fix(scrape): 整理后刮削照 MP 对作品目录刮,集只刮本次整理的文件

原地入库逐个文件排刮削,单文件刮削只写该集,剧集根目录(tvshow.nfo、海报、季海报)与季目录
(season.nfo)没有元数据。照 MP ScrapingChain.scrape_metadata_event:按作品目录分组,目录元数据照常
初始化,集只刮本次文件;复制 / 移动同样改为只刮本次文件(此前刮整个作品目录)。

- 迁移 0062:scrape_record.videos(null = 目录下全部)
- queueScrape 支持 videos;同目录排队中还没开始的刮削合成一批(有一方刮全部则刮全部)
- 作品详情「重新刮削」与文件管理刮削不变(目录下全部)

Co-Authored-By: Claude Opus 5.5 <[email protected]>

apps/server/src/modules/acquisition-organize/services/scrape.ts

@@ -77,6 +77,8 @@ export interface QueueScrapeInput extends ScrapeLocation {
   actor?: { type: string; id: string | null } | undefined;
   /** 引擎幂等键(入库后自动刮削按整理记录) */
   idempotencyKey?: string | undefined;
+  /** 只刮这些视频(库根相对路径;整理后刮削用,作品 / 季目录元数据照常初始化);缺省 = 目录下全部 */
+  videos?: readonly string[] | undefined;
 }
 
 type Row = {
@@ -93,6 +95,7 @@ type Row = {
   reason: string | null;
   files: unknown;
   organization_record_id: string | null;
+  videos: string[] | null;
   engine_run_id: string | null;
   requested_at: Date | string;
   started_at: Date | string | null;
@@ -116,6 +119,7 @@ export function scrapeRecordView(r: Row): ScrapeRecord {
     reason: r.reason,
     files: (r.files ?? []) as ScrapeFileResult[],
     organizationRecordId: r.organization_record_id,
+    videos: r.videos ?? null,
     taskId: r.engine_run_id,
     requestedAt: iso(r.requested_at)!,
     startedAt: iso(r.started_at),
@@ -157,7 +161,7 @@ export async function queueScrape(
   const mode = input.mode ?? (await scrapeSettings(d)).scrapeMode;
   const cur = await db
     .selectFrom("scrape_record")
-    .select(["id", "engine_run_id", "media_entity_id", "match"])
+    .select(["id", "engine_run_id", "media_entity_id", "match", "videos"])
     .where("root_id", "=", input.rootId)
     .where("path", "=", input.path)
     .where("mode", "=", mode)
@@ -165,10 +169,17 @@ export async function queueScrape(
     .where("requested_at", ">", new Date(Date.now() - DEDUPE_MS))
     .orderBy("requested_at", "desc")
     .executeTakeFirst();
+  const videos = input.videos?.length ? [...new Set(input.videos)].sort() : null;
   const sameMatch =
     !!cur &&
     (cur.media_entity_id ?? null) === input.match.mediaEntityId &&
     ((cur.match as ScrapeMatch | null)?.externalId ?? null) === input.match.externalId;
+  if (cur && sameMatch) {
+    // 同一目录还没开始的刮削合成一批(MP 按目标目录合批):有一方刮全部则刮全部,否则合并文件
+    const merged = cur.videos && videos ? [...new Set([...cur.videos, ...videos])].sort() : null;
+    if (JSON.stringify(merged) !== JSON.stringify(cur.videos ? [...cur.videos].sort() : null))
+      await db.updateTable("scrape_record").set({ videos: merged }).where("id", "=", cur.id).execute();
+  }
   if (cur && sameMatch && cur.engine_run_id) return { recordId: cur.id, taskId: cur.engine_run_id, deduplicated: true };
   const reused = !!cur && sameMatch;
   const rec = reused
@@ -188,6 +199,7 @@ export async function queueScrape(
           organization_record_id: input.organizationRecordId ?? null,
           actor_type: input.actor?.type ?? (input.trigger === "auto" ? "system" : null),
           actor_id: input.actor?.id ?? null,
+          videos,
         })
         .returning("id")
         .executeTakeFirstOrThrow();
@@ -362,7 +374,7 @@ export async function scrapeEntity(
   return out;
 }
 
