aps-agent/tests/golden/test_importers.py

# ============================================================
# MD-04 Excel/CSV 导入管线黄金测试
# ============================================================
from __future__ import annotations

from server.aps_domain.importers import apply_import_commit, preview_file, validate_batch
from server.state.seed import seed_world


def _next_id_factory(world):
    tables = {
        "audit": "auditEvents",
        "salesOrder": "salesOrders",
        "material": "materials",
    }
    counters: dict[str, int] = {}

    def next_id(kind: str) -> int:
        if kind not in counters:
            rows = world.get(tables.get(kind, kind + "s"), [])
            counters[kind] = max((r.get("id", 0) for r in rows if isinstance(r, dict)), default=0)
        counters[kind] += 1
        return counters[kind]

    return next_id


def test_validate_equipment_batch():
    world = seed_world()
    rows = [
        {"code": "EQ-X1", "name": "试验压接机", "capabilities": "OP-PRESS,OP-WELD",
         "opStdTime": "OP-PRESS:2,OP-WELD:5", "movable": "是", "zone": "ZONE-A",
         "availabilityRate": "0.95", "status": "RUNNING"},
    ]
    batch = validate_batch("equipment", rows, world)
    assert batch["okCount"] == 1
    assert batch["errorCount"] == 0
    assert "OP-PRESS" in batch["okRows"][0]["capabilities"]


def test_preview_csv_orders_and_commit():
    world = seed_world()
    product = next(m for m in world["materials"] if m.get("type") == "FINISHED_PRODUCT")
    before = len(world["salesOrders"])
    csv_text = (
        "订单号,客户,产品编码,数量,交期,优先级\n"
        f"SO-IMP-1,导入客户,{product['code']},50,2026-08-01,3\n"
    ).encode("utf-8")
    preview = preview_file("orders.csv", csv_text, world)
    assert preview["canCommit"] is True
    assert preview["totalOk"] >= 1
    assert preview["batches"][0]["kind"] == "orders"

    applied = apply_import_commit(world, _next_id_factory(world), preview["batches"])
    assert applied["total"] >= 1
    assert len(world["salesOrders"]) == before + 1
    assert any(o.get("customerName") == "导入客户" for o in world["salesOrders"])


def test_commit_flex_equipment():
    world = seed_world()
    batches = [{
        "kind": "equipment",
        "sheet": "设备",
        "okRows": [{
            "code": "EQ-NEW-99", "name": "新压接", "capabilities": ["OP-PRESS"],
            "opStdTime": {"OP-PRESS": 1.5}, "movable": True, "moveTimeMin": 15,
            "zone": "ZONE-A", "availabilityRate": 1.0, "status": "RUNNING",
            "adaptableMolds": [],
        }],
    }]
    applied = apply_import_commit(world, _next_id_factory(world), batches)
    assert applied["summary"].get("flexEquipment", 0) == 1
    assert any(e["code"] == "EQ-NEW-99" for e in world["flexEquipment"])


def test_xlsx_mom_template_multiblock_headers_soft_preview():
    """锐扬 MOM 收集表模板：多行标题/分块表头也能识别（round-55）。

    - 「02-工艺模型」标题行 + 第 4 行真实表头（序号/ERP品号/产品名称/单套用量/单位）
    - 「04-物料模型」物料组块（第 1 行）+ 物料主数据块（第 5 行）
    - 「07-设备模型」第 1 行表头用「财务编号/设备名称」（无可执行工序 → soft GENERAL）
    """
    import io

    import openpyxl
    from server.aps_domain.importers import preview_file
    from server.state.seed import seed_world

