feat(parser,sites): W5-D port title/zhconv/site parsing gaps found by MP differential

Title recognition (parseRelease / parseReleasePath):
- anime prefix tags (★N月新番, [国漫], 【22年日剧】), category lead bracket, `A / B` cn aliases
- `Name [01]`, `[TV 04]`, `#13`, `名称 12.mkv`, `S01 13-14`, `E13-14` episodes
- lead bracket as title vs group, trailing [Group], YTS.MX/Hall_of_C style groups, 字幕组 suffix
- `[dot.release.name][…]` unbracketing, `Season 2017` as year, leading date not airDate,
  date-only parent dir does not merge as TV
- original disc: Blu-ray + AVC/HEVC/VC-1/MPEG-2 without x26x/H.26x -> BluRay-Disc
- `{[tmdbid=;type=;s=;e=]}` explicit block, bangumi/anilist ids
- bare [1080]/[5.1], DTS-HD MA5.1, ￡ multi-group, 合集, pinyin title -> cn from subtitle,
  streaming platform table, CHS/CHT/GB/BIG5 in subtitle
zh-map: add 迴滾魷瑯臥駭; document one-to-many chars kept out.
Sites: tbody branch + header-row tightening via movieclaw-import rowSelector (18 definitions),
row-level case/hr/labels include the row itself, subtitle branch (searchType: "subtitles"),
parseSizeMp banker's rounding.
Differential snapshots regenerated; MP test sample sources registered in ledger.

Co-Authored-By: Claude Opus 5.5 <[email protected]>

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

@@ -10,16 +10,27 @@
  *   → `SiteSpider.getInfo` 与同名私有方法
  * - `_safe_query / __filter_text / __remove / __attribute_or_text / __index` → 同名方法
  * - `__is_login_or_permission_page` → `isLoginOrPermissionPage`
- * - `parse` → `SiteSpider.parse`(不含 Rust 加速分支与字幕分支)
- * - `app/foundation/size.py parse_size` → `parseSizeMp`
+ * - `parse` → `SiteSpider.parse`(不含 Rust 加速分支);`search_type=subtitles` 分支:`get_subtitle_info /
+ *   __get_language_icon / __get_report_url / __get_subtitle_field / __fill_subtitle_ids` → `getSubtitleInfo` 等同名方法
+ * - `app/foundation/size.py parse_size` → `parseSizeMp`(Python round 为银行家舍入,同样实现)
  *
  * 与 MP 的差异(均在 README 与迁移账本登记):
  * 1. 额外输出 `category_id`(原始分类 ID),`category` 仍是 MP 的媒体类型投影(REQ §4.5 原始分类不得丢失)。
  * 2. 系数 `selector` 模式用浮点而非 int()(MP 对 0.5 抛异常导致整条丢失,迁移账本 SA4)。
  * 3. 不发请求:HTTP/Cookie/UA 由 adapter 注入,本类只负责 URL 构造与 HTML 解析(便于限流与脱敏统一)。
+ * 4. 查询范围:PyQuery 在行元素上查询是 descendant-or-self(行自身也会命中),MP Rust 快路径与 cheerio `.find` 只查后代。
+ *    本项目取折中:促销 case / hr / labels 这三类"行级标记"(NexusPHP 把 free_bg、sticky 等 class 挂在 <tr> 上)
+ *    包含行自身,与 Python 一致;取值字段(safeQuery)只查后代,避免 `div:nth-child(3)` 这类选择器命中行本身取到整行文本。
  */
 import { type CheerioAPI, load } from "cheerio";
-import type { IndexerDefinition, SpiderFieldSelector, SpiderFields, SpiderFilter, SpiderListConfig } from "./schema.ts";
+import type {
+  IndexerDefinition,
+  SpiderCategoryConfig,
+  SpiderFieldSelector,
+  SpiderFields,
+  SpiderFilter,
+  SpiderListConfig,
+} from "./schema.ts";
 import { renderTemplate } from "./template.ts";
 
 type Sel = ReturnType<CheerioAPI>;
@@ -65,8 +76,18 @@ export interface SpiderTorrent {
   hit_and_run?: boolean;
   category?: MpMediaType;
   category_id?: string;
+  // ---- 字幕分支(search_type=subtitles)专有字段,键名与 MP 一致 ----
+  language_icon?: string;
+  report_url?: string;
+  language?: string;
+  uploader?: string;
+  torrent_id?: string;
+  subtitle_id?: string;
+  file_name?: string;
 }
 