-/** 整理入库成功后:设置打开或本次 Local 补全要求写本地元数据时,复用作品目录刮削任务。 */
+/** 整理入库成功后(设置打开或本次要求刮削):照 MP 对作品目录刮削,集只刮本次整理的文件。 */
 export async function scrapeAfterImport(
   d: AoDeps,
   entityId: string | null,
@@ -376,16 +388,17 @@ export async function scrapeAfterImport(
     if (!s.scrapeAfterImport && !opts.force && opts.enabled !== true) return;
     const organization = await d.ctx.db
       .selectFrom("organization_record")
-      .select(["action", "target_library_root_id"])
+      .select(["target_library_root_id"])
       .where("id", "=", recordId)
       .executeTakeFirst();
-    if (organization?.action === "in_place" && organization.target_library_root_id) {
     const match = await entityScrapeMatch(d, entityId, "library");
-      const root = await d.ctx.db
+    const root = organization?.target_library_root_id
+      ? await d.ctx.db
           .selectFrom("library_root")
           .selectAll()
           .where("id", "=", organization.target_library_root_id)
-        .executeTakeFirst();
+          .executeTakeFirst()
+      : undefined;
     if (!match || !root || root.management_mode !== "managed") return;
     const files = await d.ctx.db
       .selectFrom("organization_file_result")
@@ -395,21 +408,26 @@ export async function scrapeAfterImport(
       .where("status", "=", "succeeded")
       .orderBy("ordinal")
       .execute();
-      for (const file of files)
-        if (file.target_path)
+    // 照 MP 整理后刮削:按作品目录分组,对作品目录刮(作品 / 季的 NFO 与图片照常初始化),集只刮本次整理的文件;
+    // 视频直接在库根(推不出作品目录)时对文件本身刮。
+    const byDir = new Map<string, string[]>();
+    for (const f of files) {
+      if (!f.target_path) continue;
+      const dir = inferWorkDir(match.kind, f.target_path) ?? f.target_path;
+      byDir.set(dir, [...(byDir.get(dir) ?? []), f.target_path]);
+    }
+    for (const [dir, videos] of byDir)
       await queueScrape(d, {
         rootId: `lib:${root.id}`,
         rootPath: root.local_path,
         libraryRootId: root.id,
-            path: file.target_path,
+        path: dir,
         match,
         trigger: "auto",
         organizationRecordId: recordId,
-            idempotencyKey: `${SCRAPE_TASK}:record:${recordId}:${file.target_path}`,
+        idempotencyKey: `${SCRAPE_TASK}:record:${recordId}:${dir}`,
+        ...(dir === videos[0] ? {} : { videos }),
       });
-      return;
-    }
-    await scrapeEntity(d, entityId, { trigger: "auto", organizationRecordId: recordId });
   } catch (err) {
     d.ctx.logger.warn({ entityId, recordId, err: { message: errMessage(err) } }, "入库后自动刮削没排上");
   }
@@ -639,6 +657,8 @@ export async function runScrapeRecord(
 
     const kind = match.kind;
     const lay = await layoutOf(rootReal, rec.path, isDir, kind);
+    // 整理后刮削:目录布局照常(作品 / 季元数据),集只刮本次整理的文件
+    if (rec.videos) lay.videos = rec.videos.filter((v) => !rec.path || v === rec.path || v.startsWith(`${rec.path}/`));
     if (!lay.videos.length) return await fail("scrape_no_videos");
 
     // ── 字段:本地作品(有元数据)优先 ──

apps/server/test/acquisition-organize/scrape.test.ts

@@ -183,7 +183,8 @@ describe.skipIf(!hasPg)("刮削:NFO 与图片写进媒体库目录", () => {
     const rec = await organize(series, "Severance.S01E01.1080p.WEB-DL.H264-GRP.mkv");
     await organize(series, "Severance.S01E02.1080p.WEB-DL.H264-GRP.mkv");
     const q = h.engine.enqueued.filter((x) => x.task === SCRAPE_TASK);
-    expect(q[0]!.key).toBe(`${SCRAPE_TASK}:record:${rec}:0`);
+    // 照 MP:对作品目录刮,集只刮本次整理的文件;两批还没开始时合成一条
+    expect(q[0]!.key.startsWith(`${SCRAPE_TASK}:record:${rec}:`)).toBe(true);
     const first = await h.db
       .selectFrom("scrape_record")
       .selectAll()

apps/server/test/file-manager/organize-in-place.test.ts

@@ -3,6 +3,7 @@ import { mkdir, readdir, readFile, stat, writeFile } from "node:fs/promises";
 import { join } from "node:path";
 import { afterAll, beforeAll, describe, expect, it } from "vitest";
 import { executeRecord } from "../../src/modules/acquisition-organize/services/organize.ts";
+import { runScrapeRecord } from "../../src/modules/acquisition-organize/services/scrape.ts";
 import { createFmHarness, type FmHarness, key, pgAvailable, seedEntity } from "./harness.ts";
 