    wb = openpyxl.Workbook()
    ws1 = wb.active
    ws1.title = "02-工艺模型"
    ws1.append(["湖南锐扬创科智能科技有限公司\n物料清单", "", "", "", "", "", ""])
    ws1.append(["", "", "", "", "", "", ""])
    ws1.append(["客户代码", "", "R001", "", "", "", ""])
    ws1.append(["序号", "层次", "ERP品号", "产品名称", "单套用量", "单位",
                "制造流程(根据工艺代码表填写对应的工艺代码）"])
    ws1.append(["", "", "", "", "", "", "01", "02", "03"])
    ws1.append([1, 1, "A0050101-00280", "外壳", 1, "PCS",
                "ZP01组装", "LM01POP拉铆", "BZ01机柜包装"])
    ws1.append([2, 1, "B0050121-22551", "柜体焊接组件", 1, "PCS",
                "ZP01组装", "WELD焊接"])

    ws4 = wb.create_sheet("04-物料模型")
    ws4.append(["物料组", "物料组名称", "物料组编码", "所属工厂"])
    ws4.append(["", "示例：成品组", "CP", "湖南锐扬"])
    ws4.append(["物料主数据", "物料名称", "物料简称", "物料编码", "物料组名称", "物料组编码", "单位"])
    ws4.append(["", "外壳", "外壳", "A0050101-00280", "成品", "CP", "PCS"])
    ws4.append(["", "冷轧板", "冷轧板", "C0010102-00021", "原材料", "YCL", "KG"])

    ws7 = wb.create_sheet("07-设备模型")
    ws7.append(["设备台账管理", "序号", "财务编号", "设备名称", "设备型号"])
    ws7.append(["", "1", "A00076", "光纤激光切割机", "D-FAST1530FCCBD300"])

    buf = io.BytesIO()
    wb.save(buf)
    world = seed_world()
    preview = preview_file("湖南锐扬MOM主数据收集表.xlsx", buf.getvalue(), world, soft=True)
    assert preview["canCommit"] is True
    batches = {b["sheet"]: b for b in preview["batches"]}
    assert batches["02-工艺模型"]["okCount"] == 2
    row = batches["02-工艺模型"]["okRows"][0]
    assert row["code"] == "A0050101-00280" and row["name"] == "外壳"
    # 制造流程列 → routing 视图（round-57）
    rts = batches["02-工艺模型#routing"]
    assert rts["kind"] == "routing" and rts["okCount"] == 5
    by_key = {(r["productCode"], r["seq"]): r for r in rts["okRows"]}
    first = by_key[("A0050101-00280", 1)]
    assert first["operationCode"] == "ZP01" and first["operationName"] == "组装"
    assert by_key[("A0050101-00280", 2)]["operationName"] == "拉铆"
    assert by_key[("B0050121-22551", 2)]["operationCode"] == "WELD"
    assert batches["04-物料模型"]["okCount"] == 2
    by_code = {r["code"]: r for r in batches["04-物料模型"]["okRows"]}
    assert by_code["A0050101-00280"]["type"] == "FINISHED_PRODUCT"
    assert by_code["C0010102-00021"]["type"] == "RAW_MATERIAL"
    eq = batches["07-设备模型"]
    assert eq["okCount"] == 1
    assert eq["okRows"][0]["code"] == "A00076"
    assert eq["okRows"][0]["capabilities"] == ["GENERAL"]


def test_xlsx_ruiyang_structured_diagnostics_preserve_source_rows_and_legacy_contract():
    """锐扬 53 行语义缩比：可恢复解析噪声、结构行和唯一真实阻塞应分离。"""
    import io
    from collections import Counter

    import openpyxl
    from server.aps_domain.importers import preview_file
    from server.state.seed import seed_world

    wb = openpyxl.Workbook()

