fix(sites): integrate shared download access and register synced adapters

apps/server/src/modules/acquisition-organize/ports.ts

@@ -9,12 +9,10 @@ import {
   fetchEpisodeGroup,
   fetchRichDetail,
   fetchRichSeason,
-  getSiteAdapter,
   getSiteDefinition,
   groupSeasonAsRichSeason,
   type ProbeResult,
   qbAuthConfig,
-  type SecretResolver,
   SiteError,
   type TmdbRichDetail,
   type TmdbRichSeasonDetail,
@@ -200,20 +198,8 @@ export function createPorts(
       const siteMeta = (raw._site ?? {}) as Record<string, unknown>;
       const defId = [str(siteMeta.adapterSiteId), site.key].find((k) => k && getSiteDefinition(k));
       if (!defId) throw new SiteError(site.id, "unsupported", `站点 ${site.key} 没有可用的适配器定义`);
-      const accounts = await ctx.db
-        .selectFrom("site_account")
-        .select(["auth_kind", "secret_ref"])
-        .where("site_id", "=", site.id)
-        .where("enabled", "=", true)
-        .execute();
-      const secrets: SecretResolver = {
-        async resolve({ kind }) {
-          const want = kind === "apiKey" ? "api_key" : kind;
-          const acc = accounts.find((a) => a.auth_kind === want);
-          return (await resolveSecret(acc?.secret_ref ?? null)) ?? null;
-        },
-      };
-      const adapter = getSiteAdapter(defId, { secrets, baseUrl: site.base_url });
+      // 与搜索/详情共用站点出口、账号会话、UA 和跨进程限流;API 基址仍取站点配置。
+      const adapter = rt().siteAdapter({ ...site, key: defId });
       const t = await adapter.downloadTarget(release.site_release_id);
       if (t.kind === "torrent")
         return { kind: "file", bytes: t.bytes, ...(t.fileName ? { filename: t.fileName } : {}) };

apps/server/src/modules/sources-identity/runtime.ts

@@ -199,6 +199,13 @@ export class SourcesRuntime {
         const pick = typeof ua === "function" ? ua() : ua !== undefined ? Promise.resolve(ua) : this.defaultUserAgent();
         headers.set("user-agent", await pick);
       }
+      const credentials = init?.credentials ?? (input instanceof Request ? input.credentials : undefined);
+      if (credentials === "omit") {
+        // 签名下载不复用账号 Cookie 或浏览器 profile,包括 cf_clearance。
+        headers.delete("cookie");
+        headers.delete("authorization");
+        return base(input, { ...init, headers, credentials });
+      }
       if (!siteId) return base(input, { ...init, headers });
       const url = input instanceof Request ? input.url : String(input);
       const method = init?.method ?? (input instanceof Request ? input.method : "GET");

apps/server/src/modules/sources-identity/service/account.ts

@@ -413,10 +413,11 @@ export async function refreshSiteAccount(
   const profile = (def.umt as { profile?: Record<string, string> }).profile;
   const vipLevels = (def.umt as { vipLevels?: string[] }).vipLevels;
   // 馒头 API 基址与适配器一致:站点配置 > 定义里的 apiBaseUrl;不再由站点地址推 api.<根域>(IP / 本地地址会拼成非法 URL)
+  const siteConfig = (site.config ?? {}) as { apiBaseUrl?: string; userdataPath?: string };
   const apiUrl =
     def.umt.kind === "mteam"
-      ? (site.config as { apiBaseUrl?: string })?.apiBaseUrl || mteamApiBase(def as MteamDefinition)
-      : null;
+      ? siteConfig.apiBaseUrl || mteamApiBase(def as MteamDefinition)
+      : siteConfig.apiBaseUrl || ("api_url" in def ? def.api_url : null);
 
   const fetchUserData = async (
     cred: { cookie?: string | null; apikey?: string | null },
@@ -425,6 +426,7 @@ export async function refreshSiteAccount(
     const info = await refreshSiteUserData({
       name: site.name,
       url: site.base_url,
+      ...(siteConfig.userdataPath ? { userdataPath: siteConfig.userdataPath } : {}),
       schema: (def as { schema?: string }).schema ?? null,
       public: true,
       cookie: cred.cookie ?? null,

