#
# This file is part of Nominatim. (https://nominatim.org)
#
-# Copyright (C) 2023 by the Nominatim developer community.
+# Copyright (C) 2024 by the Nominatim developer community.
# For a full list of authors see the git log.
"""
Test for creation of token assignments from tokenized queries.
"""
import pytest
-from nominatim.api.search.query import QueryStruct, Phrase, PhraseType, BreakType, TokenType, TokenRange, Token
-from nominatim.api.search.token_assignment import yield_token_assignments, TokenAssignment, PENALTY_TOKENCHANGE
+from nominatim_api.search.query import QueryStruct, Phrase, PhraseType, BreakType, TokenType, TokenRange, Token
+from nominatim_api.search.token_assignment import yield_token_assignments, TokenAssignment, PENALTY_TOKENCHANGE
class MyToken(Token):
def get_category(self):
def make_query(*args):
- q = None
- dummy = MyToken(3.0, 45, 1, 'foo', True)
+ q = QueryStruct([Phrase(args[0][1], '')])
+ dummy = MyToken(penalty=3.0, token=45, count=1, addr_count=1,
+ lookup_word='foo', is_indexed=True)
- for btype, ptype, tlist in args:
- if q is None:
- q = QueryStruct([Phrase(ptype, '')])
- else:
- q.add_node(btype, ptype)
+ for btype, ptype, _ in args[1:]:
+ q.add_node(btype, ptype)
+ q.add_node(BreakType.END, PhraseType.NONE)
- start = len(q.nodes) - 1
- for end, ttype in tlist:
+ for start, t in enumerate(args):
+ for end, ttype in t[2]:
q.add_token(TokenRange(start, end), ttype, dummy)
- q.add_node(BreakType.END, PhraseType.NONE)
-
return q
def test_single_word_poi_search():
q = make_query((BreakType.START, PhraseType.NONE,
- [(1, TokenType.CATEGORY),
+ [(1, TokenType.NEAR_ITEM),
(1, TokenType.QUALIFIER)]))
res = list(yield_token_assignments(q))
- assert res == [TokenAssignment(category=TokenRange(0, 1))]
+ assert res == [TokenAssignment(near_item=TokenRange(0, 1))]
@pytest.mark.parametrize('btype', [BreakType.WORD, BreakType.PART, BreakType.TOKEN])
@pytest.mark.parametrize('ttype', [TokenType.POSTCODE, TokenType.COUNTRY,
- TokenType.CATEGORY, TokenType.QUALIFIER])
+ TokenType.NEAR_ITEM, TokenType.QUALIFIER])
def test_housenumber_with_only_special_terms(ttype):
q = make_query((BreakType.START, PhraseType.NONE, [(1, TokenType.HOUSENUMBER)]),
(BreakType.WORD, PhraseType.NONE, [(2, ttype)]))
address=[TokenRange(0, 1)]))
-def test_category_at_beginning():
- q = make_query((BreakType.START, PhraseType.NONE, [(1, TokenType.CATEGORY)]),
+def test_near_item_at_beginning():
+ q = make_query((BreakType.START, PhraseType.NONE, [(1, TokenType.NEAR_ITEM)]),
(BreakType.WORD, PhraseType.NONE, [(2, TokenType.PARTIAL)]))
check_assignments(yield_token_assignments(q),
TokenAssignment(penalty=0.1, name=TokenRange(1, 2),
- category=TokenRange(0, 1)))
+ near_item=TokenRange(0, 1)))
-def test_category_at_end():
+def test_near_item_at_end():
q = make_query((BreakType.START, PhraseType.NONE, [(1, TokenType.PARTIAL)]),
- (BreakType.WORD, PhraseType.NONE, [(2, TokenType.CATEGORY)]))
+ (BreakType.WORD, PhraseType.NONE, [(2, TokenType.NEAR_ITEM)]))
check_assignments(yield_token_assignments(q),
TokenAssignment(penalty=0.1, name=TokenRange(0, 1),
- category=TokenRange(1, 2)))
+ near_item=TokenRange(1, 2)))
-def test_category_in_middle():
+def test_near_item_in_middle():
q = make_query((BreakType.START, PhraseType.NONE, [(1, TokenType.PARTIAL)]),
- (BreakType.WORD, PhraseType.NONE, [(2, TokenType.CATEGORY)]),
+ (BreakType.WORD, PhraseType.NONE, [(2, TokenType.NEAR_ITEM)]),
(BreakType.WORD, PhraseType.NONE, [(3, TokenType.PARTIAL)]))
check_assignments(yield_token_assignments(q))
(BreakType.WORD, PhraseType.NONE, [(3, TokenType.PARTIAL)]))
check_assignments(yield_token_assignments(q))
+
+
+def test_qualifier_in_middle_of_phrase():
+ q = make_query((BreakType.START, PhraseType.NONE, [(1, TokenType.PARTIAL)]),
+ (BreakType.PHRASE, PhraseType.NONE, [(2, TokenType.PARTIAL)]),
+ (BreakType.WORD, PhraseType.NONE, [(3, TokenType.QUALIFIER)]),
+ (BreakType.WORD, PhraseType.NONE, [(4, TokenType.PARTIAL)]),
+ (BreakType.PHRASE, PhraseType.NONE, [(5, TokenType.PARTIAL)]))
+
+ check_assignments(yield_token_assignments(q))
+