feat(organize): 文件级标题粗核,混合包里明显不是本作品的视频降为待确认

- 同一批视频里文件名标题与作品标题 / 原名 / 别名明显不符(没有足够共同词)的排除为 title_mismatch,不按本作品命名入库; - 只在同批至少一个文件对得上作品时才判断,剧集只看带 SxxEyy 的文件,用户逐文件指定过的不核; - title_mismatch 算「本该入库却被排除」:记录部分完成、开待处理、Move 不移除下载任务。 Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_01Cxwgnn2pbzQ7rXv3rq7fei
2026-10-03 23:33:41 +08:00
1 parent b40716d74f
commit c532060c83
@@ -20,6 +20,14 @@ import { scrapeAfterImport } from "./scrape.ts";
 import { slotComplete as sharedSlotComplete } from "./subscriptions.ts";
 import { downloadTasksUnder } from "./takeover.ts";

+/** 被排除待确认的视频原因(待处理摘要里显示) */
+const EXCLUDED_TEXT: Record<string, string> = {
+  episode_unresolved: "集数待确认",
+  no_target_dir: "没有季目录",
+  naming_failed: "命名失败",
+  title_mismatch: "标题不是本作品",
+};
+
 /** 登记资产/位置/绑定 + Slot/计划/下载任务 + 记录终态,同一事务;记录行锁作 fencing */
 export async function finalizeRecord(d: AoDeps, recordId: string): Promise<ExecuteResult> {
  const db = d.ctx.db;
@@ -350,13 +358,14 @@ export async function finalizeRecord(d: AoDeps, recordId: string): Promise<Execu
      .where("id", "=", recordId)
      .execute();
    if (unresolved.length > 0) {
-      const shown = unresolved
-        .slice(0, 5)
-        .map((f) => `${f.source_path.split("/").pop()}(${f.skip_reason?.replace(/^excluded:/, "")})`);
+      const shown = unresolved.slice(0, 5).map((f) => {
+        const code = f.skip_reason?.replace(/^excluded:/, "") ?? "";
+        return `${f.source_path.split("/").pop()}（${EXCLUDED_TEXT[code] ?? code}）`;
+      });
      await upsertDecision(d, trx, {
        kind: "other",
        title: `有 ${unresolved.length} 个视频没有整理`,
-        summary: `季集、季目录或命名没能确定，文件留在原处${
+        summary: `季集、季目录、命名没能确定，或文件标题不是本作品，文件留在原处${
          rec.action === "move" ? "，下载任务保留" : ""
        }：${shown.join("、")}${unresolved.length > shown.length ? " 等" : ""}`,
        subjectType: "organization_record",
@@ -24,6 +24,7 @@ import {
  resolveTargetDirectory,
  type SourceFile,
  sidecarStem,
+  titleMismatches,
 } from "@umt/domain";
 import { organizePlanToFileResults, toDomainPathBinding } from "@umt/mapping";
 import {
@@ -637,9 +638,22 @@ export async function planImport(d: AoDeps, req: ImportRequest): Promise<PlanRes
  const episodes = mediaType === "tv" ? await planFileEpisodes(d, entity, rel, dt) : null;
  const mappedOverrides: Record<string, DomainFileOverride> = {};
  const unresolvedEpisodes: NonNullable<FrozenPlan["unresolvedEpisodes"]> = {};
+  // 文件级标题粗核:混合包 / 合集里标题明显不是本作品的视频不按本作品命名入库(待确认 title_mismatch);
+  // 用户逐文件指定过的不核
+  const mismatched = await fileTitleMismatches(
+    d,
+    entity,
+    mediaType,
+    files.filter((f) => !overrideFor(req.fileOverrides, f.path) && fileKindOf(basename(f.path)) === "video"),
+  );
  for (const f of files) {
    let parsed = parseReleasePath(segmentsWithin(f.path, sourceRoot));
    if (parsed.fileKind !== "video") continue;
+    const wrongTitle = mismatched.get(f.path);
+    if (wrongTitle !== undefined) {
+      mappedOverrides[f.path] = { exclude: { reason: "title_mismatch", detail: wrongTitle } };
+      continue;
+    }
    const ov = overrideFor(req.fileOverrides, f.path);
    let resolution: EpisodeResolution | undefined;
    if (mediaType === "tv" && (ov?.season !== undefined || ov?.episodes)) {
@@ -1221,3 +1235,32 @@ export async function emitRecord(d: AoDeps, exec: Exec, recordId: string): Promi
    mediaEntityId: r.media_entity_id,
  });
 }
+
+/** 作品标题 / 原名 / 别名 → 这批视频里标题明显不是本作品的(路径 → 文件标题) */
+async function fileTitleMismatches(
+  d: AoDeps,
+  entity: Pick<MediaEntityRow, "id" | "title" | "original_title" | "metadata">,
+  mediaType: "movie" | "tv",
+  videos: readonly { path: string }[],
+): Promise<Map<string, string>> {
+  if (videos.length < 2) return new Map();
+  const aliases = await d.ctx.db
+    .selectFrom("title_alias")
+    .select("alias")
+    .where("media_entity_id", "=", entity.id)
+    .where("content_unit_id", "is", null)
