rootandClaude Opus 5.5
ce93ea3a93
Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_01Cxwgnn2pbzQ7rXv3rq7fei
217 lines
10 KiB
TypeScript
217 lines
10 KiB
TypeScript
import { execFileSync } from "node:child_process";
|
|
import { fileURLToPath } from "node:url";
|
|
import { createMockTmdb, type MockTmdb, type TmdbFixture } from "@umt/testkit";
|
|
import { sql } from "kysely";
|
|
import { afterAll, beforeAll, describe, expect, it } from "vitest";
|
|
import { createEntityFromTmdbCandidate } from "../../src/modules/sources-identity/service/entity-from-candidate.ts";
|
|
import { refreshMetadata } from "../../src/modules/sources-identity/service/metadata.ts";
|
|
import { createHarness, type Harness, PG_URL, pgAvailable } from "./harness.ts";
|
|
|
|
// TMDB 多语言保留规则(用户规则 2026-10-06):建档 / 刷新只写简中 / 繁中 / 英文 / 原语言的别名与字段;
|
|
// scripts/tmdb-alias-language-cleanup.ts 只删 TMDB 来源、语言不在保留范围的别名。
|
|
|
|
const up = await pgAvailable();
|
|
const ROOT = fileURLToPath(new URL("../../../../", import.meta.url));
|
|
const SCRIPT = fileURLToPath(new URL("../../../../scripts/tmdb-alias-language-cleanup.ts", import.meta.url));
|
|
|
|
/** 日语原作的剧:各种语言的译名与别名 */
|
|
const AOT: TmdbFixture = {
|
|
type: "tv",
|
|
id: 1429,
|
|
original_title: "進撃の巨人",
|
|
original_language: "ja",
|
|
origin_country: ["JP"],
|
|
date: "2013-04-07",
|
|
titles: { "zh-CN": "进击的巨人", "en-US": "Attack on Titan", "ja-JP": "進撃の巨人", "zh-TW": "進擊的巨人" },
|
|
overviews: { "zh-CN": "巨人吃人的世界。", "ja-JP": "巨人が人を喰う世界。", "fr-FR": "Un monde de titans." },
|
|
popularity: 200,
|
|
alternative_titles: [
|
|
{ iso_3166_1: "JP", title: "Shingeki no Kyojin", type: "Romaji" },
|
|
{ iso_3166_1: "US", title: "AoT", type: "" },
|
|
{ iso_3166_1: "BR", title: "Ataque dos Titãs", type: "" },
|
|
{ iso_3166_1: "RU", title: "Атака на титанов", type: "" },
|
|
{ iso_3166_1: "CN", title: "巨人", type: "简称" },
|
|
{ iso_3166_1: "CN", title: "Attack Titan CN", type: "" },
|
|
{ iso_3166_1: "KR", title: "진격의 거인", type: "" },
|
|
],
|
|
extra_translations: [
|
|
{ iso_639_1: "ru", iso_3166_1: "RU", title: "Атака титанов" },
|
|
{ iso_639_1: "fr", iso_3166_1: "FR", title: "L'Attaque des Titans" },
|
|
{ iso_639_1: "ko", iso_3166_1: "KR", title: "진격의 거인" },
|
|
{ iso_639_1: "th", iso_3166_1: "TH", title: "ผ่าพิภพไททัน" },
|
|
],
|
|
external_ids: { imdb_id: "tt2560140", tvdb_id: 267440 },
|
|
poster_path: "/aot.jpg",
|
|
rich: { genres: ["动画"], status: "Ended" },
|
|
seasons: [
|
|
{
|
|
season_number: 1,
|
|
air_date: "2013-04-07",
|
|
names: { "zh-CN": "第 1 季", "en-US": "Season 1" },
|
|
episodes: [
|
|
{
|
|
episode_number: 1,
|
|
air_date: "2013-04-07",
|
|
names: { "fr-FR": "À toi, dans 2000 ans", "ja-JP": "二千年後の君へ" },
|
|
overviews: { "fr-FR": "Le mur tombe.", "ja-JP": "壁が壊される。" },
|
|
},
|
|
],
|
|
},
|
|
],
|
|
};
|
|
|
|
const ALLOWED = /^(zh(-(CN|SG|TW|HK|MO|Hans|Hant))?|en(-[A-Z]{2})?|ja(-JP|-Latn)?)$/;
|
|
|
|
describe.skipIf(!up)("TMDB 多语言保留规则", () => {
|
|
let h: Harness;
|
|
let tmdb: MockTmdb;
|
|
let id: string;
|
|
let url: string;
|
|
const actor = { type: "user" as const, id: "tester" };
|
|
|
|
const run = (...args: string[]) =>
|
|
execFileSync(Deno.execPath(), ["run", "-A", SCRIPT, ...args], {
