expected_count = sum(t.count for t in hnrs)
partials = {t.token: t.addr_count for trange in address
- for t in self.query.get_partials_list(trange)}
+ for t in self.query.get_partials_list(trange)
+ if t.is_indexed}
+
+ if not partials:
+ # can happen when none of the partials is indexed
+ return
if expected_count < 8000:
sdata.lookups.append(dbf.FieldLookup('nameaddress_vector',
yield penalty, exp_count, dbf.lookup_by_names(list(name_partials.keys()), addr_tokens)
return
- addr_count = min(t.addr_count for t in addr_partials) if addr_partials else 30000
+ addr_count = min(t.addr_count for t in addr_partials) if addr_partials else 50000
# Partial term to frequent. Try looking up by rare full names first.
name_fulls = self.query.get_tokens(name, TokenType.WORD)
if name_fulls:
if partials_indexed:
penalty += 1.2 * sum(t.penalty for t in addr_partials if not t.is_indexed)
- if fulls_count < 50000 or addr_count < 30000:
+ if fulls_count < 80000 or addr_count < 50000:
yield penalty,fulls_count / (2**len(addr_tokens)), \
self.get_full_name_ranking(name_fulls, addr_partials,
fulls_count > 30000 / max(1, len(addr_tokens)))
use_lookup: bool) -> List[dbf.FieldLookup]:
""" Create a ranking expression with full name terms and
additional address lookup. When 'use_lookup' is true, then
- address lookups will use the index, when the occurences are not
+ address lookups will use the index, when the occurrences are not
too many.
"""
# At this point drop unindexed partials from the address.
# This might yield wrong results, nothing we can do about that.
if use_lookup:
addr_restrict_tokens = []
- addr_lookup_tokens = []
- for t in addr_partials:
- if t.is_indexed:
- if t.addr_count > 20000:
- addr_restrict_tokens.append(t.token)
- else:
- addr_lookup_tokens.append(t.token)
+ addr_lookup_tokens = [t.token for t in addr_partials if t.is_indexed]
else:
addr_restrict_tokens = [t.token for t in addr_partials if t.is_indexed]
addr_lookup_tokens = []