X-Git-Url: https://git.openstreetmap.org./nominatim.git/blobdiff_plain/3f72ca4bcab2e0f6f0f6db89c7c2659d06858885..c7d80a2cc8cacb7dba95f023c2f480d25f7bf6b1:/test/python/api/search/test_db_search_builder.py?ds=sidebyside diff --git a/test/python/api/search/test_db_search_builder.py b/test/python/api/search/test_db_search_builder.py index e3feff0d..5d984014 100644 --- a/test/python/api/search/test_db_search_builder.py +++ b/test/python/api/search/test_db_search_builder.py @@ -9,11 +9,11 @@ Tests for creating abstract searches from token assignments. """ import pytest -from nominatim.api.search.query import Token, TokenRange, BreakType, PhraseType, TokenType, QueryStruct, Phrase -from nominatim.api.search.db_search_builder import SearchBuilder -from nominatim.api.search.token_assignment import TokenAssignment -from nominatim.api.types import SearchDetails -import nominatim.api.search.db_searches as dbs +from nominatim_api.search.query import Token, TokenRange, BreakType, PhraseType, TokenType, QueryStruct, Phrase +from nominatim_api.search.db_search_builder import SearchBuilder +from nominatim_api.search.token_assignment import TokenAssignment +from nominatim_api.types import SearchDetails +import nominatim_api.search.db_searches as dbs class MyToken(Token): def get_category(self): @@ -31,7 +31,9 @@ def make_query(*args): for end, ttype, tinfo in tlist: for tid, word in tinfo: q.add_token(TokenRange(start, end), ttype, - MyToken(0.5 if ttype == TokenType.PARTIAL else 0.0, tid, 1, word, True)) + MyToken(penalty=0.5 if ttype == TokenType.PARTIAL else 0.0, + token=tid, count=1, addr_count=1, + lookup_word=word, is_indexed=True)) return q @@ -313,6 +315,64 @@ def test_name_only_search_with_category(): assert search.qualifiers.values == [('foo', 'bar')] +def test_name_with_near_item_search_with_category_mismatch(): + q = make_query([(1, TokenType.NEAR_ITEM, [(88, 'g')])], + [(2, TokenType.PARTIAL, [(1, 'a')]), + (2, TokenType.WORD, [(100, 'a')])]) + builder = SearchBuilder(q, SearchDetails.from_kwargs({'categories': [('foo', 'bar')]})) + + searches = list(builder.build(TokenAssignment(name=TokenRange(1, 2), + near_item=TokenRange(0, 1)))) + + assert len(searches) == 0 + + +def test_name_with_near_item_search_with_category_match(): + q = make_query([(1, TokenType.NEAR_ITEM, [(88, 'g')])], + [(2, TokenType.PARTIAL, [(1, 'a')]), + (2, TokenType.WORD, [(100, 'a')])]) + builder = SearchBuilder(q, SearchDetails.from_kwargs({'categories': [('foo', 'bar'), + ('this', 'that')]})) + + searches = list(builder.build(TokenAssignment(name=TokenRange(1, 2), + near_item=TokenRange(0, 1)))) + + assert len(searches) == 1 + search = searches[0] + + assert isinstance(search, dbs.NearSearch) + assert isinstance(search.search, dbs.PlaceSearch) + + +def test_name_with_qualifier_search_with_category_mismatch(): + q = make_query([(1, TokenType.QUALIFIER, [(88, 'g')])], + [(2, TokenType.PARTIAL, [(1, 'a')]), + (2, TokenType.WORD, [(100, 'a')])]) + builder = SearchBuilder(q, SearchDetails.from_kwargs({'categories': [('foo', 'bar')]})) + + searches = list(builder.build(TokenAssignment(name=TokenRange(1, 2), + qualifier=TokenRange(0, 1)))) + + assert len(searches) == 0 + + +def test_name_with_qualifier_search_with_category_match(): + q = make_query([(1, TokenType.QUALIFIER, [(88, 'g')])], + [(2, TokenType.PARTIAL, [(1, 'a')]), + (2, TokenType.WORD, [(100, 'a')])]) + builder = SearchBuilder(q, SearchDetails.from_kwargs({'categories': [('foo', 'bar'), + ('this', 'that')]})) + + searches = list(builder.build(TokenAssignment(name=TokenRange(1, 2), + qualifier=TokenRange(0, 1)))) + + assert len(searches) == 1 + search = searches[0] + + assert isinstance(search, dbs.PlaceSearch) + assert search.qualifiers.values == [('this', 'that')] + + def test_name_only_search_with_countries(): q = make_query([(1, TokenType.PARTIAL, [(1, 'a')]), (1, TokenType.WORD, [(100, 'a')])]) @@ -337,14 +397,14 @@ def make_counted_searches(name_part, name_full, address_part, address_full, q.add_node(BreakType.END, PhraseType.NONE) q.add_token(TokenRange(0, 1), TokenType.PARTIAL, - MyToken(0.5, 1, name_part, 'name_part', True)) + MyToken(0.5, 1, name_part, 1, 'name_part', True)) q.add_token(TokenRange(0, 1), TokenType.WORD, - MyToken(0, 101, name_full, 'name_full', True)) + MyToken(0, 101, name_full, 1, 'name_full', True)) for i in range(num_address_parts): q.add_token(TokenRange(i + 1, i + 2), TokenType.PARTIAL, - MyToken(0.5, 2, address_part, 'address_part', True)) + MyToken(0.5, 2, address_part, 1, 'address_part', True)) q.add_token(TokenRange(i + 1, i + 2), TokenType.WORD, - MyToken(0, 102, address_full, 'address_full', True)) + MyToken(0, 102, address_full, 1, 'address_full', True)) builder = SearchBuilder(q, SearchDetails()) @@ -362,8 +422,8 @@ def test_infrequent_partials_in_name(): assert len(search.lookups) == 2 assert len(search.rankings) == 2 - assert set((l.column, l.lookup_type) for l in search.lookups) == \ - {('name_vector', 'lookup_all'), ('nameaddress_vector', 'restrict')} + assert set((l.column, l.lookup_type.__name__) for l in search.lookups) == \ + {('name_vector', 'LookupAll'), ('nameaddress_vector', 'Restrict')} def test_frequent_partials_in_name_and_address(): @@ -374,10 +434,10 @@ def test_frequent_partials_in_name_and_address(): assert all(isinstance(s, dbs.PlaceSearch) for s in searches) searches.sort(key=lambda s: s.penalty) - assert set((l.column, l.lookup_type) for l in searches[0].lookups) == \ - {('name_vector', 'lookup_any'), ('nameaddress_vector', 'restrict')} - assert set((l.column, l.lookup_type) for l in searches[1].lookups) == \ - {('nameaddress_vector', 'lookup_all'), ('name_vector', 'lookup_all')} + assert set((l.column, l.lookup_type.__name__) for l in searches[0].lookups) == \ + {('name_vector', 'LookupAny'), ('nameaddress_vector', 'Restrict')} + assert set((l.column, l.lookup_type.__name__) for l in searches[1].lookups) == \ + {('nameaddress_vector', 'LookupAll'), ('name_vector', 'LookupAll')} def test_too_frequent_partials_in_name_and_address(): @@ -388,5 +448,5 @@ def test_too_frequent_partials_in_name_and_address(): assert all(isinstance(s, dbs.PlaceSearch) for s in searches) searches.sort(key=lambda s: s.penalty) - assert set((l.column, l.lookup_type) for l in searches[0].lookups) == \ - {('name_vector', 'lookup_any'), ('nameaddress_vector', 'restrict')} + assert set((l.column, l.lookup_type.__name__) for l in searches[0].lookups) == \ + {('name_vector', 'LookupAny'), ('nameaddress_vector', 'Restrict')}