|
|
cwd: ROOT,
|
|
env: { ...process.env, DATABASE_URL: url, NO_COLOR: "1" },
|
|
encoding: "utf8",
|
|
});
|
|
const aliases = () =>
|
|
h.db
|
|
.selectFrom("title_alias")
|
|
.select(["alias", "language", "kind", "source", "locked"])
|
|
.where("media_entity_id", "=", id)
|
|
.orderBy("alias")
|
|
.execute();
|
|
|
|
beforeAll(async () => {
|
|
tmdb = await createMockTmdb({ fixtures: [AOT] });
|
|
h = await createHarness("tmdblang", { tmdbUrl: tmdb.url, now: () => new Date("2026-10-06T00:00:00Z") });
|
|
const schema = (await sql<{ s: string }>`select current_schema() as s`.execute(h.db)).rows[0]!.s;
|
|
const u = new URL(PG_URL);
|
|
u.searchParams.set("options", `-c search_path=${schema},public`);
|
|
url = u.toString();
|
|
// 用户在语言链里放了 fr-FR(不是原语言,应被跳过)与 ja-JP(是原语言,保留)
|
|
await h.db
|
|
.insertInto("settings")
|
|
.values({ key: "metadata", value: JSON.stringify({ languages: ["fr-FR", "zh-CN", "ja-JP", "en-US"] }) } as never)
|
|
.execute();
|
|
});
|
|
afterAll(async () => {
|
|
await h?.close();
|
|
await tmdb?.close();
|
|
});
|
|
|
|
it("建档:别名只含保留语言,不写空语言的 other;罗马字记 ja-Latn", async () => {
|
|
const r = await createEntityFromTmdbCandidate(h.db, () => h.rt.tmdb(), { externalId: "1429", kind: "tv" }, actor);
|
|
id = r.mediaEntityId;
|
|
const list = await aliases();
|
|
const texts = list.map((a) => a.alias);
|
|
for (const bad of [
|
|
"Атака титанов",
|
|
"Атака на титанов",
|
|
"L'Attaque des Titans",
|
|
"진격의 거인",
|
|
"ผ่าพิภพไททัน",
|
|
"Ataque dos Titãs",
|
|
"Attack Titan CN",
|
|
])
|
|
expect(texts).not.toContain(bad);
|
|
expect(list.find((a) => a.alias === "進撃の巨人")).toMatchObject({ kind: "original", language: "ja" });
|
|
expect(list.find((a) => a.alias === "Shingeki no Kyojin")).toMatchObject({
|
|
kind: "romanized",
|
|
language: "ja-Latn",
|
|
});
|
|
expect(list.find((a) => a.alias === "巨人")).toMatchObject({ kind: "other", language: "zh-CN" });
|
|
expect(list.find((a) => a.alias === "AoT")).toMatchObject({ language: "en-US" });
|
|
expect(list.filter((a) => a.kind === "other" && a.language === null)).toEqual([]);
|
|
for (const a of list) if (a.language !== null) expect(a.language).toMatch(ALLOWED);
|
|
});
|
|
|
|
it("刷新:语言链去掉 fr-FR,标题 / 简介 / 集名不取法语;别名仍只含保留语言", async () => {
|
|
const res = await refreshMetadata(h.rt, id, { full: true });
|
|
expect(res.status).toBe("refreshed");
|
|
expect(res.titles).toMatchObject({ title: "进击的巨人", language: "zh-CN" });
|
|
const fields = await h.db
|
|
.selectFrom("metadata_field")
|
|
.leftJoin("content_unit", "content_unit.id", "metadata_field.content_unit_id")
|
|
.select(["metadata_field.field", "metadata_field.language", "metadata_field.value"])
|
|
.where((eb) =>
|
|
eb.or([eb("metadata_field.media_entity_id", "=", id), eb("content_unit.media_entity_id", "=", id)]),
|
|
)
|
|
.where("metadata_field.language", "is not", null)
|
|
.execute();
|
|
expect(fields.length).toBeGreaterThan(0);
|
|
for (const f of fields) expect(f.language).toMatch(ALLOWED);
|
|
const ep = await h.db
|
|
.selectFrom("content_unit")
|
|
.select(["title", "metadata"])
|
|
.where("media_entity_id", "=", id)
|
|
.where("kind", "=", "episode")
