compat_urllib_parse_urlparse,
compat_urllib_request,
compat_urlparse,
+ compat_xpath,
shlex_quote,
)
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 = { }
+ 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
parser.close()
return parser.attrs
+
def clean_html(html):
"""Clean an HTML snippet into a readable string"""
if not for_subprocess and sys.platform == 'win32' and sys.getwindowsversion()[0] >= 5:
return s
+ # Jython assumes filenames are Unicode strings though reported as Python 2.x compatible
+ if sys.platform.startswith('java'):
+ return s
+
return s.encode(get_subprocess_encoding(), 'ignore')
raise OSError('Unlocking file failed: %r' % ctypes.FormatError())
else:
- import fcntl
+ # Some platforms, such as Jython, is missing fcntl
+ try:
+ import fcntl
- def _lock_file(f, exclusive):
- fcntl.flock(f, fcntl.LOCK_EX if exclusive else fcntl.LOCK_SH)
+ def _lock_file(f, exclusive):
+ fcntl.flock(f, fcntl.LOCK_EX if exclusive else fcntl.LOCK_SH)
- def _unlock_file(f):
- fcntl.flock(f, fcntl.LOCK_UN)
+ def _unlock_file(f):
+ fcntl.flock(f, fcntl.LOCK_UN)
+ except ImportError:
+ UNSUPPORTED_MSG = 'file locking is not supported on this platform'
+
+ def _lock_file(f, exclusive):
+ raise IOError(UNSUPPORTED_MSG)
+
+ def _unlock_file(f):
+ raise IOError(UNSUPPORTED_MSG)
class locked_file(object):
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):
def setproctitle(title):
assert isinstance(title, compat_str)
+
+ # ctypes in Jython is not complete
+ # http://bugs.jython.org/issue2148
+ if sys.platform.startswith('java'):
+ return
+
try:
libc = ctypes.cdll.LoadLibrary('libc.so.6')
except OSError:
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