feat(api): product families backfill script

2026-08-28 12:36:39 +08:00
parent d20a933e34
commit d63f1a6f35
4 changed files with 159 additions and 76 deletions
@@ -22,7 +22,8 @@
    "prisma:migrate": "prisma migrate dev",
    "prisma:studio": "prisma studio",
    "configure:product-center-icons": "ts-node prisma/configure-product-center-icons.ts",
-    "import:product-detail": "ts-node prisma/import-product-detail.ts"
+    "import:product-detail": "ts-node prisma/import-product-detail.ts",
+    "backfill:product-families": "ts-node prisma/backfill-product-families.ts"
  },
  "dependencies": {
    "@nestjs/axios": "^3.0.1",
@@ -0,0 +1,71 @@
+/**
+ * 产品族回填脚本（一次性 / 幂等）：
+ *   1. 全量 OriginGood 回填四个链接名解析列；
+ *   2. auto-group 全量建族并挂成员（每族建立即重算）；
+ *   3. 输出统计与不可解析清单。
+ *
+ * 运行：pnpm --filter @inkreach/api backfill:product-families
+ * 幂等性：重复执行时步骤 1 数据不变、步骤 2 候选为空（familyId=null 过滤）。
+ */
+import { PrismaService } from '../src/prisma/prisma.service';
+import { FamilyRecomputeService } from '../src/product-families/family-recompute.service';
+import { ProductFamiliesService } from '../src/product-families/product-families.service';
+import { parseOriginName } from '../src/product-families/origin-name.parser';
+
+async function main() {
+  const prisma = new PrismaService();
+  await prisma.onModuleInit();
+  const recompute = new FamilyRecomputeService(prisma);
+  const families = new ProductFamiliesService(prisma, recompute);
+
+  // ---- 1. 解析列回填（分批） ----
+  const BATCH = 100;
+  let parsed = 0;
+  const unparsable: string[] = [];
+  for (;;) {
+    const batch = await prisma.originGood.findMany({
+      orderBy: { id: 'asc' },
+      take: BATCH,
+      skip: parsed,
+      select: { id: true, goodName: true },
+    });
+    if (batch.length === 0) break;
+    for (const og of batch) {
+      const p = parseOriginName(og.goodName);
+      if (!p.skuCode && !p.craftLabel) unparsable.push(`#${og.id} ${og.goodName ?? ''}`);
+      await prisma.originGood.update({
+        where: { id: og.id },
+        data: {
+          skuCode: p.skuCode,
+          logisticsLabel: p.logisticsLabel,
+          craftLabel: p.craftLabel,
+          warehouseLabel: p.warehouseLabel,
+        },
+      });
+    }
+    parsed += batch.length;
+  }
+  console.log(`[1/2] parsed ${parsed} origin goods (${unparsable.length} without sku/craft)`);
+
+  // ---- 2. 自动建族（含逐族重算） ----
+  const result = await families.autoGroup(true);
+  console.log(`[2/2] created ${result.applied} families`);
+
+  // ---- 统计 ----
+  const total = await prisma.productFamily.count();
+  const withMatrix = await prisma.productFamily.count({ where: { priceMatrix: { not: null } } });
+  const members = await prisma.originGood.count({ where: { familyId: { not: null } } });
+  const stale = await prisma.productFamily.count({ where: { stale: true } });
+  console.log(`stats: families=${total}, materialized=${withMatrix}, members=${members}, stale=${stale}`);
+  if (unparsable.length) {
+    console.log('unparsable names:');
+    for (const line of unparsable) console.log(`  - ${line}`);
+  }
+
+  await prisma.$disconnect();
+}
+
+main().catch((err) => {
+  console.error(err);
+  process.exit(1);
+});
@@ -132,7 +132,7 @@ export class ProductFamiliesService {
  async autoGroup(apply: boolean) {
    const candidates = await this.prisma.originGood.findMany({
      where: { familyId: null, delisted: false },
-      select: { id: true, goodName: true, goodImage: true },
+      select: { id: true, goodName: true, goodImage: true, source: true },
      orderBy: { id: 'asc' },
    });
    const groups = new Map<string, typeof candidates>();
@@ -160,10 +160,16 @@ export class ProductFamiliesService {
    let applied = 0;
    for (const members of groups.values()) {
      const parsed = parseOriginName(members[0].goodName);