+    .execute();
+  const meta = (entity.metadata ?? {}) as Record<string, { value?: unknown } | undefined>;
+  const names = [
+    entity.title,
+    entity.original_title,
+    meta.title?.value,
+    meta.originalTitle?.value,
+    ...aliases.map((a) => a.alias),
+  ].filter((n): n is string => typeof n === "string" && n.trim().length > 0);
+  return titleMismatches(
+    videos.map((v) => ({ path: v.path, name: basename(v.path) })),
+    names,
+    mediaType,
+  );
+}
@@ -0,0 +1,84 @@
+import { existsSync } from "node:fs";
+import { createTempTree, type TempTree } from "@umt/testkit";
+import { afterAll, beforeAll, describe, expect, it } from "vitest";
+import { executeRecord, planImport } from "../../src/modules/acquisition-organize/services/organize.ts";
+import { createHarness, type Harness, pgAvailable, seedBase, seedEntity } from "./harness.ts";
+
+// 文件级标题粗核:合集 / 混合包里标题明显不是本作品的视频不按本作品命名入库,
+// 降为待确认(排除原因 title_mismatch),记录部分完成并开待处理;用户逐文件指定过的不核。
+
+const hasPg = await pgAvailable();
+const SIZE = 200_000;
+const PACK = "Nolan.Pack";
+
+describe.skipIf(!hasPg)("整理:文件标题明显不是本作品时降为待确认", () => {
+  let h: Harness;
+  let lib: TempTree;
+  let src: TempTree;
+  let entity = "";
+
+  beforeAll(async () => {
+    lib = await createTempTree({}, { sentinel: true, prefix: "umt-tclib-" });
+    src = await createTempTree(
+      {
+        [`${PACK}/Inception.2010.1080p.BluRay.x264-GRP.mkv`]: SIZE,
+        [`${PACK}/Interstellar.2014.1080p.BluRay.x264-GRP.mkv`]: SIZE,
+      },
+      { prefix: "umt-tcsrc-" },
+    );
+    h = await createHarness("title_check");
+    await seedBase(h.db, { libraryPath: lib.root, cachePath: src.root, qbUrl: "http://127.0.0.1:9" });
+    await h.setSettings("organize", { scrapeAfterImport: false });
+    entity = await seedEntity(h.db, { kind: "movie", title: "盗梦空间", year: 2010 });
+    await h.db.updateTable("media_entity").set({ original_title: "Inception" }).where("id", "=", entity).execute();
+  }, 60_000);
+  afterAll(async () => {
+    await h?.close();
+    for (const t of [lib, src]) await t?.cleanup();
+  });
+
+  it("合集里的别的电影:排除 title_mismatch,不入库,记录部分完成并开待处理", async () => {
+    const plan = await planImport(h.d, {
+      sourcePaths: [src.path(PACK)],
+      mediaEntityId: entity,
+      action: "copy",
+      trigger: "auto",
+      actorType: "system",
+    });
+    expect(plan.status).toBe("running");
+    const ex = await executeRecord(h.d, plan.recordId!);
+    expect(ex.status).toBe("partial");
+    const files = await h.db
+      .selectFrom("organization_file_result")
+      .select(["source_path", "status", "skip_reason", "target_path"])
+      .where("organization_record_id", "=", plan.recordId!)
+      .execute();
+    const other = files.find((f) => f.source_path.includes("Interstellar"))!;
+    expect(other).toMatchObject({ status: "skipped", skip_reason: "excluded:title_mismatch", target_path: null });
+    const mine = files.find((f) => f.source_path.includes("Inception"))!;
+    expect(mine.status).toBe("succeeded");
+    expect(existsSync(src.path(`${PACK}/Interstellar.2014.1080p.BluRay.x264-GRP.mkv`))).toBe(true);
+    const decision = await h.db
+      .selectFrom("decision")
+      .select(["summary", "status"])
+      .where("dedupe_key", "=", `organize_excluded:${plan.recordId}`)
+      .executeTakeFirstOrThrow();
+    expect(decision.status).toBe("open");
+    expect(decision.summary).toContain("标题不是本作品");
+  });
+
+  it("用户逐文件指定过的不核", async () => {
+    const plan = await planImport(h.d, {
+      sourcePaths: [src.path(PACK)],
+      mediaEntityId: entity,
+      action: "copy",
+      trigger: "manual",
+      actorType: "user",
+      dryRun: true,
+      fileOverrides: {
+        [src.path(`${PACK}/Interstellar.2014.1080p.BluRay.x264-GRP.mkv`)]: { contentUnitIds: [] },
+      },
+    });
+    expect(plan.plan!.organize.excluded.some((x) => x.reason === "title_mismatch")).toBe(false);
+  });
+});