    # 02-工艺模型：第 5 行是制造流程二级序号表头，不是 materials 数据。
    ws2 = wb.active
    ws2.title = "02-工艺模型"
    ws2.append(["湖南锐扬创科智能科技有限公司\n物料清单"] + [""] * 27)
    ws2.append([""] * 28)
    ws2.append(["客户代码", "", "R001"] + [""] * 25)
    ws2.append([
        "序号", "层次", "ERP品号", "客户图号", "产品名称", "客户版本",
        "单套用量", "出数/底数", "单位", "品号属性", "仓库代码", "材质",
        "厚度", "展开长*宽", "面积", "表面处理",
        "制造流程(根据工艺代码表填写对应的工艺代码）",
    ] + [""] * 11)
    for col in range(1, 17):
        ws2.merge_cells(start_row=4, start_column=col, end_row=5, end_column=col)
    ws2.merge_cells("Q4:AB4")
    for offset in range(12):
        ws2.cell(row=5, column=17 + offset, value=f"{offset + 1:02d}")
    material_row = [None] * 28
    material_row[0] = 1
    material_row[1] = 1
    material_row[2] = "MAT-001"
    material_row[4] = "合成外壳"
    material_row[6] = 1
    material_row[8] = "PCS"
    material_row[16] = "ZP01组装"
    ws2.append(material_row)

    # 07-设备模型：2 条重复 code 别名噪声、4 条嵌入 section、1 条真实缺码。
    ws7 = wb.create_sheet("07-设备模型")
    ws7.append([
        "设备台账管理", "序号", "财务编号", "固定资产编号", "设备名称",
        "设备型号", "厂内编号", "出厂编号", "位置/一级保养责任人", "供应商",
        "购置年月", "设备类型", "设备状态", "备注", "", "设备状态",
    ])
    ws7.append(["", 1, "FIN-VALID", "FA-VALID", "厂内编码设备", "MODEL-A",
                "INTERNAL-VALID", "SERIAL-A", "A区", "供应商A", None,
                "生产设备", "使用中"])
    ws7.append(["", 2, "FIN-001", "FA-001", "财务编码设备一", "MODEL-B",
                None, "SERIAL-B", "B区", "供应商B", None, "生产设备", "使用中"])
    ws7.append(["", 3, "FIN-002", "FA-002", "财务编码设备二", "MODEL-C",
                None, "SERIAL-C", "C区", "供应商C", None, "生产设备", "使用中"])
    ws7.append(["", 4, None, None, "唯一缺码设备", "MODEL-D",
                None, "SERIAL-D", "D区", "供应商D", None, "生产设备", "使用中"])
    ws7.merge_cells("A1:A5")

    ws7.append(["备品备件管理", "备件名称", "备件编码", "规格型号", "单位", "备注"])
    ws7.append([None] * 16)
    ws7.merge_cells("A6:A7")
    ws7.append(["设备部件管理", "部件名称", "部件编码", "规格型号", "单位", "备注"])
    ws7.append([None] * 16)
    ws7.merge_cells("A8:A9")
    ws7.append(["设备类型", "类型名称", "类型编码", "备注"])
    ws7.append(["", "生产设备", "SBLX001"])
    ws7.append([None] * 16)
    ws7.merge_cells("A10:A12")

    buf = io.BytesIO()
    wb.save(buf)
    preview = preview_file(
        "湖南锐扬MOM主数据收集表.xlsx",
        buf.getvalue(),
        seed_world(),
        soft=True,
    )

    # 兼容性：原 preview/batch 字段必须继续存在。
    assert {"filename", "batches", "totalOk", "totalErrors", "canCommit"} <= preview.keys()
    batches = {b["sheet"]: b for b in preview["batches"]}
    materials = batches["02-工艺模型"]
    equipment = batches["07-设备模型"]
    legacy_batch_fields = {
        "sheet", "kind", "fieldMap", "headersRaw", "okRows", "errors",
        "warnings", "okCount", "errorCount",
    }
    assert legacy_batch_fields <= materials.keys()
    assert legacy_batch_fields <= equipment.keys()
    assert isinstance(materials["diagnostics"], list)
    assert isinstance(equipment["diagnostics"], list)

    # 02 row 5：SECONDARY_SEQUENCE_HEADER 必须忽略且保留真实 Excel 行号。
    secondary = [
        d for d in materials["diagnostics"]
        if d.get("code") == "SECONDARY_SEQUENCE_HEADER"
    ]
    assert len(secondary) == 1
    assert secondary[0]["classification"] == "ignorable_section"
    assert secondary[0]["excelRow"] == 5
    assert materials["okCount"] == 1
    assert materials["errorCount"] == 0

