safety-screen/Nominatim
1
0
Fork 0
forked from hans/Nominatim
Code Pull Requests Activity

replace BreakType enum with simple char constants

Browse Source
This commit is contained in:
Sarah Hoffmann
2025-02-21 09:57:48 +01:00
parent 9bf1428d81
commit 4577669213
8 changed files with 150 additions and 147 deletions
Download Patch File Download Diff File Expand all files Collapse all files

149
test/python/api/search/test_token_assignment.py
View File

@@ -9,7 +9,8 @@ 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.query import QueryStruct, Phrase, PhraseType, TokenType, TokenRange, Token
import nominatim_api.search.query as qmod
from nominatim_api.search.token_assignment import yield_token_assignments, TokenAssignment, PENALTY_TOKENCHANGE
class MyToken(Token):
@@ -24,7 +25,7 @@ def make_query(*args):
for btype, ptype, _ in args[1:]:
q.add_node(btype, ptype)
q.add_node(BreakType.END, PhraseType.NONE)
q.add_node(qmod.BREAK_END, PhraseType.NONE)
for start, t in enumerate(args):
for end, ttype in t[2]:
@@ -44,13 +45,13 @@ def check_assignments(actual, *expected):
def test_query_with_missing_tokens():
q = QueryStruct([Phrase(PhraseType.NONE, '')])
q.add_node(BreakType.END, PhraseType.NONE)
q.add_node(qmod.BREAK_END, PhraseType.NONE)
assert list(yield_token_assignments(q)) == []
def test_one_word_query():
q = make_query((BreakType.START, PhraseType.NONE,
q = make_query((qmod.BREAK_START, PhraseType.NONE,
[(1, TokenType.PARTIAL),
(1, TokenType.WORD),
(1, TokenType.HOUSENUMBER)]))
@@ -60,7 +61,7 @@ def test_one_word_query():
def test_single_postcode():
q = make_query((BreakType.START, PhraseType.NONE,
q = make_query((qmod.BREAK_START, PhraseType.NONE,
[(1, TokenType.POSTCODE)]))
res = list(yield_token_assignments(q))
@@ -68,7 +69,7 @@ def test_single_postcode():
def test_single_country_name():
q = make_query((BreakType.START, PhraseType.NONE,
q = make_query((qmod.BREAK_START, PhraseType.NONE,
[(1, TokenType.COUNTRY)]))
res = list(yield_token_assignments(q))
@@ -76,7 +77,7 @@ def test_single_country_name():
def test_single_word_poi_search():
q = make_query((BreakType.START, PhraseType.NONE,
q = make_query((qmod.BREAK_START, PhraseType.NONE,
[(1, TokenType.NEAR_ITEM),
(1, TokenType.QUALIFIER)]))
@@ -84,9 +85,9 @@ def test_single_word_poi_search():
assert res == [TokenAssignment(near_item=TokenRange(0, 1))]
@pytest.mark.parametrize('btype', [BreakType.WORD, BreakType.PART, BreakType.TOKEN])
@pytest.mark.parametrize('btype', [qmod.BREAK_WORD, qmod.BREAK_PART, qmod.BREAK_TOKEN])
def test_multiple_simple_words(btype):
q = make_query((BreakType.START, PhraseType.NONE, [(1, TokenType.PARTIAL)]),
q = make_query((qmod.BREAK_START, PhraseType.NONE, [(1, TokenType.PARTIAL)]),
(btype, PhraseType.NONE, [(2, TokenType.PARTIAL)]),
(btype, PhraseType.NONE, [(3, TokenType.PARTIAL)]))
@@ -106,8 +107,8 @@ def test_multiple_simple_words(btype):
def test_multiple_words_respect_phrase_break():
q = make_query((BreakType.START, PhraseType.NONE, [(1, TokenType.PARTIAL)]),
(BreakType.PHRASE, PhraseType.NONE, [(2, TokenType.PARTIAL)]))
q = make_query((qmod.BREAK_START, PhraseType.NONE, [(1, TokenType.PARTIAL)]),
(qmod.BREAK_PHRASE, PhraseType.NONE, [(2, TokenType.PARTIAL)]))
check_assignments(yield_token_assignments(q),
TokenAssignment(name=TokenRange(0, 1),
@@ -117,8 +118,8 @@ def test_multiple_words_respect_phrase_break():
def test_housenumber_and_street():
q = make_query((BreakType.START, PhraseType.NONE, [(1, TokenType.HOUSENUMBER)]),
(BreakType.PHRASE, PhraseType.NONE, [(2, TokenType.PARTIAL)]))
q = make_query((qmod.BREAK_START, PhraseType.NONE, [(1, TokenType.HOUSENUMBER)]),
(qmod.BREAK_PHRASE, PhraseType.NONE, [(2, TokenType.PARTIAL)]))
check_assignments(yield_token_assignments(q),
