feat(identity): source-aware short-drama enrichment with snippet mode and hit rate (META-2)

- classify enrich sources: encyclopedia, platform pages (hongguo, kuaikaw, video platforms), douban (search snippet only) and other sites' search snippets (snippet mode, pages never opened); one extractor per source, fields carry source kind and confidence, merged by noisy-OR across hosts - new cast field (applied as MediaCredit with placeholder ids) - robots.txt check (cached per host 1 day) and login-wall/captcha give-up - per-enrichment source attempts + GET /identity/enrich-stats, 30-day hit-rate table under settings -> metadata -> identity - setting identity.enrichSnippetAllSites (default on); default domains add baike.com and kuaikaw.cn - replay fixtures from the real run on the 4 user-library short dramas (sanitized Brave responses, hongguo page, robots.txt) Co-Authored-By: Claude Opus 5.5 <[email protected]>
2026-09-30 18:40:19 +08:00
1 parent 081db4651c
commit 9489e40e0c
@@ -4,6 +4,7 @@ import {
  type EnrichApplyInput,
  EnrichApplyResult,
  type EnrichRequest,
+  EnrichStats,
  ExternalIdentity,
  type IdentityCandidateReason,
  type IdentityCandidateStatus,
@@ -28,6 +29,7 @@ export const identityKeys = {
  identities: (f: IdentityFilters) => ["identity", "external", f] as const,
  revisions: (id: string) => ["identity", "revisions", id] as const,
  aliases: (f: AliasFilters) => ["identity", "aliases", f] as const,
+  enrichStats: (days: number) => ["identity", "enrich-stats", days] as const,
 };

 export type PoolFilters = { status?: IdentityCandidateStatus; reason?: IdentityCandidateReason; q?: string };
@@ -159,6 +161,15 @@ export function useDeleteAlias() {
  });
 }

+/** 元数据补全命中率(最近 N 天,按来源类别;META-2) */
+export function useEnrichStats(days = 30) {
+  return useQuery({
+    queryKey: identityKeys.enrichStats(days),
+    queryFn: ({ signal }) => getParsed(EnrichStats, "/identity/enrich-stats", { days }, signal),
+    staleTime: 60_000,
+  });
+}
+
 /** 元数据补全建议(metadata.enrich;不改作品,同一标题 7 天内返回上次结果) */
 export function useEnrichPool() {
  return useMutation({
@@ -10,13 +10,21 @@ import { Button } from "~/components/ui/button.tsx";
 import { Label } from "~/components/ui/label.tsx";
 import { Switch } from "~/components/ui/switch.tsx";
 import { errorMessage } from "~/lib/api.ts";
-import { ENRICH_FIELD_LABEL, ENRICH_SKIPPED, EVIDENCE_LEVEL } from "./labels.ts";
+import { ENRICH_FIELD_LABEL, ENRICH_SKIPPED, ENRICH_SOURCE_KIND, EVIDENCE_LEVEL } from "./labels.ts";
 import { Chip } from "./parts.tsx";

 // META-1 待识别资源 → 元数据补全:先看建议(网页搜索 + 白名单网页 + Agent 抽取,不改作品),再应用(建本地临时作品并关联)。

 type Suggestion = NonNullable<MetadataEnrichOutput["suggestion"]>;
-type Value = { value: unknown; sourceUrl: string | null; evidenceLevel: string; verbatim: boolean };
+type Value = {
+  value: unknown;
+  sourceUrl: string | null;
+  evidenceLevel: string;
+  verbatim: boolean;
+  sourceKind?: string;
+  confidence?: number;
+  hosts?: number;
+};

