"""入库、周四归月与聚合口径的单元测试。""" from __future__ import annotations import csv from datetime import date from decimal import Decimal from pathlib import Path import pytest from toolsvc.customs.engine.rules import Rules from toolsvc.customs.engine.storage import Store, ingest, week_rows RULES_DIR = Path(__file__).resolve().parents[1] / "toolsvc" / "customs" / "rules" HEADER = ["I_E_FLAG", "HS_CODE", "YEAR", "WEEK", "MONORSUM", "COUNT", "G_MODEL", "ORIGIN_COUNTRY", "TRADE_AREA_CODE", "TRADE_COUNTRY", "DISTINATE_PORT"] @pytest.fixture(scope="module") def rules(): return Rules(RULES_DIR / "数据映射规则.xlsx", RULES_DIR / "进出口数据字段需求20260819.xlsx") def write_csv(path: Path, rows: list[list[str]]) -> Path: with path.open("w", encoding="utf-8", newline="") as handle: writer = csv.writer(handle) writer.writerow(HEADER) writer.writerows(rows) return path def config_for(path: Path, out: Path, **overrides) -> dict: return { "input_csv": str(path), "encoding": "utf-8", "delimiter": ",", "directions": ["I"], "cutoff_date": None, "start_year": 2025, "end_year": 2026, "country_code_system": "ISO", "confirmed_complete_years": [], "allow_partial_year_basket": True, **overrides, } def sample_row(hs="8532241000", year="2025", week="1", amount="100.50", quantity="10", model="4|0|多层|陶瓷|MURATA|GRM188|||"): return ["I", hs, year, week, amount, quantity, model, "JPN", "KOR", "JPN", "JPN501"] def test_thursday_month_mapping_is_iso_week_based(): assert date.fromisocalendar(2025, 13, 4) == date(2025, 3, 27) rows = {row[1]: row for row in week_rows(2025, 2026) if row[0] == 2025} assert rows[13][6] == 3 # 归属 3 月 assert rows[5][6] == 1 # 2025-W05 周四 1-30,归 1 月 assert rows[18][6] == 5 # 2025-W18 周四 5-01,归 5 月 def test_ingest_maps_week_to_month_and_counts(tmp_path, rules): path = write_csv(tmp_path / "in.csv", [sample_row(week="1"), sample_row(week="5", amount="200.00")]) database, stats = ingest(config_for(path, tmp_path), rules, tmp_path) try: assert stats["total"] == 2 assert stats["included"] == 2 assert stats["min_month"] == "2025-01" store = Store(database) total = store.total({"direction": "I"}) assert total["amount"] == Decimal("300.50") assert total["n"] == 2 months = {row["month"]: row for row in store.aggregate(("month",), {"direction": "I"})} assert set(months) == {"2025-01"} # W01 与 W05 的周四都在 1 月 finally: database.close() def test_ingest_excludes_out_of_scope_rows(tmp_path, rules): rows = [ sample_row(), # 正常 sample_row(hs="85322410"), # 8 位 HS sample_row(hs="8517700000"), # 不在项目范围 sample_row(week="60"), # 非法周 sample_row(year="2031", week="2"), # 晚于截止日期(默认今天) ["E", "8532241000", "2025", "1", "10", "1", "4|0|多层|陶瓷|MURATA|GRM188|||", "JPN", "KOR", "JPN", "JPN501"], # 方向不符 ] path = write_csv(tmp_path / "in.csv", rows) database, stats = ingest(config_for(path, tmp_path, cutoff_date="2026-12-31"), rules, tmp_path) try: assert stats["total"] == 6 assert stats["included"] == 1 # 只有第一行在范围内;2031 年仍晚于 2026-12-31 assert stats["excluded"] == 5 finally: database.close() def test_ingest_records_numeric_alerts_without_dropping_rows(tmp_path, rules): path = write_csv(tmp_path / "in.csv", [sample_row(amount="0"), sample_row(amount="-5", quantity="abc")]) database, stats = ingest(config_for(path, tmp_path), rules, tmp_path) try: assert stats["included"] == 2 assert stats["alerts"] >= 2 issues = database.execute("SELECT field,reason FROM issues").fetchall() assert any("零值或负数" in row["reason"] for row in issues) assert any("无法识别" in row["reason"] for row in issues) finally: database.close() def test_ingest_rejects_missing_required_column(tmp_path, rules): path = tmp_path / "bad.csv" path.write_text("I_E_FLAG,HS_CODE\nI,8532241000\n", encoding="utf-8") with pytest.raises(ValueError, match="缺少必需字段"): ingest(config_for(path, tmp_path), rules, tmp_path) def test_store_rejects_unknown_dimension(rules, tmp_path): path = write_csv(tmp_path / "in.csv", [sample_row()]) database, _stats = ingest(config_for(path, tmp_path), rules, tmp_path) try: with pytest.raises(ValueError): list(Store(database).aggregate(("nope",), {})) finally: database.close() def test_chinese_headers_are_aliased(tmp_path, rules): path = tmp_path / "cn.csv" path.write_text( "进出口,HS编码,年份,周,金额,数量,品牌/型号\nI,8532241000,2025,3,88.80,7,4|0|多层|陶瓷|MURATA|GRM188|||\n", encoding="utf-8", ) database, stats = ingest(config_for(path, tmp_path), rules, tmp_path) try: assert stats["included"] == 1 finally: database.close()