+export type SpiderSearchType = "torrents" | "subtitles";
+
 /** MP `select_media_categories` */
 export function selectMediaCategories(category: IndexerDefinition["category"], mtype?: MediaKind) {
   if (!category) return [];
@@ -86,7 +107,15 @@ export function resolveCategoryMediaType(value: unknown, category: IndexerDefini
   return matches.length === 1 ? MEDIA_VALUE[matches[0]!] : "未知";
 }
 
-/** MP `size_tools.parse_size`:二进制单位,round。 */
+/** Python 3 `round()`:恰好 .5 时取偶数(银行家舍入) */
+function roundHalfEven(v: number): number {
+  const f = Math.floor(v);
+  const diff = v - f;
+  if (diff === 0.5) return f % 2 === 0 ? f : f + 1;
+  return Math.round(v);
+}
+
+/** MP `size_tools.parse_size`:二进制单位,round(银行家舍入)。 */
 export function parseSizeMp(text: string | number | null | undefined): number {
   if (!text) return 0;
   const s = String(text);
@@ -101,7 +130,7 @@ export function parseSizeMp(text: string | number | null | undefined): number {
   else if (normalized.includes("GB") || normalized.includes("GIB")) v *= 1024 ** 3;
   else if (normalized.includes("MB") || normalized.includes("MIB")) v *= 1024 ** 2;
   else if (normalized.includes("KB") || normalized.includes("KIB")) v *= 1024;
-  return Math.round(v);
+  return roundHalfEven(v);
 }
 
 /** strptime 结果(MP dateparse 过滤器返回 datetime) */
@@ -193,6 +222,8 @@ export interface SiteSpiderQuery {
   cat?: string;
   /** 站点内部页码(MP 0 起) */
   page?: number;
+  /** MP `search_type`:subtitles 时改用 `def.subtitles` 的配置 */
+  searchType?: SpiderSearchType;
 }
 
 export class SiteSpider {
@@ -204,8 +235,13 @@ export class SiteSpider {
   readonly domain: string;
   readonly resultNum: number;
   readonly siteMediaType: MediaKind | undefined;
+  readonly searchType: SpiderSearchType;
   private list: SpiderListConfig;
   private fields: SpiderFields;
+  private searchCfg: IndexerDefinition["search"];
+  private batchCfg: IndexerDefinition["batch"];
+  private browseCfg: IndexerDefinition["browse"];
+  private categoryCfg: SpiderCategoryConfig | undefined;
   private requestedResultMediaType: MediaKind | undefined;
   private $!: CheerioAPI;
   private info: SpiderTorrent = {};
@@ -217,26 +253,42 @@ export class SiteSpider {
     this.cat = q.cat;
     this.page = q.page ?? 0;
     this.siteMediaType = fromAgent(def.media_type);
-    this.list = def.torrents.list;
-    this.fields = def.torrents.fields ?? {};
-    if (!q.keyword && def.browse) {
-      this.list = def.browse.list ?? this.list;
-      this.fields = def.browse.fields ?? this.fields;
+    this.searchType = q.searchType ?? "torrents";
+    if (this.searchType === "subtitles") {
+      const sc = def.subtitles ?? {};
+      this.searchCfg = sc.search;
+      this.batchCfg = sc.batch;
+      this.browseCfg = sc.browse;
+      this.categoryCfg = sc.category;
+      this.list = sc.list ?? { selector: "" };
+      this.fields = sc.fields ?? {};
+      this.resultNum = Number(sc.result_num ?? def.result_num ?? 100);
+    } else {
+      this.searchCfg = def.search;
+      this.batchCfg = def.batch;
+      this.browseCfg = def.browse;
+      this.categoryCfg = def.category;
+      this.list = def.torrents.list;
+      this.fields = def.torrents.fields ?? {};
+      if (!q.keyword && def.browse) {
+        this.list = def.browse.list ?? this.list;
+        this.fields = def.browse.fields ?? this.fields;
+      }
+      this.resultNum = Number(def.result_num ?? 100);
     }
     this.domain = def.domain.endsWith("/") ? def.domain : `${def.domain}/`;
-    this.resultNum = Number(def.result_num ?? 100);
   }
 