|
|
.executeTakeFirstOrThrow();
|
|
// 集名:fr-FR 被跳过,ja-JP(原语言)可用
|
|
expect(ep.title).toBe("二千年後の君へ");
|
|
expect((ep.metadata as Record<string, { language: string | null }>).name?.language).toBe("ja-JP");
|
|
const list = await aliases();
|
|
expect(list.filter((a) => a.kind === "other" && a.language === null)).toEqual([]);
|
|
for (const a of list) if (a.language !== null) expect(a.language).toMatch(ALLOWED);
|
|
expect(list.map((a) => a.alias)).not.toContain("L'Attaque des Titans");
|
|
});
|
|
|
|
it("清理脚本:dry-run 统计、apply 只删 TMDB 来源的不合规别名,人工 / 站点 / 锁定的保留", async () => {
|
|
// 旧版本写下的数据:各语言译名(translated)、空语言 other;以及人工 / 站点 / 锁定的同类文本
|
|
const add = (alias: string, language: string | null, kind: string, source = "tmdb", locked = false) =>
|
|
sql`insert into title_alias (media_entity_id, content_unit_id, alias, normalized, language, kind, source, locked)
|
|
values (${id}, null, ${alias}, ${alias.toLowerCase()}, ${language}, ${kind}, ${source}, ${locked})`.execute(
|
|
h.db,
|
|
);
|
|
await add("Атака титанов", "ru-RU", "translated");
|
|
await add("Ataque dos Titãs", "pt-BR", "translated");
|
|
await add("L'Attaque des Titans", "fr-FR", "translated");
|
|
await add("Атака на титанов", null, "other");
|
|
await add("진격의 거인", null, "other");
|
|
await add("Ataque Titan", null, "other"); // 拉丁:按英文保留
|
|
await add("ผ่าพิภพไททัน", "th-TH", "translated", "user"); // 人工:不动
|
|
await add("Атака гигантов", "ru-RU", "site_title", "site"); // 站点:不动
|
|
await add("Attaque verrouillée", "fr-FR", "translated", "tmdb", true); // 锁定:不动
|
|
// 字段来源里的法语值(旧语言链写下的):只报告
|
|
await sql`insert into metadata_field (media_entity_id, content_unit_id, field, value, language, source, source_revision, fetched_at)
|
|
values (${id}, null, 'tagline', '"Le mur"'::jsonb, 'fr-FR', 'tmdb', 'x', now())`.execute(h.db);
|
|
|
|
const before = (await aliases()).length;
|
|
const out = run("--dry-run");
|
|
expect(out).toContain("DRY-RUN");
|
|
expect(out).toMatch(/删除 5 条/);
|
|
expect(out).toContain("ru-RU / translated:1 条");
|
|
expect(out).toContain("(空) / other:2 条");
|
|
expect(out).toContain("不动(非 tmdb 来源或已锁定):");
|
|
expect(out).toContain("## 字段来源(tmdb)语言不在保留范围:1 条");
|
|
expect(out).toContain("fr-FR / 作品·tagline:1 条");
|
|
expect((await aliases()).length).toBe(before);
|
|
|
|
const applied = run("--apply");
|
|
expect(applied).toContain("已删除 5 条别名");
|
|
const after = await aliases();
|
|
const texts = after.map((a) => a.alias);
|
|
for (const gone of ["Атака титанов", "Ataque dos Titãs", "L'Attaque des Titans", "Атака на титанов", "진격의 거인"])
|
|
expect(texts).not.toContain(gone);
|
|
for (const kept of [
|
|
"Ataque Titan",
|
|
"ผ่าพิภพไททัน",
|
|
"Атака гигантов",
|
|
"Attaque verrouillée",
|
|
"進撃の巨人",
|
|
"Shingeki no Kyojin",
|
|
])
|
|
expect(texts).toContain(kept);
|
|
expect(after.length).toBe(before - 5);
|
|
const audit = await h.db
|
|
.selectFrom("audit_event")
|
|
.select(["payload"])
|
|
.where("action", "=", "metadata.tmdb_alias_language_cleanup")
|
|
.executeTakeFirstOrThrow();
|
|
expect((audit.payload as { deleted: number }).deleted).toBe(5);
|
|
// 再跑一遍没有要删的
|
|
expect(run()).toContain("删除 0 条");
|
|
});
|
|
});
|