X-Git-Url: https://git.openstreetmap.org./nominatim.git/blobdiff_plain/c01386b5b4960df099815f357434f0b944b97b55..e53eb79923517a44060464d86d0803e8f0a80238:/nominatim/api/search/db_search_builder.py diff --git a/nominatim/api/search/db_search_builder.py b/nominatim/api/search/db_search_builder.py index 2a3153be..e27a24d6 100644 --- a/nominatim/api/search/db_search_builder.py +++ b/nominatim/api/search/db_search_builder.py @@ -5,9 +5,9 @@ # Copyright (C) 2023 by the Nominatim developer community. # For a full list of authors see the git log. """ -Convertion from token assignment to an abstract DB search. +Conversion from token assignment to an abstract DB search. """ -from typing import Optional, List, Tuple, Iterator +from typing import Optional, List, Tuple, Iterator, Dict import heapq from nominatim.api.types import SearchDetails, DataLayer @@ -15,7 +15,7 @@ from nominatim.api.search.query import QueryStruct, Token, TokenType, TokenRange from nominatim.api.search.token_assignment import TokenAssignment import nominatim.api.search.db_search_fields as dbf import nominatim.api.search.db_searches as dbs -from nominatim.api.logging import log +import nominatim.api.search.db_search_lookups as lookups def wrap_near_search(categories: List[Tuple[str, str]], @@ -90,12 +90,14 @@ class SearchBuilder: if sdata is None: return - categories = self.get_search_categories(assignment) + near_items = self.get_near_items(assignment) + if near_items is not None and not near_items: + return # impossible compbination of near items and category parameter if assignment.name is None: - if categories and not sdata.postcodes: - sdata.qualifiers = categories - categories = None + if near_items and not sdata.postcodes: + sdata.qualifiers = near_items + near_items = None builder = self.build_poi_search(sdata) elif assignment.housenumber: hnr_tokens = self.query.get_tokens(assignment.housenumber, @@ -103,18 +105,23 @@ class SearchBuilder: builder = self.build_housenumber_search(sdata, hnr_tokens, assignment.address) else: builder = self.build_special_search(sdata, assignment.address, - bool(categories)) + bool(near_items)) else: builder = self.build_name_search(sdata, assignment.name, assignment.address, - bool(categories)) + bool(near_items)) - if categories: - penalty = min(categories.penalties) - categories.penalties = [p - penalty for p in categories.penalties] + if near_items: + penalty = min(near_items.penalties) + near_items.penalties = [p - penalty for p in near_items.penalties] for search in builder: - yield dbs.NearSearch(penalty, categories, search) + search_penalty = search.penalty + search.penalty = 0.0 + yield dbs.NearSearch(penalty + assignment.penalty + search_penalty, + near_items, search) else: - yield from builder + for search in builder: + search.penalty += assignment.penalty + yield search def build_poi_search(self, sdata: dbf.SearchData) -> Iterator[dbs.AbstractSearch]: @@ -146,7 +153,7 @@ class SearchBuilder: sdata.lookups = [dbf.FieldLookup('nameaddress_vector', [t.token for r in address for t in self.query.get_partials_list(r)], - 'restrict')] + lookups.Restrict)] penalty += 0.2 yield dbs.PostcodeSearch(penalty, sdata) @@ -156,14 +163,28 @@ class SearchBuilder: """ Build a simple address search for special entries where the housenumber is the main name token. """ - partial_tokens: List[int] = [] - for trange in address: - partial_tokens.extend(t.token for t in self.query.get_partials_list(trange)) + sdata.lookups = [dbf.FieldLookup('name_vector', [t.token for t in hnrs], lookups.LookupAny)] + 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)} + + if expected_count < 8000: + sdata.lookups.append(dbf.FieldLookup('nameaddress_vector', + list(partials), lookups.Restrict)) + elif len(partials) != 1 or list(partials.values())[0] < 10000: + sdata.lookups.append(dbf.FieldLookup('nameaddress_vector', + list(partials), lookups.LookupAll)) + else: + addr_fulls = [t.token for t + in self.query.get_tokens(address[0], TokenType.WORD)] + if len(addr_fulls) > 5: + return + sdata.lookups.append( + dbf.FieldLookup('nameaddress_vector', addr_fulls, lookups.LookupAny)) - sdata.lookups = [dbf.FieldLookup('name_vector', [t.token for t in hnrs], 'lookup_any'), - dbf.FieldLookup('nameaddress_vector', partial_tokens, 'lookup_all') - ] - yield dbs.PlaceSearch(0.05, sdata, sum(t.count for t in hnrs)) + sdata.housenumbers = dbf.WeightedStrings([], []) + yield dbs.PlaceSearch(0.05, sdata, expected_count) def build_name_search(self, sdata: dbf.SearchData, @@ -187,72 +208,99 @@ class SearchBuilder: be searched for. This takes into account how frequent the terms are and tries to find a lookup that optimizes index use. """ - penalty = 0.0 # extra penalty currently unused - - name_partials = self.query.get_partials_list(name) - exp_name_count = min(t.count for t in