apps/server/test/site-fetch-credentials.test.ts

@@ -0,0 +1,22 @@
+import { expect, it, vi } from "vitest";
+import { SourcesRuntime } from "../src/modules/sources-identity/runtime.ts";
+
+it("签名下载的 credentials omit 不复用账号或浏览器Cookie", async () => {
+  let sent = new Headers();
+  const rt = new SourcesRuntime({} as never, {
+    fetch: async (_input, init) => {
+      sent = new Headers(init?.headers);
+      return new Response("subtitle bytes");
+    },
+  });
+  const browser = vi.spyOn(rt.cloudflare, "handle");
+  const fetch = rt.siteFetch({ id: "original-site", proxy: false }, undefined, "Site User Agent");
+  const headers = { cookie: "account-cookie", authorization: "account-token" };
+  const url = "https://signed-cdn.example/file?credential=ephemeral";
+  const response = await fetch(url, { headers, credentials: "omit" });
+  expect(await response.text()).toBe("subtitle bytes");
+  expect(sent.has("cookie")).toBe(false);
+  expect(sent.has("authorization")).toBe(false);
+  expect(sent.get("user-agent")).toBe("Site User Agent");
+  expect(browser).not.toHaveBeenCalled();
+});

apps/server/test/sources-identity/mteam-tasks.test.ts

@@ -1,6 +1,7 @@
 import { TaskRegistry } from "@umt/tasks";
 import { createMockMteamApi, type MockMteamApi } from "@umt/testkit";
 import { afterAll, beforeAll, describe, expect, it } from "vitest";
+import { createPorts } from "../../src/modules/acquisition-organize/ports.ts";
 import { snapshotSite } from "../../src/modules/sources-identity/service/snapshot.ts";
 import { createHarness, type Harness, pgAvailable } from "./harness.ts";
 
@@ -50,6 +51,27 @@ describe.skipIf(!up)("M-Team 快照与任务注册", () => {
     expect(JSON.stringify(rows)).not.toContain(api.apiKey);
   });
 
+  it("下载种子沿用配置的 API 基址，并与站点访问共用 PG 限流桶", async () => {
+    const site = await h.db
+      .updateTable("site")
+      .set({ rate_limit: JSON.stringify({ minIntervalMs: 10, burst: 1 }) })
+      .where("id", "=", siteId)
+      .returningAll()
+      .executeTakeFirstOrThrow();
+    const release = await h.db
+      .selectFrom("source_release")
+      .selectAll()
+      .where("site_id", "=", siteId)
+      .executeTakeFirstOrThrow();
+    const before = api.issuedTokens;
+    const source = await createPorts(h.ctx).torrentSource({ release, site });
+    expect(source.kind).toBe("file");
+    expect(api.issuedTokens).toBe(before + 1);
+    expect(
+      await h.db.selectFrom("rate_limit_bucket").select("key").where("key", "=", `site:${siteId}`).executeTakeFirst(),
+    ).toBeTruthy();
+  });
+
   it("key 失效:账号标失效 + 待处理,快照结果 failed", async () => {
     api.setFailMode("banned_key");
     const r = await snapshotSite(h.rt, { siteId, kind: "manual" });

apps/server/vitest.config.ts

@@ -23,6 +23,7 @@ const UNIT = [
   "test/mapping-compat.test.ts",
   "test/modules.test.ts",
   "test/sse.test.ts",
+  "test/site-fetch-credentials.test.ts",
   "test/static.test.ts",
   "test/system-resources.test.ts",
 ];

apps/web/app/components/subtitles/subtitle-dialog.tsx

@@ -49,7 +49,7 @@ export const SUBTITLE_LANGUAGE_LABEL: Record<SubtitleLanguage, string> = {
   ko: "韩文",
   unknown: "未知",
 };
-const SOURCE_LABEL = { upload: "上传", subhd: "SubHD", zimuku: "字幕库" } as const;
+const SOURCE_LABEL = { upload: "上传", subhd: "SubHD", zimuku: "字幕库", mteam: "馒头" } as const;
 const SKIP = "__skip__";
 const ACCEPT = ".srt,.ass,.ssa,.vtt,.sub,.sbv,.zip";
 

packages/adapters/src/sites/definitions/definitions.test.ts

@@ -32,9 +32,9 @@ const PUBLIC = ["anibt", "animegarden", "bangumimoe", "mikan", "nyaa"];
 describe("站点定义:加载时 zod 校验", () => {
   const ids = Object.keys(MOVIECLAW_DEFINITION_JSON).sort();
 