TokenAssignment(name=TokenRange(1, 2),
@@ -128,8 +129,8 @@ def test_housenumber_and_street():
def test_housenumber_and_street_backwards():
q = make_query((BreakType.START, PhraseType.NONE, [(1, TokenType.PARTIAL)]),
(BreakType.PHRASE, PhraseType.NONE, [(2, TokenType.HOUSENUMBER)]))
q = make_query((qmod.BREAK_START, PhraseType.NONE, [(1, TokenType.PARTIAL)]),
(qmod.BREAK_PHRASE, PhraseType.NONE, [(2, TokenType.HOUSENUMBER)]))
check_assignments(yield_token_assignments(q),
TokenAssignment(name=TokenRange(0, 1),
@@ -139,10 +140,10 @@ def test_housenumber_and_street_backwards():
def test_housenumber_and_postcode():
q = make_query((BreakType.START, PhraseType.NONE, [(1, TokenType.PARTIAL)]),
(BreakType.WORD, PhraseType.NONE, [(2, TokenType.HOUSENUMBER)]),
(BreakType.WORD, PhraseType.NONE, [(3, TokenType.PARTIAL)]),
(BreakType.WORD, PhraseType.NONE, [(4, TokenType.POSTCODE)]))
q = make_query((qmod.BREAK_START, PhraseType.NONE, [(1, TokenType.PARTIAL)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(2, TokenType.HOUSENUMBER)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(3, TokenType.PARTIAL)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(4, TokenType.POSTCODE)]))
check_assignments(yield_token_assignments(q),
TokenAssignment(penalty=pytest.approx(0.3),
@@ -156,10 +157,10 @@ def test_housenumber_and_postcode():
postcode=TokenRange(3, 4)))
def test_postcode_and_housenumber():
q = make_query((BreakType.START, PhraseType.NONE, [(1, TokenType.PARTIAL)]),
(BreakType.WORD, PhraseType.NONE, [(2, TokenType.POSTCODE)]),
(BreakType.WORD, PhraseType.NONE, [(3, TokenType.PARTIAL)]),
(BreakType.WORD, PhraseType.NONE, [(4, TokenType.HOUSENUMBER)]))
q = make_query((qmod.BREAK_START, PhraseType.NONE, [(1, TokenType.PARTIAL)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(2, TokenType.POSTCODE)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(3, TokenType.PARTIAL)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(4, TokenType.HOUSENUMBER)]))
check_assignments(yield_token_assignments(q),
TokenAssignment(penalty=pytest.approx(0.3),
@@ -174,10 +175,10 @@ def test_postcode_and_housenumber():
def test_country_housenumber_postcode():
q = make_query((BreakType.START, PhraseType.NONE, [(1, TokenType.COUNTRY)]),
(BreakType.WORD, PhraseType.NONE, [(2, TokenType.PARTIAL)]),
(BreakType.WORD, PhraseType.NONE, [(3, TokenType.HOUSENUMBER)]),
(BreakType.WORD, PhraseType.NONE, [(4, TokenType.POSTCODE)]))
q = make_query((qmod.BREAK_START, PhraseType.NONE, [(1, TokenType.COUNTRY)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(2, TokenType.PARTIAL)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(3, TokenType.HOUSENUMBER)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(4, TokenType.POSTCODE)]))
check_assignments(yield_token_assignments(q))
@@ -185,27 +186,27 @@ def test_country_housenumber_postcode():
@pytest.mark.parametrize('ttype', [TokenType.POSTCODE, TokenType.COUNTRY,
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)]))
q = make_query((qmod.BREAK_START, PhraseType.NONE, [(1, TokenType.HOUSENUMBER)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(2, ttype)]))
check_assignments(yield_token_assignments(q))
@pytest.mark.parametrize('ttype', [TokenType.POSTCODE, TokenType.HOUSENUMBER, TokenType.COUNTRY])
def test_multiple_special_tokens(ttype):
q = make_query((BreakType.START, PhraseType.NONE, [(1, ttype)]),
(BreakType.PHRASE, PhraseType.NONE, [(2, TokenType.PARTIAL)]),
(BreakType.PHRASE, PhraseType.NONE, [(3, ttype)]))
q = make_query((qmod.BREAK_START, PhraseType.NONE, [(1, ttype)]),
(qmod.BREAK_PHRASE, PhraseType.NONE, [(2, TokenType.PARTIAL)]),
(qmod.BREAK_PHRASE, PhraseType.NONE, [(3, ttype)]))
check_assignments(yield_token_assignments(q))
def test_housenumber_many_phrases():