 const skip = !(await pgAvailable());
@@ -100,6 +101,37 @@ describe.skipIf(skip)("file manager in-place import", () => {
       .execute();
     expect(units.some((u) => u.season_number === 2 && u.episode_number === 7)).toBe(true);
   });
+  it("剧集原地入库后照 MP 刮削:剧集根目录与季目录写元数据,集只刮本次登记的", async () => {
+    const title = `Show-${randomUUID().slice(0, 8)}`;
+    const entity = await seedEntity(h.db, { title, kind: "series", year: 2022 });
+    const at = new Date().toISOString();
+    await h.db
+      .updateTable("media_entity")
+      .set({
+        metadata: JSON.stringify({ title: { value: title, source: "tmdb", language: "zh-CN", at, locked: false } }),
+      })
+      .where("id", "=", entity)
+      .execute();
+    await mkdir(join(h.lib, title, "Season 1"), { recursive: true });
+    const ep1 = `${title}/Season 1/original.name.S01E01.mkv`;
+    const ep2 = `${title}/Season 1/original.name.S01E02.mkv`;
+    for (const p of [ep1, ep2]) await writeFile(join(h.lib, p), "video");
+    const r = await queue([{ root: h.libRoot, path: ep1, mediaEntityId: entity }], { scrapeAfterImport: true });
+    expect((await executeRecord(h.module.deps().ao, r.recordId)).status).toBe("succeeded");
+    const recs = await h.db
+      .selectFrom("scrape_record")
+      .select(["id", "path", "videos"])
+      .where("organization_record_id", "=", r.recordId)
+      .execute();
+    expect(recs).toMatchObject([{ path: title, videos: [ep1] }]);
+    await runScrapeRecord(h.module.deps().ao, recs[0]!.id);
+    const show = await readdir(join(h.lib, title));
+    const season = await readdir(join(h.lib, title, "Season 1"));
+    expect(show).toContain("tvshow.nfo");
+    expect(season).toContain("season.nfo");
+    expect(season).toContain("original.name.S01E01.nfo");
+    expect(season).not.toContain("original.name.S01E02.nfo");
+  });
   it("scrapes only the newly registered file, retaining its organization association on resume", async () => {
     const s = await source("Scrape");
     await h.db
@@ -118,10 +150,13 @@ describe.skipIf(skip)("file manager in-place import", () => {
     await executeRecord(h.module.deps().ao, r.recordId);
     const scrapes = await h.db
       .selectFrom("scrape_record")
-      .select(["path", "organization_record_id"])
+      .select(["path", "organization_record_id", "videos"])
       .where("organization_record_id", "=", r.recordId)
       .execute();
-    expect(scrapes).toEqual([{ path: s.path, organization_record_id: r.recordId }]);
+    // 照 MP:对作品目录刮(目录海报 / 背景照常),视频只刮本次登记的这一个
+    expect(scrapes).toEqual([
+      { path: s.path.split("/").slice(0, -1).join("/"), organization_record_id: r.recordId, videos: [s.path] },
+    ]);
     expect(await readFile(join(h.lib, s.path.split("/")[0]!, "Unrelated.mkv"), "utf8")).toBe("unrelated video");
   });
   it("rejects an unregistered download directory without submitting a task", async () => {

apps/web/app/components/files/mock.ts

@@ -765,6 +765,7 @@ function mockScrape(b: {
       reason: null,
       files,
       organizationRecordId: null,
+      videos: null,
       taskId: `run-${seq}`,
       requestedAt: now,
       startedAt: now,

packages/contracts/src/scrape.ts

@@ -61,6 +61,8 @@ export const ScrapeRecord = z.object({
   files: z.array(ScrapeFileResult),
   /** 触发它的整理记录(入库后自动刮削) */
   organizationRecordId: Id.nullable(),
+  /** 只刮这些视频(整理后刮削:作品 / 季目录照常,集只刮本次整理的);null = 目录下全部 */
+  videos: z.array(z.string()).nullable(),
   /** 引擎 run id */
   taskId: z.string().nullable(),
   requestedAt: Timestamp,
@@ -82,6 +84,7 @@ export const ScrapeRecordRow = z.object({
   trigger: ScrapeTrigger,
   mode: ScrapeMode,
   organizationRecordId: Id.nullable(),
+  videos: z.array(z.string()).nullable(),
   engineRunId: z.string().nullable(),
   actorType: z.string().nullable(),
   actorId: z.string().nullable(),

packages/db/migrations/0062_scrape_record_videos.ts

@@ -0,0 +1,13 @@
+import type { Kysely } from "kysely";
+import { execDdl } from "../src/migration-helpers.ts";
+
+// 整理后刮削照 MP:对作品目录刮削(作品 / 季的 NFO 与图片照常初始化),集只刮本次整理的文件。
+// videos 为库根相对路径;null = 按目录布局刮全部视频(文件管理、作品详情的手动刮削)。
+// biome-ignore lint/suspicious/noExplicitAny: 迁移不依赖应用表类型
+export async function up(db: Kysely<any>): Promise<void> {
+  await execDdl(db, "alter table scrape_record add column if not exists videos text[]");
+}
+// biome-ignore lint/suspicious/noExplicitAny: 同上
+export async function down(db: Kysely<any>): Promise<void> {
+  await execDdl(db, "alter table scrape_record drop column if exists videos");
+}

packages/db/src/types.ts

@@ -1362,6 +1362,7 @@ export interface ScrapeRecordTable {
   files: Generated<Json>;
   created_at: GeneratedTimestamptz;
   updated_at: GeneratedTimestamptz;
+  videos: string[] | null;
 }
 
 export interface SecretTable {