-  it("MovieClaw v0.27.0 全部 23 站转换结果 + 公开 BT 站均在注册表中", () => {
+  it("MovieClaw v0.27.0 全部 23 站、公开 BT 站与手写 Milkie 定义均在注册表中", () => {
     expect(ids).toHaveLength(23);
-    expect(listSiteDefinitions().map((d) => d.id)).toEqual([...ids, ...PUBLIC].sort());
+    expect(listSiteDefinitions().map((d) => d.id)).toEqual([...ids, ...PUBLIC, "milkie"].sort());
   });
 
   it.each(ids)("%s:schema 校验通过,provenance 指向 MovieClaw v0.27.0 MIT", (id) => {

packages/adapters/src/sites/definitions/index.ts

@@ -8,6 +8,7 @@ import type { MteamDefinition } from "../mteam/definition.ts";
 import type { IndexerDefinition } from "../spider/schema.ts";
 import type { SpecializedSiteDefinition } from "../spiders/types.ts";
 import { MOVIECLAW_DEFINITION_JSON } from "./manifest.generated.ts";
+import milkieJson from "./milkie.json" with { type: "json" };
 import { PUBLIC_SITE_DEFINITION_JSON } from "./public.ts";
 import { parseIndexerDefinition, parseMteamDefinition, parseSpecializedSiteDefinition } from "./schema.ts";
 
@@ -27,13 +28,12 @@ export function loadIndexerDefinition(json: unknown): IndexerDefinition {
   return parseIndexerDefinition(json);
 }
 
-/** 随产品发布的全部定义(已校验),键为站点 id:MovieClaw 转换的 PT 站 + 手写的公开 BT 站 */
+/** 随产品发布的全部定义(已校验),键为站点 id:MovieClaw 转换站、手写 API 站与公开 BT 站 */
 export const BUILTIN_SITE_DEFINITIONS: Readonly<Record<string, SiteDefinition>> = Object.freeze(
   Object.fromEntries(
-    Object.entries({ ...MOVIECLAW_DEFINITION_JSON, ...PUBLIC_SITE_DEFINITION_JSON }).map(([id, json]) => [
-      id,
-      loadSiteDefinition(json),
-    ]),
+    Object.entries({ ...MOVIECLAW_DEFINITION_JSON, ...PUBLIC_SITE_DEFINITION_JSON, milkie: milkieJson }).map(
+      ([id, json]) => [id, loadSiteDefinition(json)],
+    ),
   ),
 );
 

packages/adapters/src/sites/definitions/milkie.json

@@ -0,0 +1,25 @@
+{
+  "id": "milkie",
+  "name": "Milkie",
+  "domain": "https://milkie.cc/",
+  "parser": "Milkie",
+  "enabled": false,
+  "provenance": {
+    "source": "handwritten",
+    "ref": "v3.2.4",
+    "commit": "3b156a545f14fea0285cd3ed3ea4aa49d97faa57",
+    "license": "GPL-3.0-only",
+    "file": "app/modules/indexer/spider/milkie.py",
+    "note": "按公开 API 行为重写;凭据由 SecretResolver 注入,未做真实站点验收"
+  },
+  "umt": {
+    "version": "2026-11-05",
+    "kind": "api",
+    "authSupported": ["apikey"],
+    "displayName": "Milkie",
+    "timezone": "+00:00",
+    "rateLimit": { "minIntervalMs": 2000, "burst": 2 },
+    "capabilities": { "list": true, "search": true, "detail": false, "download": true, "rss": false },
+    "categoryNames": { "1": "电影", "2": "电视剧", "3": "音乐" }
+  }
+}