name_partials) - addr_partials = [] - for trange in address: - addr_partials.extend(self.query.get_partials_list(trange)) - addr_tokens = [t.token for t in addr_partials] - partials_indexed = all(t.is_indexed for t in name_partials) \ - and all(t.is_indexed for t in addr_partials) + penalty = 0.0 # extra penalty + name_partials = {t.token: t for t in self.query.get_partials_list(name)} - if (len(name_partials) > 3 or exp_name_count < 1000) and partials_indexed: - # Lookup by name partials, use address partials to restrict results. - lookup = [dbf.FieldLookup('name_vector', - [t.token for t in name_partials], 'lookup_all')] - if addr_tokens: - lookup.append(dbf.FieldLookup('nameaddress_vector', addr_tokens, 'restrict')) - yield penalty, exp_name_count, lookup - return + addr_partials = [t for r in address for t in self.query.get_partials_list(r)] + addr_tokens = list({t.token for t in addr_partials}) - exp_addr_count = min(t.count for t in addr_partials) if addr_partials else exp_name_count - if exp_addr_count < 1000 and partials_indexed: - # Lookup by address partials and restrict results through name terms. - # Give this a small penalty because lookups in the address index are - # more expensive - yield penalty + exp_addr_count/5000, exp_addr_count,\ - [dbf.FieldLookup('name_vector', [t.token for t in name_partials], 'restrict'), - dbf.FieldLookup('nameaddress_vector', addr_tokens, 'lookup_all')] + partials_indexed = all(t.is_indexed for t in name_partials.values()) \ + and all(t.is_indexed for t in addr_partials) + exp_count = min(t.count for t in name_partials.values()) / (2**(len(name_partials) - 1)) + + if (len(name_partials) > 3 or exp_count < 8000) and partials_indexed: + 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 # Partial term to frequent. Try looking up by rare full names first. name_fulls = self.query.get_tokens(name, TokenType.WORD) - rare_names = list(filter(lambda t: t.count < 1000, name_fulls)) - # At this point drop unindexed partials from the address. - # This might yield wrong results, nothing we can do about that. - if not partials_indexed: - addr_tokens = [t.token for t in addr_partials if t.is_indexed] - log().var_dump('before', penalty) - penalty += 1.2 * sum(t.penalty for t in addr_partials if not t.is_indexed) - log().var_dump('after', penalty) - if rare_names: - # Any of the full names applies with all of the partials from the address - lookup = [dbf.FieldLookup('name_vector', [t.token for t in rare_names], 'lookup_any')] - if addr_tokens: - lookup.append(dbf.FieldLookup('nameaddress_vector', addr_tokens, 'restrict')) - yield penalty, sum(t.count for t in rare_names), lookup + if name_fulls: + fulls_count = sum(t.count for t in name_fulls) + if len(name_partials) == 1: + penalty += min(0.5, max(0, (exp_count - 50 * fulls_count) / (2000 * fulls_count))) + 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: + 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))) # To catch remaining results, lookup by name and address # We only do this if there is a reasonable number of results expected. - if min(exp_name_count, exp_addr_count) < 10000: - if all(t.is_indexed for t in name_partials): - lookup = [dbf.FieldLookup('name_vector', - [t.token for t in name_partials], 'lookup_all')] - else: - # we don't have the partials, try with the non-rare names - non_rare_names = [t.token for t in name_fulls if t.count >= 1000] - if not non_rare_names: - return - lookup = [dbf.FieldLookup('name_vector', non_rare_names, 'lookup_any')] - if addr_tokens: - lookup.append(dbf.FieldLookup('nameaddress_vector', addr_tokens, 'lookup_all')) - yield penalty + 0.1 * max(0, 5 - len(name_partials) - len(addr_tokens)),\ - min(exp_name_count, exp_addr_count), lookup - - - def get_name_ranking(self, trange: TokenRange) -> dbf.FieldRanking: + exp_count = exp_count / (2**len(addr_tokens)) if addr_tokens else exp_count + if exp_count < 10000 and addr_count < 20000\ + and all(t.is_indexed for t in name_partials.values()): + penalty += 0.35 * max(1 if name_fulls else 0.1, + 5 - len(name_partials) - len(addr_tokens)) + yield penalty, exp_count,\ + self.get_name_address_ranking(list(name_partials.keys()), addr_partials) + + + def get_name_address_ranking(self, name_tokens: List[int], + addr_partials: List[Token]) -> List[dbf.FieldLookup]: + """ Create a ranking expression looking up by name and address. + """ + lookup = [dbf.FieldLookup('name_vector', name_tokens, lookups.LookupAll)] + + 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) + + if addr_restrict_tokens: + lookup.append(dbf.FieldLookup('nameaddress_vector', + addr_restrict_tokens, lookups.Restrict)) + if addr_lookup_tokens: + lookup.append(dbf.FieldLookup('nameaddress_vector', + addr_lookup_tokens, lookups.LookupAll)) + + return lookup + + + def get_full_name_ranking(self, name_fulls: List[Token], addr_partials: List[Token], + 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 + 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) + else: + addr_restrict_tokens = [t.token for t in addr_partials if t.is_indexed] + addr_lookup_tokens = [] + + return dbf.lookup_by_any_name([t.token for t in name_fulls], + addr_restrict_tokens, addr_lookup_tokens) + + + def get_name_ranking(self, trange: TokenRange, + db_field: str = 'name_vector') -> dbf.FieldRanking: """ Create a ranking expression for a name term in the given range. """ name_fulls = self.query.get_tokens(trange, TokenType.WORD) @@ -261,7 +309,7 @@ class SearchBuilder: # Fallback, sum of penalty for partials name_partials = self.query.get_partials_list(trange) default = sum(t.penalty for t in name_partials) + 0.2 - return dbf.FieldRanking('name_vector', default, ranks) + return dbf.FieldRanking(db_field, default, ranks) def get_addr_ranking(self, trange: TokenRange) -> dbf.FieldRanking: @@ -319,11 +367,9 @@ class SearchBuilder: sdata = dbf.SearchData() sdata.penalty = assignment.penalty if assignment.country: - tokens = self.query.get_tokens(assignment.country, TokenType.COUNTRY) - if self.details.countries: - tokens = [t for t in tokens if t.lookup_word in self.details.countries] - if not tokens: - return None + tokens = self.get_country_tokens(assignment.country) + if not tokens: + return None sdata.set_strings('countries', tokens) elif self.details.countries: sdata.countries = dbf.WeightedStrings(self.details.countries, @@ -337,34 +383,70 @@ class SearchBuilder: self.query.get_tokens(assignment.postcode, TokenType.POSTCODE)) if assignment.qualifier: - sdata.set_qualifiers(self.query.get_tokens(assignment.qualifier, - TokenType.QUALIFIER)) + tokens = self.get_qualifier_tokens(assignment.qualifier) + if not tokens: + return None + sdata.set_qualifiers(tokens) + elif self.details.categories: + sdata.qualifiers = dbf.WeightedCategories(self.details.categories, + [0.0] * len(self.details.categories)) if assignment.address: - sdata.set_ranking([self.get_addr_ranking(r) for r in assignment.address]) + if not assignment.name and assignment.housenumber: + # housenumber search: the first item needs to be handled like + # a name in ranking or penalties are not comparable with + # normal searches. + sdata.set_ranking([self.get_name_ranking(assignment.address[0], + db_field='nameaddress_vector')] + + [self.get_addr_ranking(r) for r in assignment.address[1:]]) + else: + sdata.set_ranking([self.get_addr_ranking(r) for r in assignment.address]) else: sdata.rankings = [] return sdata - def get_search_categories(self, - assignment: TokenAssignment) -> Optional[dbf.WeightedCategories]: - """ Collect tokens for category search or use the categories - requested per parameter. - Returns None if no category search is requested. + def get_country_tokens(self, trange: TokenRange) -> List[Token]: + """ Return the list of country tokens for the given range, + optionally filtered by the country list from the details + parameters. """ - if assignment.category: - tokens = [t for t in self.query.get_tokens(assignment.category, - TokenType.CATEGORY) - if not self.details.categories - or t.get_category() in self.details.categories] - return dbf.WeightedCategories([t.get_category() for t in tokens], - [t.penalty for t in tokens]) + tokens = self.query.get_tokens(trange, TokenType.COUNTRY) + if self.details.countries: + tokens = [t for t in tokens if t.lookup_word in self.details.countries] + + return tokens + + def get_qualifier_tokens(self, trange: TokenRange) -> List[Token]: + """ Return the list of qualifier tokens for the given range, + optionally filtered by the qualifier list from the details + parameters. + """ + tokens = self.query.get_tokens(trange, TokenType.QUALIFIER) if self.details.categories: - return dbf.WeightedCategories(self.details.categories, - [0.0] * len(self.details.categories)) + tokens = [t for t in tokens if t.get_category() in self.details.categories] + + return tokens + + + def get_near_items(self, assignment: TokenAssignment) -> Optional[dbf.WeightedCategories]: + """ Collect tokens for near items search or use the categories + requested per parameter. + Returns None if no category search is requested. + """ + if assignment.near_item: + tokens: Dict[Tuple[str, str], float] = {} + for t in self.query.get_tokens(assignment.near_item, TokenType.NEAR_ITEM): + cat = t.get_category() + # The category of a near search will be that of near_item. + # Thus, if search is restricted to a category parameter, + # the two sets must intersect. + if (not self.details.categories or cat in self.details.categories)\ + and t.penalty < tokens.get(cat, 1000.0): + tokens[cat] = t.penalty + return dbf.WeightedCategories(list(tokens.keys()), list(tokens.values())) return None