 const hostOf = (u: string | null) => {
  if (!u) return null;
@@ -69,6 +77,12 @@ function FieldRow({ k, s, enrichmentId }: { k: EnrichFieldName; s: Suggestion; e
            <div className="flex flex-wrap items-center gap-1.5">
              <Chip>{EVIDENCE_LEVEL[v.evidenceLevel] ?? v.evidenceLevel}</Chip>
              {v.verbatim ? null : <Chip>智能体归纳</Chip>}
+              {v.sourceKind ? <Chip>{ENRICH_SOURCE_KIND[v.sourceKind] ?? v.sourceKind}</Chip> : null}
+              {v.confidence !== undefined ? (
+                <span className="text-meta tabular-nums text-muted-foreground">
+                  置信度 {Math.round(v.confidence * 100)}%{v.hosts && v.hosts > 1 ? ` · ${v.hosts} 个网站` : ""}
+                </span>
+              ) : null}
              <SourceLink url={v.sourceUrl} />
            </div>
          </div>
@@ -84,6 +98,7 @@ const FIELDS: EnrichFieldName[] = [
  "year",
  "platform",
  "episodeCount",
+  "cast",
  "genre",
  "synopsis",
  "posterUrl",
@@ -134,7 +149,8 @@ export function EnrichSection({ item }: { item: IdentityPoolItem }) {
    >
      {!out ? (
        <p className="text-caption text-muted-foreground">
-          TMDB 查不到时，按标题搜索网页，只读平台和百科网站，抽取字段并附出处。先看建议，确认后才建作品。
+          TMDB
+          查不到时，按标题搜索网页：读平台和百科网站，其它网站只用搜索摘要，抽取字段并附出处和置信度。先看建议，确认后才建作品。
        </p>
      ) : (
        <div className="space-y-3">
@@ -162,6 +178,7 @@ export function EnrichSection({ item }: { item: IdentityPoolItem }) {
                    <SourceLink url={x.url} />
                    <span>
                      {x.fetched ? "已读取" : x.skipped ? (ENRICH_SKIPPED[x.skipped] ?? x.skipped) : "未读取"}
+                      {x.snippetUsed ? "，用了搜索摘要" : ""}
                    </span>
                  </li>
                ))}
@@ -281,6 +281,16 @@ export const ENRICH_FIELD_LABEL: Record<string, string> = {
  synopsis: "简介",
  posterUrl: "海报",
  genre: "类型",
+  cast: "主演",
+};
+
+/** 补全来源类别(META-2) */
+export const ENRICH_SOURCE_KIND: Record<string, string> = {
+  encyclopedia: "百科",
+  platform: "平台页",
+  douban: "豆瓣摘要",
+  snippet: "搜索摘要",
+  other: "其它网站",
 };

 /** 补全来源链接被跳过的原因 */
@@ -290,6 +300,9 @@ export const ENRICH_SKIPPED: Record<string, string> = {
  irrelevant: "标题对不上",
  limit: "超出读取条数",
  error: "读取失败",
+  snippet_only: "只用搜索摘要",
+  robots: "网站不允许抓取",
+  blocked: "要求登录或验证",
 };