packages/adapters/src/sites/definitions/schema.ts

@@ -216,6 +216,8 @@ export const IndexerDefinitionSchema = z.strictObject({
 export const MteamDefinitionSchema = z.strictObject({
   ...BaseDefinitionShape,
   parser: z.literal("mTorrent"),
+  api_domain: z.string().min(1).optional(),
+  api_host: z.string().min(1).optional(),
   timeout: z.number().positive(),
   umt: z.strictObject({
     version: z.string().min(1),
@@ -242,6 +244,7 @@ export const SpecializedSiteDefinitionSchema = z.strictObject({
     "TNodeSpider",
     "TorrentLeech",
     "Yema",
+    "Milkie",
     "Nyaa",
     "Mikan",
     "AniBT",
@@ -251,6 +254,8 @@ export const SpecializedSiteDefinitionSchema = z.strictObject({
   public: z.boolean().optional(),
   category: CategoryConfigSchema.optional(),
   api_url: z.string().url().optional(),
+  api_domain: z.string().min(1).optional(),
+  api_host: z.string().min(1).optional(),
   umt: z.strictObject({
     version: z.string().min(1),
     kind: z.literal("api"),
@@ -258,6 +263,7 @@ export const SpecializedSiteDefinitionSchema = z.strictObject({
     timezone: TimezoneSchema,
     rateLimit: RateLimitSchema,
     capabilities: CapabilitiesSchema,
+    authSupported: z.array(z.enum(["cookie", "credential", "apikey"])).optional(),
     categoryNames: z.record(z.string(), z.string()).optional(),
     apiBaseUrl: z.string().url().optional(),
   }),

packages/adapters/src/sites/index.ts

@@ -150,6 +150,9 @@ export {
   type SitePromotion,
   type SiteRelease,
   type SiteReleaseDetail,
+  type SiteSubtitle,
+  type SiteSubtitleFile,
+  type SiteSubtitlePage,
   staticSecretResolver,
 } from "./types.ts";
 export * from "./user-parsers/index.ts";

packages/adapters/src/sites/mteam/adapter.ts

@@ -186,7 +186,12 @@ export function createMteamAdapter(input: MteamDefinition, deps: SiteAdapterDeps
         throw new SiteError(def.id, "rate_limited", `genDlToken ${id}:未返回下载链接(可能未认证或达到下载频率限制)`);
       }
       // MP:涉及解析地址的不使用 Cookie 下载;直链含一次性签名,只在内存中使用
-      const res = await siteFetch(http, url, { method: "GET", ...(opts.signal ? { signal: opts.signal } : {}) });
+      const res = await siteFetch(http, url, {
+        method: "GET",
+        credentials: "omit",
+        referrerPolicy: "no-referrer",
+        ...(opts.signal ? { signal: opts.signal } : {}),
+      });
       if (!res.ok) {
         await res.body?.cancel().catch(() => {});
         throw new SiteError(

packages/adapters/src/sites/mteam/mteam.test.ts

@@ -17,11 +17,20 @@ afterEach(async () => {
   api = undefined;
 });
 
-function adapterFor(m: MockMteamApi, key: string | null = m.apiKey, limiter?: SiteRateLimiter): SiteAdapter {
+function adapterFor(
+  m: MockMteamApi,
+  key: string | null = m.apiKey,
+  limiter?: SiteRateLimiter,
+  requests?: RequestInit[],
+): SiteAdapter {
   return createMteamAdapter(mteamDefinition, {
     baseUrl: m.baseUrl,
     secrets: staticSecretResolver(key ? { mteam: { apiKey: key } } : {}),
     limiter: limiter ?? createSiteRateLimiter({ minIntervalMs: 0 }),
+    fetch: (input, init) => {
+      requests?.push(init ?? {});
+      return globalThis.fetch(input, init);
+    },
   });
 }
 
@@ -129,13 +138,15 @@ describe("M-Team API 适配器(mock)", () => {
 
   it("下载令牌:现场 genDlToken → 一次性直链 → .torrent,infohash 校验", async () => {
     api = await createMockMteamApi();
-    const a = adapterFor(api);
+    const requests: RequestInit[] = [];
+    const a = adapterFor(api, api.apiKey, undefined, requests);
     const t = await a.downloadTarget("912233");
     expect(api.issuedTokens).toBe(1);
     expect(api.requests.map((r) => r.path.split("?")[0])).toEqual(["/api/torrent/genDlToken", "/api/rss/dlv2"]);
     // MP:genDlToken 的 id 走 query 参数;直链下载不带 x-api-key
     expect(api.requests[0]!.path).toBe("/api/torrent/genDlToken?id=912233");
     expect(api.requests[1]!.headers["x-api-key"]).toBeUndefined();
+    expect(requests[1]).toMatchObject({ credentials: "omit", referrerPolicy: "no-referrer" });
     expect(t.kind).toBe("torrent");
     if (t.kind === "torrent") expect(t.infoHash).toBe(api.infoHashOf("912233"));
     // 每次下载都重新签发令牌
@@ -162,15 +173,27 @@ describe("M-Team API 适配器(mock)", () => {
 
