id')
+ 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:
@@ -3534,10 +3548,13 @@ class GeoUtils(object):
}
@classmethod
- def random_ipv4(cls, code):
- block = cls._country_ip_map.get(code.upper())
- if not block:
- return None
+ 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))