def checkIndex(self, maxID):
"""Compute a list of all UIDs less than or equal to maxID that
are not in the db.
"""
# I had originally suspected that the performance hit of
# returning a huge list in the case of empty indexes would be
# substantial, but testing with a 120,000 msg index indicates
# that performance is fine and that the space overhead is quite
# reasonable. If that were not the case, this could be optimized
# by calculating the maximum document ID in the index and only
# scanning up to the minimum of maxID and the max ID in the
# index, assuming that were using the same document IDs in the
# index as in atop.
missingUIDs = []
for uid in xrange(maxID + 1):
term = makePairForWrite("UID", str(uid))
if not self.db.term_exists(term):
missingUIDs.append(uid)
return missingUIDs
def get_documents(self, uid):
""" return a list of remapped UIDs corresponding to the actual UID given
"""
docTerm = makePairForWrite("UID", str(uid))
candidates = self.search(RawQuery(docTerm))
return [int(c["uid"]) for c in candidates]