    # 重复 code 别名：最后一个非空编码优先；后续空列不能覆盖已有编码。
    equipment_by_code = {row["code"]: row for row in equipment["okRows"]}
    assert {"INTERNAL-VALID", "FA-001", "FA-002"} <= equipment_by_code.keys()
    assert equipment["okCount"] == 3
    assert equipment["errorCount"] == 1
    assert preview["totalErrors"] == 1

    # 真实 47/5/1 的小型缩比：2 条 parser_noise、5 条结构忽略、1 条 blocking。
    all_diagnostics = materials["diagnostics"] + equipment["diagnostics"]
    primary_classes = Counter(
        d.get("classification") for d in all_diagnostics
        if d.get("classification") in {"parser_noise", "ignorable_section", "blocking"}
    )
    assert primary_classes == Counter({
        "parser_noise": 2,
        "ignorable_section": 5,
        "blocking": 1,
    })
    assert {
        d["excelRow"] for d in equipment["diagnostics"]
        if d.get("classification") == "parser_noise"
    } == {3, 4}
    assert {
        d["excelRow"] for d in equipment["diagnostics"]
        if d.get("classification") == "ignorable_section"
    } == {6, 8, 10, 11}
    blocking = [
        d for d in equipment["diagnostics"]
        if d.get("classification") == "blocking"
    ]
    assert len(blocking) == 1 and blocking[0]["excelRow"] == 5

    # soft 模式下 GENERAL warning 与成功行共存，不能把成功行重新算成 error。
    assert all(equipment_by_code[code]["capabilities"] == ["GENERAL"]
               for code in ("INTERNAL-VALID", "FA-001", "FA-002"))
    assert any("GENERAL" in str(warning) for warning in equipment["warnings"])
    general_diagnostics = [
        d for d in equipment["diagnostics"]
        if d.get("classification") == "warning" and "GENERAL" in str(d)
    ]
    assert general_diagnostics
    assert all(isinstance(d.get("excelRow"), int) for d in all_diagnostics)

    # __source__ 只供内部追踪，不能泄漏到兼容的 okRows 领域行。
    assert all(
        "__source__" not in row
        for batch in preview["batches"]
        for row in batch["okRows"]
    )


def test_xlsx_flat_header_row_0_behavior_unchanged():
    """平铺表第 1 行即表头：探测不得改变既有行为（round-55 回归）。"""
    import io

    import openpyxl
    from server.aps_domain.importers import preview_file
    from server.state.seed import seed_world

    wb = openpyxl.Workbook()
    ws = wb.active
    ws.title = "物料"
    ws.append(["编码", "名称", "类型", "单位", "库存"])
    ws.append(["M-FLAT-1", "平铺物料", "FINISHED_PRODUCT", "件", "10"])
    buf = io.BytesIO()
    wb.save(buf)
    world = seed_world()
    preview = preview_file("物料.xlsx", buf.getvalue(), world, soft=True)
    assert preview["canCommit"] is True
    assert preview["batches"][0]["okRows"][0]["code"] == "M-FLAT-1"


def test_order_csv_rush_column_maps_to_is_rush():
    """订单表「加急」列 → isRush（round-58，现场订单表常用列）。"""
    from server.aps_domain.importers import preview_file
    from server.state.seed import seed_world

    world = seed_world()
    product = next(m for m in world["materials"] if m.get("type") == "FINISHED_PRODUCT")
    csv_text = (
        "订单号,客户,产品编码,数量,交期,优先级,加急\n"
        f"SO-RUSH-1,现场客户,{product['code']},10,2026-08-10,3,是\n"
    ).encode("utf-8")
    preview = preview_file("订单.csv", csv_text, world, soft=True)
    assert preview["totalOk"] >= 1
    row = preview["batches"][0]["okRows"][0]
    assert row["isRush"] is True
    assert row["orderNo"] == "SO-RUSH-1"