 export const UNIT_KIND: Record<string, string> = { season: "季", episode: "集", special: "特别篇", part: "分段" };
@@ -246,14 +246,19 @@ export function CreditsStrip({ cast, crew }: { cast: MediaCredit[]; crew: MediaC
      {people.map((p) => (
        <li key={p.key} className="flex w-20 shrink-0 flex-col items-center text-center">
          <Avatar path={p.profilePath} name={p.name} />
-          <a
-            href={`https://www.themoviedb.org/person/${p.id}`}
-            target="_blank"
-            rel="noreferrer noopener"
-            className="mt-1.5 line-clamp-2 text-meta font-medium break-words text-foreground hover:underline"
-          >
-            {p.name}
-          </a>
+          {p.id > 0 ? (
+            <a
+              href={`https://www.themoviedb.org/person/${p.id}`}
+              target="_blank"
+              rel="noreferrer noopener"
+              className="mt-1.5 line-clamp-2 text-meta font-medium break-words text-foreground hover:underline"
+            >
+              {p.name}
+            </a>
+          ) : (
+            // 元数据补全写的主演没有 TMDB 人物 ID(负数占位),不链接
+            <span className="mt-1.5 line-clamp-2 text-meta font-medium break-words text-foreground">{p.name}</span>
+          )}
          {p.role ? <span className="line-clamp-2 text-micro break-words text-muted-foreground">{p.role}</span> : null}
        </li>
      ))}
@@ -1,6 +1,9 @@
 import { DEFAULT_SETTINGS } from "@umt/contracts";
 import { toast } from "sonner";
 import { useClearImageCache, useClearTmdbCache, useImageCache, useTmdbCache } from "~/api/cache.ts";
+import { useEnrichStats } from "~/api/identity.ts";
+import { DataTable } from "~/components/console/data-table.tsx";
+import { ENRICH_SOURCE_KIND } from "~/components/library/labels.ts";
 import { Button } from "~/components/ui/button.tsx";
 import { errorMessage } from "~/lib/api.ts";
 import { secretPatch } from "~/lib/secret.ts";
@@ -212,7 +215,14 @@ const IDENTITY_CARD: CardSpec = {
      name: "identity.enrichDomains",
      label: "补全可读取的网站",
      kind: "textarea",
-      hint: "每行一个域名，含子域名；只放平台和百科类网站。已登记站点的域名一律不读",
+      hint: "每行一个域名，含子域名；只放平台和百科类网站。豆瓣只用搜索摘要，不读页面；robots.txt 不允许的页面不读。已登记站点的域名一律不读",
+      wide: true,
+    },
+    {
+      name: "identity.enrichSnippetAllSites",
+      label: "也用其它网站的搜索摘要",
+      kind: "switch",
+      hint: "短剧多半只有聚合站收录。打开后，白名单以外网站的搜索摘要也用来抽年份、集数、主演、简介（只看搜索引擎给的摘要，不打开这些网页）",
      wide: true,
    },
  ],
@@ -221,6 +231,7 @@ const IDENTITY_CARD: CardSpec = {
    "identity.autoEnrich": s.identity.autoEnrich,
    "identity.enrichDailyBudget": n(s.identity.enrichDailyBudget),
    "identity.enrichDomains": s.identity.enrichDomains.join("\n"),
+    "identity.enrichSnippetAllSites": s.identity.enrichSnippetAllSites ?? true,
  }),
  toPatch: (v) => ({
    identity: {
@@ -231,10 +242,59 @@ const IDENTITY_CARD: CardSpec = {
        .split(/[\s,，]+/)
        .map((d) => d.trim().toLowerCase())
        .filter(Boolean),
+      enrichSnippetAllSites: v["identity.enrichSnippetAllSites"] === true,
    },
  }),
+  extra: () => <EnrichStatsTable />,
 };

+type StatRow = { kind: string; withResults: number; hits: number; rate: number | null };
+
+/** 最近 30 天补全命中率(按来源类别,META-2) */
+function EnrichStatsTable() {
+  const q = useEnrichStats(30);
+  const d = q.data;
+  return (
+    <div className="mb-3 space-y-2">
+      <p className="text-meta text-muted-foreground tabular-nums">
+        {d
+          ? `最近 ${d.days} 天元数据补全 ${d.total} 次，出了建议 ${d.suggested} 次${d.tracked < d.total ? `；有来源明细的 ${d.tracked} 次` : ""}`
+          : "读取补全命中率…"}
+      </p>
+      {d && d.tracked > 0 ? (
+        <DataTable<StatRow>
+          rows={d.sources}
+          rowKey={(r) => r.kind}
+          empty={{ title: "还没有补全记录" }}
+          columns={[
+            { key: "kind", header: "来源", cell: (r) => ENRICH_SOURCE_KIND[r.kind] ?? r.kind },
+            {
+              key: "withResults",
+              header: "有结果",
+              align: "right",
+              cell: (r) => <span className="tabular-nums">{r.withResults}</span>,
+            },
+            {
+              key: "hits",
+              header: "命中",
+              align: "right",
+              cell: (r) => <span className="tabular-nums">{r.hits}</span>,
+            },
+            {
+              key: "rate",
+              header: "命中率",
+              align: "right",
+              cell: (r) => (
+                <span className="tabular-nums">{r.rate === null ? "—" : `${Math.round(r.rate * 100)}%`}</span>
+              ),
+            },
+          ]}
+        />
+      ) : null}
+    </div>
+  );
+}
+
 export function MetadataSection() {
  return (
    <SettingsGate>