   it("非 JSON 维护页 → site_maintenance;业务限流/HTTP 429 → rate_limited 并顺延限流器", async () => {
     api = await createMockMteamApi();
-    const limiter = createSiteRateLimiter({ minIntervalMs: 0 });
+    let clock = 0;
+    const waits: number[] = [];
+    const limiter = createSiteRateLimiter({
+      minIntervalMs: 0,
+      now: () => clock,
+      sleep: async (ms) => {
+        waits.push(ms);
+        clock += ms;
+      },
+    });
     const a = adapterFor(api, api.apiKey, limiter);
     api.setFailMode("non_json");
     await expectSiteError(a.list(1), "site_maintenance");
     api.setFailMode("rate_limit");
-    await expectSiteError(a.list(1), "rate_limited");
+    const business = await expectSiteError(a.list(1), "rate_limited");
+    expect(business.retryAfterMs).toBe(3600_000);
+    expect(limiter.tryAcquire()).toBe(false);
     api.setFailMode("http_429");
     const e = await expectSiteError(a.list(1), "rate_limited");
     expect(e.retryAfterMs).toBe(5000);
+    expect(waits).toEqual([3600_000]);
     expect(limiter.tryAcquire()).toBe(false);
   });
 

packages/adapters/src/sites/spider/adapter.ts

@@ -17,7 +17,7 @@ import { inspectTorrentIsolated } from "../../isolation.ts";
 import { redactText } from "../../rss/redact.ts";
 import { challengeOutcomeText, challengeRuleOfUnrecognizedPage } from "../challenge.ts";
 import { buildHttpContext, type SiteAdapterDeps } from "../deps.ts";
-import { buildFailureSample, type SiteRequestKind } from "../failure-sample.ts";
+import { buildFailureSample, redactPageHtml, type SiteRequestKind } from "../failure-sample.ts";
 import { readBody, readText, siteFetch } from "../http.ts";
 import {
   type DownloadTarget,
@@ -155,7 +155,7 @@ export function createSpiderSiteAdapter(input: IndexerDefinition, deps: SiteAdap
   /** 页面认不出来时的错误:像人机验证就报 challenge,否则 parse_error;都附页面样本 */
   function unrecognized(page: FetchedPage, what: string, kind: SiteRequestKind, parseMessage: string): SiteError {
     const rule = challengeRuleOfUnrecognizedPage(page.html);
-    const title = redactText(loadHtml(page.html)("title").first().text().replace(/\s+/g, " ").trim()).slice(0, 120);
+    const title = redactPageHtml(loadHtml(page.html)("title").first().text()).replace(/\s+/g, " ").trim().slice(0, 120);
     const message = rule
       ? `${what}：${PAGE_LABEL[kind]}需要人机验证（Cloudflare 等），${challengeOutcomeText("unhandled")}`
       : `${parseMessage}（HTTP ${page.status}${title ? `，页面标题：${title}` : ""}；页面样本见站点页技术信息）`;

packages/adapters/src/sites/spider/pagination-rss.test.ts

@@ -1,7 +1,7 @@
 import { createMockFetch } from "@umt/testkit";
 import { describe, expect, it } from "vitest";
 import { createSiteRateLimiter } from "../rate-limit.ts";
-import { staticSecretResolver } from "../types.ts";
+import { SiteError, staticSecretResolver } from "../types.ts";
 import { createSpiderSiteAdapter } from "./adapter.ts";
 import type { IndexerDefinition } from "./schema.ts";
 import { SiteSpider } from "./spider.ts";
@@ -78,13 +78,19 @@ describe("通用索引XML与本地分页", () => {
 
