Files
UHD_Media_Tool/apps/web/app/components/settings/metadata.tsx
T
rootandClaude Opus 5.5 9489e40e0c 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

312 lines
11 KiB
TypeScript
Raw Blame History

This file contains ambiguous Unicode characters
This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.
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";
import { str } from "./resource-section.tsx";
import { type CardSpec, n, SettingsCard, SettingsGate } from "./settings-card.tsx";
// 元数据:TMDB API Key(只写,存 secret)、Base URL、语言回退链
const DEFAULT_LANGUAGES = DEFAULT_SETTINGS.metadata.languages;
/** "zh-CN, zh_sg en" → ["zh-CN","zh-SG","en"];不合法的项丢弃,全空回缺省 */
function parseLanguages(v: string): string[] {
const out = v
.split(/[,,、\s]+/)
.map((x) => x.trim().replace("_", "-"))
.filter(Boolean)
.map((x) => {
const [l, r] = x.split("-");
return r ? `${l!.toLowerCase()}-${r.toUpperCase()}` : l!.toLowerCase();
})
.filter((x) => /^[a-z]{2,3}(-[A-Z]{2})?$/.test(x));
return out.length ? [...new Set(out)].slice(0, 8) : [...DEFAULT_LANGUAGES];
}
const TMDB_CARD: CardSpec = {
anchor: "tmdb",
title: "TMDB",
description: "留空时用环境变量 TMDB_API_KEY / TMDB_ACCESS_TOKEN / TMDB_BASE_URL。",
fields: [
{
name: "metadata.tmdbApiKey",
label: "API Key",
kind: "secret",
secret: { set: (v) => v["metadata.tmdbApiKey:set"] === true, clearable: true },
hint: "v3 API Key 或 v4 Read Access Token",
wide: true,
},
{
name: "metadata.tmdbBaseUrl",
label: "Base URL",
kind: "url",
placeholder: "https://api.themoviedb.org",
hint: "自建反代时填写;留空用官方地址",
wide: true,
},
{
name: "metadata.languages",
label: "语言回退链",
kind: "text",
placeholder: DEFAULT_LANGUAGES.join(", "),
hint: "标题、简介和每集信息按顺序取第一个有值的语言;中文名取链中第一个中文值。逗号分隔,如 zh-CN, zh-SG",
wide: true,
},
{
name: "metadata.includeSpecials",
label: "包含特别篇(季 0)",
kind: "switch",
hint: "全量刷新和元数据回刷时,TMDB 有季 0 就一并拉取每集信息。特别篇不计入正片季的总集数和订阅完成判定",
wide: true,
},
],
fromSettings: (s) => ({
"metadata.tmdbApiKey": "",
"metadata.tmdbApiKey:set": !!s.metadata.tmdbApiKeySecretRef,
"metadata.tmdbBaseUrl": s.metadata.tmdbBaseUrl ?? "",
"metadata.languages": (s.metadata.languages ?? DEFAULT_LANGUAGES).join(", "),
"metadata.includeSpecials": s.metadata.includeSpecials !== false,
}),
toPatch: (v) => {
const key = secretPatch(v["metadata.tmdbApiKey"]);
return {
metadata: {
tmdbBaseUrl: str(v["metadata.tmdbBaseUrl"]) || null,
languages: parseLanguages(str(v["metadata.languages"])),
includeSpecials: v["metadata.includeSpecials"] === true,
...(key !== undefined ? { tmdbApiKey: key } : {}),
},
};
},
};
// TMDB 缓存(对照 MP:META_CACHE_EXPIRE / CONF.tmdb / EMPTY_RESULT_CACHE_TTL)
const TMDB_CACHE_CARD: CardSpec = {
anchor: "tmdb-cache",
title: "TMDB 缓存",
description:
"请求响应与识别结果存在数据库里,三个进程共用。超过条数上限时删最久未用的条目;有效期在「自动化 → 存储与保留」里设。",
fields: [{ name: "cache.tmdbMaxSize", label: "条数上限", kind: "number", hint: "请求缓存和识别缓存各自计数" }],
fromSettings: (s) => ({
"cache.tmdbMaxSize": n(s.cache.tmdbMaxSize),
}),
toPatch: (v) => ({
cache: {
tmdbMaxSize: Number(v["cache.tmdbMaxSize"]),
},
}),
extra: () => <TmdbCacheStatus />,
};
function TmdbCacheStatus() {
const q = useTmdbCache();
const clear = useClearTmdbCache();
const c = q.data?.counts;
return (
<div className="mb-2 flex flex-wrap items-center gap-2 text-meta text-muted-foreground">
