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]>

apps/web/app/components/library/identity-enrich.tsx

@@ -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>
                 ))}

apps/web/app/components/library/labels.ts

@@ -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" };