q = make_query((BreakType.START, PhraseType.NONE, [(1, TokenType.PARTIAL)]),
(BreakType.PHRASE, PhraseType.NONE, [(2, TokenType.PARTIAL)]),
(BreakType.PHRASE, PhraseType.NONE, [(3, TokenType.PARTIAL)]),
(BreakType.PHRASE, PhraseType.NONE, [(4, TokenType.HOUSENUMBER)]),
(BreakType.WORD, PhraseType.NONE, [(5, TokenType.PARTIAL)]))
q = make_query((qmod.BREAK_START, PhraseType.NONE, [(1, TokenType.PARTIAL)]),
(qmod.BREAK_PHRASE, PhraseType.NONE, [(2, TokenType.PARTIAL)]),
(qmod.BREAK_PHRASE, PhraseType.NONE, [(3, TokenType.PARTIAL)]),
(qmod.BREAK_PHRASE, PhraseType.NONE, [(4, TokenType.HOUSENUMBER)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(5, TokenType.PARTIAL)]))
check_assignments(yield_token_assignments(q),
TokenAssignment(penalty=0.1,
@@ -220,8 +221,8 @@ def test_housenumber_many_phrases():
def test_country_at_beginning():
q = make_query((BreakType.START, PhraseType.NONE, [(1, TokenType.COUNTRY)]),
(BreakType.WORD, PhraseType.NONE, [(2, TokenType.PARTIAL)]))
q = make_query((qmod.BREAK_START, PhraseType.NONE, [(1, TokenType.COUNTRY)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(2, TokenType.PARTIAL)]))
check_assignments(yield_token_assignments(q),
TokenAssignment(penalty=0.1, name=TokenRange(1, 2),
@@ -229,8 +230,8 @@ def test_country_at_beginning():
def test_country_at_end():
q = make_query((BreakType.START, PhraseType.NONE, [(1, TokenType.PARTIAL)]),
(BreakType.WORD, PhraseType.NONE, [(2, TokenType.COUNTRY)]))
q = make_query((qmod.BREAK_START, PhraseType.NONE, [(1, TokenType.PARTIAL)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(2, TokenType.COUNTRY)]))
check_assignments(yield_token_assignments(q),
TokenAssignment(penalty=0.1, name=TokenRange(0, 1),
@@ -238,16 +239,16 @@ def test_country_at_end():
def test_country_in_middle():
q = make_query((BreakType.START, PhraseType.NONE, [(1, TokenType.PARTIAL)]),
(BreakType.WORD, PhraseType.NONE, [(2, TokenType.COUNTRY)]),
(BreakType.WORD, PhraseType.NONE, [(3, TokenType.PARTIAL)]))
q = make_query((qmod.BREAK_START, PhraseType.NONE, [(1, TokenType.PARTIAL)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(2, TokenType.COUNTRY)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(3, TokenType.PARTIAL)]))
check_assignments(yield_token_assignments(q))
def test_postcode_with_designation():
q = make_query((BreakType.START, PhraseType.NONE, [(1, TokenType.POSTCODE)]),
(BreakType.PHRASE, PhraseType.NONE, [(2, TokenType.PARTIAL)]))
q = make_query((qmod.BREAK_START, PhraseType.NONE, [(1, TokenType.POSTCODE)]),
(qmod.BREAK_PHRASE, PhraseType.NONE, [(2, TokenType.PARTIAL)]))
check_assignments(yield_token_assignments(q),
TokenAssignment(penalty=0.1, name=TokenRange(1, 2),
@@ -257,8 +258,8 @@ def test_postcode_with_designation():
def test_postcode_with_designation_backwards():
q = make_query((BreakType.START, PhraseType.NONE, [(1, TokenType.PARTIAL)]),
(BreakType.PHRASE, PhraseType.NONE, [(2, TokenType.POSTCODE)]))
q = make_query((qmod.BREAK_START, PhraseType.NONE, [(1, TokenType.PARTIAL)]),
(qmod.BREAK_PHRASE, PhraseType.NONE, [(2, TokenType.POSTCODE)]))
check_assignments(yield_token_assignments(q),
TokenAssignment(name=TokenRange(0, 1),
@@ -268,8 +269,8 @@ def test_postcode_with_designation_backwards():
def test_near_item_at_beginning():
q = make_query((BreakType.START, PhraseType.NONE, [(1, TokenType.NEAR_ITEM)]),
(BreakType.WORD, PhraseType.NONE, [(2, TokenType.PARTIAL)]))
q = make_query((qmod.BREAK_START, PhraseType.NONE, [(1, TokenType.NEAR_ITEM)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(2, TokenType.PARTIAL)]))
check_assignments(yield_token_assignments(q),
TokenAssignment(penalty=0.1, name=TokenRange(1, 2),
@@ -277,8 +278,8 @@ def test_near_item_at_beginning():
def test_near_item_at_end():