   it("无法识别的200页面错误包含安全标题", async () => {
     const mock = createMockFetch({
-      "/search": { body: "<html><head><title>Search temporarily unavailable</title></head><body>Later</body></html>" },
+      "/search": {
+        body: "<html><head><title>Search temporarily unavailable passkey=private-key session=session-secret</title></head><body>Later</body></html>",
+      },
     });
     const adapter = createSpiderSiteAdapter(def, {
       fetch: mock.fetch,
       secrets: staticSecretResolver({}),
       limiter: createSiteRateLimiter({ minIntervalMs: 0 }),
     });
-    await expect(adapter.search("special")).rejects.toThrow("HTTP 200，页面标题：Search temporarily unavailable");
+    const error = await adapter.search("special").catch((e: unknown) => e);
+    expect(error).toBeInstanceOf(SiteError);
+    expect((error as SiteError).message).toContain("HTTP 200，页面标题：Search temporarily unavailable");
+    expect((error as SiteError).message).not.toMatch(/private-key|session-secret/);
+    expect((error as SiteError).sample?.message).not.toMatch(/private-key|session-secret/);
   });
 });

packages/adapters/src/sites/spiders/common.ts

@@ -2,7 +2,7 @@
  * SPDX-License-Identifier: GPL-3.0-only
  * 移植自 MoviePilot(https://github.com/jxxghp/MoviePilot,GPLv3),本文件按 GPLv3 条款提供。
  * @mp-source app/chain/download/submission.py @ e6820a8
- * @mp-source app/foundation/temporal.py @ e6820a8
+ * @mp-source app/foundation/temporal.py @ 3b156a5
  *
  * 专用 spider 共用 IO 与工具:
  * - `submission._resolve_indirect_download_url` → `resolveIndirectDownloadUrl`(配置对象在内存中构造,不再 base64 编进 enclosure)

packages/adapters/src/sites/spiders/types.ts

@@ -58,6 +58,7 @@ export interface SpecializedSiteDefinition {
     timezone: string;
     rateLimit: { minIntervalMs: number; burst: number };
     capabilities: SiteCapabilities;
+    authSupported?: Array<"cookie" | "credential" | "apikey">;
     categoryNames?: Record<string, string>;
     /** 覆盖 API 基址(测试指向 mock,或站点换域) */
     apiBaseUrl?: string;

packages/adapters/src/sites/types.ts

@@ -170,6 +170,31 @@ export interface SiteCallOptions {
   signal?: AbortSignal;
 }
 
+/** API 站字幕结果只保存稳定 ID;下载签名在 downloadSubtitle 时现场获取。 */
+export interface SiteSubtitle {
+  id: string;
+  title: string;
+  fileName: string | null;
+  language: string | null;
+  sizeBytes: number | null;
+  uploadedAt: string | null;
+  downloads: number | null;
+  uploader: string | null;
+  torrentId: string | null;
+  detailUrl: string | null;
+}
+
+export interface SiteSubtitlePage {
+  items: SiteSubtitle[];
+  page: number;
+  totalPages: number | null;
+}
+
+export interface SiteSubtitleFile {
+  bytes: Uint8Array;
+  fileName: string | null;
+}
+
 export interface SiteAdapter {
   readonly id: string;
   readonly kind: SiteKind;
@@ -182,6 +207,9 @@ export interface SiteAdapter {
   detail(siteReleaseId: string, opts?: SiteCallOptions): Promise<SiteReleaseDetail>;
   /** 下载前现场取目标:会复核登录态/令牌,校验返回确为 .torrent */
   downloadTarget(siteReleaseId: string, opts?: SiteCallOptions): Promise<DownloadTarget>;
+  searchSubtitles?(query: string, opts?: SiteCallOptions & { page?: number }): Promise<SiteSubtitlePage>;
+  listSubtitles?(torrentId: string, opts?: SiteCallOptions): Promise<SiteSubtitle[]>;
+  downloadSubtitle?(subtitleId: string, opts?: SiteCallOptions): Promise<SiteSubtitleFile>;
   /** 不抛错;把失败归类为状态 */
   healthCheck(opts?: SiteCallOptions): Promise<SiteHealth>;
 }

packages/adapters/src/sites/user-parsers/registry.ts

@@ -1,7 +1,7 @@
 /**
  * SPDX-License-Identifier: GPL-3.0-only
  * 移植自 MoviePilot(https://github.com/jxxghp/MoviePilot,GPLv3),本文件按 GPLv3 条款提供。
- * @mp-source app/modules/indexer/__init__.py @ e6820a8
+ * @mp-source app/modules/indexer/__init__.py @ 3b156a5
  *
  * 站点用户信息解析器注册表与刷新入口。
  * - `IndexerModule.__init__` 里 `ModuleHelper.load('app.modules.indexer.parser', filter=有 schema)` → `USER_PARSER_CLASSES`