+      const fallbackCode =
+        members[0].source === 'CUSTOM' ? `CUSTOM-${members[0].id}` : null;
      const family = await this.prisma.productFamily.create({
        data: {
          familyName: parsed.productName ?? parsed.country ?? originGroupKey(members[0].goodName),
-          familyCode: parsed.skuCode ? await this.ensureUniqueCode(parsed.skuCode) : null,
+          familyCode: parsed.skuCode
+            ? await this.ensureUniqueCode(parsed.skuCode)
+            : fallbackCode
+              ? await this.ensureUniqueCode(fallbackCode)
+              : null,
          familyImage: members[0].goodImage ?? null,
          primaryOriginGoodId: members[0].id,
        },
@@ -7,8 +7,9 @@ import {
 } from './sync.service';
 import { SdsClientService } from './sds-client.service';
 import { PrismaService } from '../prisma/prisma.service';
+import { FamilyRecomputeService } from '../product-families/family-recompute.service';

-describe('SyncService', () => {
+describe('SyncService', () => {
  let service: SyncService;
  let sds: jest.Mocked<SdsClientService>;
  let prisma: PrismaService;
@@ -17,22 +18,23 @@ describe('SyncService', () => {

  beforeAll(async () => {
    const sdsMock: Partial<SdsClientService> = {
-      fetchCategoryTree: jest.fn(),
-      fetchProductsPage: jest.fn(),
-      fetchProductDetail: jest.fn(async (goodId: string | number) => ({ id: goodId })),
+      fetchCategoryTree: jest.fn(),
+      fetchProductsPage: jest.fn(),
+      fetchProductDetail: jest.fn(async (goodId: string | number) => ({ id: goodId })),
    };
    const moduleRef = await Test.createTestingModule({
      imports: [ConfigModule.forRoot({ isGlobal: true })],
      providers: [
        SyncService,
        { provide: SdsClientService, useValue: sdsMock },
+        { provide: FamilyRecomputeService, useValue: { enqueue: jest.fn() } },
        PrismaService,
      ],
    }).compile();
-    service = moduleRef.get(SyncService);
-    jest
-      .spyOn(service, 'syncConfiguredProductDetails')
-      .mockResolvedValue({ synced: 0, failed: 0 });
+    service = moduleRef.get(SyncService);
+    jest
+      .spyOn(service, 'syncConfiguredProductDetails')
+      .mockResolvedValue({ synced: 0, failed: 0 });
    sds = moduleRef.get(SdsClientService) as jest.Mocked<SdsClientService>;
    prisma = moduleRef.get(PrismaService);
    await prisma.onModuleInit();
@@ -275,68 +277,71 @@ describe('SyncService', () => {
      expect(logs.length).toBeGreaterThan(0);
    });
  });
-});
-
-describe('SyncService product detail scopes', () => {
-  const originGoods = [
-    { id: 1n, sdsGoodId: 'all-1' },
-    { id: 2n, sdsGoodId: 'all-2' },
-  ];
-
-  function createService() {
-    const prisma = {
-      originGood: { findMany: jest.fn().mockResolvedValue(originGoods) },
-    } as unknown as PrismaService;
-    const sds = {
-      fetchProductDetail: jest.fn(async (goodId: string) => ({ id: goodId })),
-    } as unknown as SdsClientService;
-    const scopedService = new SyncService(prisma, sds);
-    jest
-      .spyOn(scopedService as any, 'persistProductDetail')
-      .mockResolvedValue(undefined);
-    return { scopedService, prisma, sds };
-  }
-
-  it('manual detail sync selects every active origin product', async () => {
-    const { scopedService, prisma, sds } = createService();
-    const result = await scopedService.syncAllProductDetails();
-
-    expect(prisma.originGood.findMany).toHaveBeenCalledWith(
-      expect.objectContaining({ where: { delisted: false, source: 'SDS' } }),
-    );
-    expect(sds.fetchProductDetail).toHaveBeenCalledTimes(2);
-    expect(result).toEqual({ total: 2, synced: 2, failed: 0 });
-  });
-
-  it('hourly detail refresh remains limited to configured products', async () => {
