refactor!: make addrkit a multi-city address normalizer

- remove the events/outages module, ArcGIS adapter, portal config and event statuses - add a configurable city dictionary/parser and make City part of StreetKey - seed: read city/city_name, namespace IDs by city, keep structures distinct - fix reliability: schema v2 with fail-closed load, deterministic ordering, input limits, config validation, atomic DB publish, LLM response caps - add read-only shadow mode comparing the old ES path with addrkit - breaking: registry.db must be re-seeded; config has no portal section
2026-10-27 14:35:51 +07:00
parent a0b4cb2b14
commit 17ff5c591f
38 changed files with 2978 additions and 1164 deletions
@@ -0,0 +1,141 @@
+package address
+
+import "testing"
+
+func testDictionary() *Dictionary {
+	return NewDictionary([]City{
+		{Name: "Новосибирск", Aliases: []string{"нск", "новосиб", "новосибирская область", "нсо"}},
+		{Name: "Бердск"},
+	})
+}
+
+func TestParseCityFromAliases(t *testing.T) {
+	parser := NewParser(testDictionary(), "")
+
+	tests := []struct {
+		raw     string
+		city    string
+		nameKey string
+	}{
+		{"нск, 2-я Станционная улица 1", "Новосибирск", "станционная"},
+		{"НСК Ленина 5", "Новосибирск", "ленина"},
+		{"Новосибирская область, Станционная 1", "Новосибирск", "станционная"},
+		{"Бердск, Ленина 5", "Бердск", "ленина"},
+		{"г. Новосибирск, ул. Ленина 5", "Новосибирск", "ленина"},
+	}
+	for _, tt := range tests {
+		got := parser.Parse(tt.raw)
+		if got.City != tt.city {
+			t.Errorf("%q: City = %q, want %q", tt.raw, got.City, tt.city)
+		}
+		if got.NameKey != tt.nameKey {
+			t.Errorf("%q: NameKey = %q, want %q", tt.raw, got.NameKey, tt.nameKey)
+		}
+	}
+}
+
+func TestParseCityIsPartOfStreetKey(t *testing.T) {
+	parser := NewParser(testDictionary(), "")
+
+	nsk := parser.Parse("Новосибирск, Ленина 5")
+	berdsk := parser.Parse("Бердск, Ленина 5")
+	if nsk.StreetKey == berdsk.StreetKey {
+		t.Fatalf("same street key for different cities: %q", nsk.StreetKey)
+	}
+	if nsk.CityKey != "новосибирск" || berdsk.CityKey != "бердск" {
+		t.Fatalf("city keys = %q / %q", nsk.CityKey, berdsk.CityKey)
+	}
+}
+
+func TestParseDefaultCity(t *testing.T) {
+	parser := NewParser(testDictionary(), "Новосибирск")
+	got := parser.Parse("Ленина 5")
+	if got.City != "Новосибирск" || got.CityKey != "новосибирск" {
+		t.Fatalf("City = %q / %q", got.City, got.CityKey)
+	}
+
+	// The default city is also recognised when the user spells it out, even
+	// if the dictionary contains only aliases.
+	explicit := NewParser(DefaultDictionary(), "Новосибирск").Parse("г. Новосибирск, Ленина 5")
+	if explicit.City != "Новосибирск" || explicit.NameKey != "ленина" {
+		t.Fatalf("explicit default city: %+v", explicit)
+	}
+}
+
+func TestParseInCityOverridesDetectedCity(t *testing.T) {
+	parser := NewParser(testDictionary(), "")
+	got := parser.ParseInCity("Бердск, Ленина 5", "Новосибирск")
+	if got.City != "Новосибирск" {
+		t.Fatalf("City = %q, want forced Новосибирск", got.City)
+	}
+	if got.NameKey != "ленина" {
+		t.Fatalf("NameKey = %q", got.NameKey)
+	}
+}
+
+func TestParseWithoutCityKeepsCityKeyEmpty(t *testing.T) {
+	parser := NewParser(testDictionary(), "")
+	got := parser.Parse("Ленина 5")
+	if got.City != "" || got.CityKey != "" {
+		t.Fatalf("City = %q / %q, want empty", got.City, got.CityKey)
+	}
+}
+
+func TestNormalizeTextStillLeavesUnknownCityTokens(t *testing.T) {
+	// A city that is not in the dictionary must not be silently dropped: the
+	// resolver should fail rather than guess another city.
+	parser := NewParser(testDictionary(), "")
+	got := parser.Parse("Томск, Ленина 5")
+	if got.City != "" {
+		t.Fatalf("City = %q, want empty for unknown city", got.City)
+	}
+	if got.NameKey == "ленина" {
+		t.Fatalf("unknown city token disappeared from the street name")
+	}
+}
+
+func TestMatchCity(t *testing.T) {
+	dict := testDictionary()
+	cases := []struct {
+		query string
+		want  string
+		ok    bool
+	}{
+		{"нск", "Новосибирск", true},
+		{"г. Новосибирск", "Новосибирск", true},
+		{"Новосибирская область", "Новосибирск", true},
+		{"Бердск", "Бердск", true},
+		{"Бердск, Ленина 5", "", false},
+		{"Томск", "", false},
+	}
+	for _, tt := range cases {
+		got, ok := dict.MatchCity(tt.query)
+		if ok != tt.ok || got.Name != tt.want {
+			t.Errorf("MatchCity(%q) = %+v, %v; want %q, %v", tt.query, got, ok, tt.want, tt.ok)
+		}
+	}
+}
+
+func FuzzParse(f *testing.F) {
+	for _, seed := range []string{
+		"",
+		"2-я Станционная улица 1/1 стр.",
+		"нск, Ленина 5",
+		"Бердск, Ленина 5 к2",
+		"г. Новосибирск, ул. Кошурникова, д. 11/1 кв. 5",
+		"\x00\xff",
+	} {
+		f.Add(seed)
+	}
+	parser := NewParser(testDictionary(), "")
+	f.Fuzz(func(t *testing.T, raw string) {
+		first := parser.Parse(raw)
+		second := parser.Parse(raw)
+		if first.StreetKey != second.StreetKey {
+			t.Fatalf("non-deterministic street key for %q: %q vs %q", raw, first.StreetKey, second.StreetKey)
+		}
+		if NormalizeText(raw) != NormalizeText(NormalizeText(raw)) {
+			t.Fatalf("NormalizeText is not idempotent for %q", raw)
+		}
+	})
+}