q = make_query((BreakType.START, PhraseType.NONE, [(1, TokenType.PARTIAL)]),
(BreakType.WORD, PhraseType.NONE, [(2, TokenType.NEAR_ITEM)]))
q = make_query((qmod.BREAK_START, PhraseType.NONE, [(1, TokenType.PARTIAL)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(2, TokenType.NEAR_ITEM)]))
check_assignments(yield_token_assignments(q),
TokenAssignment(penalty=0.1, name=TokenRange(0, 1),
@@ -286,17 +287,17 @@ def test_near_item_at_end():
def test_near_item_in_middle():
q = make_query((BreakType.START, PhraseType.NONE, [(1, TokenType.PARTIAL)]),
(BreakType.WORD, PhraseType.NONE, [(2, TokenType.NEAR_ITEM)]),
(BreakType.WORD, PhraseType.NONE, [(3, TokenType.PARTIAL)]))
q = make_query((qmod.BREAK_START, PhraseType.NONE, [(1, TokenType.PARTIAL)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(2, TokenType.NEAR_ITEM)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(3, TokenType.PARTIAL)]))
check_assignments(yield_token_assignments(q))
def test_qualifier_at_beginning():
q = make_query((BreakType.START, PhraseType.NONE, [(1, TokenType.QUALIFIER)]),
(BreakType.WORD, PhraseType.NONE, [(2, TokenType.PARTIAL)]),
(BreakType.WORD, PhraseType.NONE, [(3, TokenType.PARTIAL)]))
q = make_query((qmod.BREAK_START, PhraseType.NONE, [(1, TokenType.QUALIFIER)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(2, TokenType.PARTIAL)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(3, TokenType.PARTIAL)]))
check_assignments(yield_token_assignments(q),
@@ -308,11 +309,11 @@ def test_qualifier_at_beginning():
def test_qualifier_after_name():
q = make_query((BreakType.START, PhraseType.NONE, [(1, TokenType.PARTIAL)]),
(BreakType.WORD, PhraseType.NONE, [(2, TokenType.PARTIAL)]),
(BreakType.WORD, PhraseType.NONE, [(3, TokenType.QUALIFIER)]),
(BreakType.WORD, PhraseType.NONE, [(4, TokenType.PARTIAL)]),
(BreakType.WORD, PhraseType.NONE, [(5, TokenType.PARTIAL)]))
q = make_query((qmod.BREAK_START, PhraseType.NONE, [(1, TokenType.PARTIAL)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(2, TokenType.PARTIAL)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(3, TokenType.QUALIFIER)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(4, TokenType.PARTIAL)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(5, TokenType.PARTIAL)]))
check_assignments(yield_token_assignments(q),
@@ -325,27 +326,27 @@ def test_qualifier_after_name():
def test_qualifier_before_housenumber():
q = make_query((BreakType.START, PhraseType.NONE, [(1, TokenType.QUALIFIER)]),
(BreakType.WORD, PhraseType.NONE, [(2, TokenType.HOUSENUMBER)]),
(BreakType.WORD, PhraseType.NONE, [(3, TokenType.PARTIAL)]))
q = make_query((qmod.BREAK_START, PhraseType.NONE, [(1, TokenType.QUALIFIER)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(2, TokenType.HOUSENUMBER)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(3, TokenType.PARTIAL)]))
check_assignments(yield_token_assignments(q))
def test_qualifier_after_housenumber():
q = make_query((BreakType.START, PhraseType.NONE, [(1, TokenType.HOUSENUMBER)]),
(BreakType.WORD, PhraseType.NONE, [(2, TokenType.QUALIFIER)]),
(BreakType.WORD, PhraseType.NONE, [(3, TokenType.PARTIAL)]))
q = make_query((qmod.BREAK_START, PhraseType.NONE, [(1, TokenType.HOUSENUMBER)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(2, TokenType.QUALIFIER)]),
(qmod.BREAK_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)]))
q = make_query((qmod.BREAK_START, PhraseType.NONE, [(1, TokenType.PARTIAL)]),
(qmod.BREAK_PHRASE, PhraseType.NONE, [(2, TokenType.PARTIAL)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(3, TokenType.QUALIFIER)]),
(qmod.BREAK_WORD, PhraseType.NONE, [(4, TokenType.PARTIAL)]),
(qmod.BREAK_PHRASE, PhraseType.NONE, [(5, TokenType.PARTIAL)]))
check_assignments(yield_token_assignments(q))