import zlib
from .compat import (
+ compat_HTMLParser,
compat_basestring,
compat_chr,
compat_etree_fromstring,
compat_urllib_parse_urlparse,
compat_urllib_request,
compat_urlparse,
+ compat_xpath,
shlex_quote,
)
def find_xpath_attr(node, xpath, key, val=None):
""" Find the xpath xpath[@key=val] """
assert re.match(r'^[a-zA-Z_-]+$', key)
- if val:
- assert re.match(r'^[a-zA-Z0-9@\s:._-]*$', val)
expr = xpath + ('[@%s]' % key if val is None else "[@%s='%s']" % (key, val))
return node.find(expr)
else:
def find_xpath_attr(node, xpath, key, val=None):
- # 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):
+ 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:
def xpath_element(node, xpath, name=None, fatal=False, default=NO_DEFAULT):
def _find_xpath(xpath):
- if sys.version_info < (2, 7): # Crazy 2.6
- xpath = xpath.encode('ascii')
- return node.find(xpath)
+ return node.find(compat_xpath(xpath))
if isinstance(xpath, (str, compat_str)):
n = _find_xpath(xpath)
return unescapeHTML(res)
+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="&"
+ 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()
+ parser.feed(html_element)
+ parser.close()
+ return parser.attrs
+
+
def clean_html(html):
"""Clean an HTML snippet into a readable string"""
'%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',
+ '%b %dst %Y %I:%M',
+ '%b %dnd %Y %I:%M',
+ '%b %dth %Y %I:%M',
'%Y %m %d',
'%Y-%m-%d',
'%Y/%m/%d',
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
'Yb': 1000 ** 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
- 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 month_by_name(name):
class OnDemandPagedList(PagedList):
- def __init__(self, pagefunc, pagesize):
+ def __init__(self, pagefunc, pagesize, use_cache=False):
self._pagefunc = pagefunc
self._pagesize = pagesize
+ self._use_cache = use_cache
+ if use_cache:
+ self._cache = {}
def getslice(self, start=0, end=None):
res = []
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
"""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),
try:
struct.pack('!I', 0)
except TypeError:
- # In Python 2.6 (and some 2.7 versions), struct requires a bytes argument
+ # In Python 2.6 and 2.7.x < 2.7.7, struct requires a bytes argument
+ # See https://bugs.python.org/issue19099
def struct_pack(spec, *args):
if isinstance(spec, compat_str):
spec = spec.encode('ascii')
return compat_urllib_parse.urlencode(*args, **kargs).encode('ascii')
+def update_url_query(url, query):
+ parsed_url = compat_urlparse.urlparse(url)
+ qs = compat_parse_qs(parsed_url.query)
+ qs.update(query)
+ qs = encode_dict(qs)
+ return compat_urlparse.urlunparse(parsed_url._replace(
+ query=compat_urllib_parse.urlencode(qs, True)))
+
+
def encode_dict(d, encoding='utf-8'):
def encode(v):
return v.encode(encoding) if isinstance(v, compat_basestring) else v
return {
'3gpp': '3gp',
+ 'smptett+xml': 'tt',
+ 'srt': 'srt',
+ 'ttaf+xml': 'dfxp',
'ttml+xml': 'ttml',
+ 'vtt': 'vtt',
'x-flv': 'flv',
'x-mp4-fragmented': 'mp4',
'x-ms-wmv': 'wmv',
payload = int(binascii.hexlify(data[::-1]), 16)
encrypted = pow(payload, exponent, modulus)
return '%x' % encrypted
+
+
+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(
+ r"}\('(.+)',(\d+),(\d+),'([^']+)'\.split\('\|'\)",
+ 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)