<span className="tabular-nums">
{c ? `识别 ${c.recognize} 条 · 请求 ${c.request} 条 · 上限各 ${q.data?.maxSize} 条` : "读取中…"}
</span>
<Button
type="button"
variant="ghost"
size="xs"
className="hover:text-destructive"
disabled={clear.isPending}
onClick={async () => {
try {
const r = await clear.mutateAsync();
toast.success(`已清空 ${r.deleted} 条 TMDB 缓存`);
} catch (err) {
toast.error(errorMessage(err, "清空失败"));
}
}}
>
清空
</Button>
</div>
);
}
// 图片代理与缓存(对照 MP GLOBAL_IMAGE_CACHE / GLOBAL_IMAGE_CACHE_DAYS)
const IMAGES_CARD: CardSpec = {
anchor: "images",
title: "图片代理与缓存",
description: "海报经服务端从 TMDB 取回。开启缓存后保存到数据目录,下次直接读取。",
fields: [
{ name: "cache.globalImageCache", label: "缓存图片", kind: "switch" },
{ name: "cache.globalImageCacheDays", label: "保留天数", kind: "number", hint: "超过天数的图片每天清理一次" },
],
fromSettings: (s) => ({
"cache.globalImageCache": s.cache.globalImageCache,
"cache.globalImageCacheDays": n(s.cache.globalImageCacheDays),
}),
toPatch: (v) => ({
cache: {
globalImageCache: v["cache.globalImageCache"] === true,
globalImageCacheDays: Number(v["cache.globalImageCacheDays"]),
},
}),
extra: () => <ImageCacheStatus />,
};
function ImageCacheStatus() {
const q = useImageCache();
const clear = useClearImageCache();
const d = q.data;
return (
<div className="mb-2 flex flex-wrap items-center gap-2 text-meta text-muted-foreground">
<span className="tabular-nums">
{d ? `${d.files} 张 · ${(d.usedBytes / 1024 / 1024).toFixed(1)} MiB` : "读取中…"}
</span>
<Button
type="button"
variant="ghost"
size="xs"
className="hover:text-destructive"
disabled={clear.isPending}
onClick={async () => {
try {
const r = await clear.mutateAsync();
toast.success(`已删除 ${r.deleted} 张缓存图片`);
} catch (err) {
toast.error(errorMessage(err, "清空失败"));
}
}}
>
清空
</Button>
</div>
);
}
// 身份识别(DEC-1 惰性识别,对照 MP:RSS / 列表资源只在匹配订阅或用户搜索时识别)
const IDENTITY_CARD: CardSpec = {
anchor: "identity",
title: "资源识别",
description: "默认只在资源命中订阅、作品别名或新作品规则,或你手动识别和搜索时查 TMDB;其余放进未识别,不进待处理。",
fields: [
{
name: "identity.eagerTmdbLookup",
label: "每条新资源都查 TMDB",
kind: "switch",
hint: "每条新资源都查一次 TMDB。RSS 量大时每天可能上千次请求,站点独有的短剧和综艺多半查不到",
wide: true,
},
{
name: "identity.autoEnrich",
label: "TMDB 查不到时自动补全元数据",
kind: "switch",
hint: "资源命中订阅或新作品规则、TMDB 却查不到时,搜索网页出一份补全建议(每条资源一次,不建作品)。每次最多读 3 个网页,调用一次模型服务",
wide: true,
},
{
name: "identity.enrichDailyBudget",
label: "元数据补全每日上限",
kind: "number",
hint: "手动和自动合计;7 天内同一标题直接用上次结果,不计次。0 = 停用",
},
{
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。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,
},
],
fromSettings: (s) => ({
"identity.eagerTmdbLookup": s.identity.eagerTmdbLookup,
"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: {
eagerTmdbLookup: v["identity.eagerTmdbLookup"] === true,
autoEnrich: v["identity.autoEnrich"] === true,
enrichDailyBudget: Number(v["identity.enrichDailyBudget"]),
enrichDomains: String(v["identity.enrichDomains"] ?? "")
.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>
{(s) => (
<>
<SettingsCard section="metadata" spec={TMDB_CARD} settings={s} />
<SettingsCard section="metadata" spec={TMDB_CACHE_CARD} settings={s} />
<SettingsCard section="metadata" spec={IDENTITY_CARD} settings={s} />
<SettingsCard section="metadata" spec={IMAGES_CARD} settings={s} />
</>
)}
</SettingsGate>
);
}