-    const { scopedService, prisma } = createService();
-    await scopedService.syncConfiguredProductDetails();
-
-    expect(prisma.originGood.findMany).toHaveBeenCalledWith(
-      expect.objectContaining({
-        where: { delisted: false, source: 'SDS', goods: { some: {} } },
-      }),
-    );
-  });
-
-  it('keeps hourly category/product sync separate from the daily detail sync', async () => {
-    const { scopedService } = createService();
-    const categories = jest.spyOn(scopedService, 'syncCategories').mockResolvedValue({
-      inserted: 0, updated: 0, total: 0, deletedStale: 0,
-    });
-    const products = jest.spyOn(scopedService, 'syncProducts').mockResolvedValue({
-      inserted: 0, updated: 0, total: 0, leafCategories: 0, delisted: 0,
-    });
-    const details = jest.spyOn(scopedService, 'syncProductDetails').mockResolvedValue({
-      total: 0, synced: 0, failed: 0,
-    });
-
-    await scopedService.hourlyCron();
-    expect(categories).toHaveBeenCalledTimes(1);
-    expect(products).toHaveBeenCalledTimes(1);
-    expect(details).not.toHaveBeenCalled();
-
-    await scopedService.dailyProductDetailCron();
-    expect(details).toHaveBeenCalledTimes(1);
-  });
-});
+});
+
+describe('SyncService product detail scopes', () => {
+  const originGoods = [
+    { id: 1n, sdsGoodId: 'all-1' },
+    { id: 2n, sdsGoodId: 'all-2' },
+  ];
+
+  function createService() {
+    const prisma = {
+      originGood: { findMany: jest.fn().mockResolvedValue(originGoods) },
+    } as unknown as PrismaService;
+    const sds = {
+      fetchProductDetail: jest.fn(async (goodId: string) => ({ id: goodId })),
+    } as unknown as SdsClientService;
+    const familyRecompute = {
+      enqueue: jest.fn(),
+    } as unknown as FamilyRecomputeService;
+    const scopedService = new SyncService(prisma, sds, familyRecompute);
+    jest
+      .spyOn(scopedService as any, 'persistProductDetail')
+      .mockResolvedValue(undefined);
+    return { scopedService, prisma, sds };
+  }
+
+  it('manual detail sync selects every active origin product', async () => {
+    const { scopedService, prisma, sds } = createService();
+    const result = await scopedService.syncAllProductDetails();
+
+    expect(prisma.originGood.findMany).toHaveBeenCalledWith(
+      expect.objectContaining({ where: { delisted: false, source: 'SDS' } }),
+    );
+    expect(sds.fetchProductDetail).toHaveBeenCalledTimes(2);
+    expect(result).toEqual({ total: 2, synced: 2, failed: 0 });
+  });
+
+  it('hourly detail refresh remains limited to configured products', async () => {
+    const { scopedService, prisma } = createService();
+    await scopedService.syncConfiguredProductDetails();
+
+    expect(prisma.originGood.findMany).toHaveBeenCalledWith(
+      expect.objectContaining({
+        where: { delisted: false, source: 'SDS', goods: { some: {} } },
+      }),
+    );
+  });
+
+  it('keeps hourly category/product sync separate from the daily detail sync', async () => {
+    const { scopedService } = createService();
+    const categories = jest.spyOn(scopedService, 'syncCategories').mockResolvedValue({
+      inserted: 0, updated: 0, total: 0, deletedStale: 0,
+    });
+    const products = jest.spyOn(scopedService, 'syncProducts').mockResolvedValue({
+      inserted: 0, updated: 0, total: 0, leafCategories: 0, delisted: 0,
+    });
+    const details = jest.spyOn(scopedService, 'syncProductDetails').mockResolvedValue({
+      total: 0, synced: 0, failed: 0,
+    });
+
+    await scopedService.hourlyCron();
+    expect(categories).toHaveBeenCalledTimes(1);
+    expect(products).toHaveBeenCalledTimes(1);
+    expect(details).not.toHaveBeenCalled();
+
+    await scopedService.dailyProductDetailCron();
+    expect(details).toHaveBeenCalledTimes(1);
+  });
+});