import email.utils
import errno
import gzip
+import itertools
import io
import json
import locale
def find_xpath_attr(node, xpath, key, val):
""" Find the xpath xpath[@key=val] """
assert re.match(r'^[a-zA-Z]+$', key)
- assert re.match(r'^[a-zA-Z0-9@\s]*$', val)
+ assert re.match(r'^[a-zA-Z0-9@\s:._]*$', val)
expr = xpath + u"[@%s='%s']" % (key, val)
return node.find(expr)
else:
return None
-def fix_xml_all_ampersand(xml_str):
+def fix_xml_ampersands(xml_str):
"""Replace all the '&' by '&' in XML"""
- return xml_str.replace(u'&', u'&')
+ return re.sub(
+ r'&(?!amp;|lt;|gt;|apos;|quot;|#x[0-9a-fA-F]{,4};|#[0-9]{,4};)',
+ u'&',
+ xml_str)
def setproctitle(title):
except OSError:
return False
return exe
+
+
+class PagedList(object):
+ def __init__(self, pagefunc, pagesize):
+ self._pagefunc = pagefunc
+ self._pagesize = pagesize
+
+ def __len__(self):
+ # This is only useful for tests
+ return len(self.getslice())
+
+ def getslice(self, start=0, end=None):
+ res = []
+ for pagenum in itertools.count(start // self._pagesize):
+ firstid = pagenum * self._pagesize
+ nextfirstid = pagenum * self._pagesize + self._pagesize
+ if start >= nextfirstid:
+ continue
+
+ page_results = list(self._pagefunc(pagenum))
+
+ startv = (
+ start % self._pagesize
+ if firstid <= start < nextfirstid
+ else 0)
+
+ endv = (
+ ((end - 1) % self._pagesize) + 1
+ if (end is not None and firstid <= end <= nextfirstid)
+ else None)
+
+ if startv != 0 or endv is not None:
+ page_results = page_results[startv:endv]
+ res.extend(page_results)
+
+ # A little optimization - if current page is not "full", ie. does
+ # not contain page_size videos then we can assume that this page
+ # is the last one - there are no more ids on further pages -
+ # i.e. no need to query again.
+ if len(page_results) + startv < self._pagesize:
+ break
+
+ # If we got the whole page, but the next page is not interesting,
+ # break out early as well
+ if end == nextfirstid:
+ break
+ return res