   /** MP `__can_search` */
   canSearch(): boolean {
     if (this.siteMediaType && this.mtype && this.siteMediaType !== this.mtype) return false;
-    return Boolean(this.def.search && this.domain);
+    return Boolean(this.searchCfg && this.domain);
   }
 
   /** MP `__format_search_word` */
   private formatSearchWord(word: string): string {
-    if (word && /^tt\d+$/.test(word) && this.def.search?.imdbid_format) {
-      return this.def.search.imdbid_format
+    if (word && /^tt\d+$/.test(word) && this.searchCfg?.imdbid_format) {
+      return this.searchCfg.imdbid_format
         .replace(/\{keyword\}/g, word)
         .replace(/\{imdbid\}/g, word)
         .replace(/\{imdbid_num\}/g, word.slice(2));
@@ -246,7 +298,7 @@ export class SiteSpider {
 
   /** MP `__get_search_url` */
   getSearchUrl(): string {
-    const search = this.def.search ?? { paths: [] };
+    const search = this.searchCfg ?? { paths: [] };
     const paths = search.paths ?? [];
     let torrentspath = "";
     let typedPathSelected = false;
@@ -272,9 +324,9 @@ export class SiteSpider {
       let searchWord: string;
       let searchMode: string;
       if (Array.isArray(this.keyword)) {
-        if (this.def.batch) {
-          const delimiter = this.def.batch.delimiter || " ";
-          const spaceReplace = this.def.batch.space_replace || " ";
+        if (this.batchCfg) {
+          const delimiter = this.batchCfg.delimiter || " ";
+          const spaceReplace = this.batchCfg.space_replace || " ";
           searchWord = this.keyword.map((k) => String(k).replace(/ /g, spaceReplace)).join(delimiter);
         } else {
           searchWord = this.keyword.join(" ");
@@ -296,7 +348,7 @@ export class SiteSpider {
           notnewword: 1,
         };
         for (const [k, v] of Object.entries(indexerParams)) params[k] = String(v).replace(/\{keyword\}/g, searchWord);
-        const category = this.def.category;
+        const category = this.categoryCfg;
         if (category) {
           const cats = selectMediaCategories(category, this.mtype);
           const allowed = this.cat ? new Set(this.cat.split(",")) : null;
@@ -333,9 +385,9 @@ export class SiteSpider {
       }
     } else {
       let page = this.page || 0;
-      if (this.def.browse && !typedPathSelected) {
-        torrentspath = this.def.browse.path;
-        if (this.def.browse.start) page = Number(this.def.browse.start) + Number(this.page || 0);
+      if (this.browseCfg && !typedPathSelected) {
+        torrentspath = this.browseCfg.path;
+        if (this.browseCfg.start) page = Number(this.browseCfg.start) + Number(this.page || 0);
       } else if (this.page && !torrentspath.includes("{page}")) {
         torrentspath += `${torrentspath.includes("?") ? "&" : "?"}page=${this.page}`;
       }
@@ -369,7 +421,7 @@ export class SiteSpider {
     const rows = $(normalizeSelector(this.list.selector ?? ""));
     rows.each((i, el) => {
       if (i >= this.resultNum) return false;
-      const info = this.getInfo($(el));
+      const info = this.searchType === "subtitles" ? this.getSubtitleInfo($(el)) : this.getInfo($(el));
       if (Object.keys(info).length > 0) out.push(this.applyRequestedMediaType(info));
       return undefined;
     });
@@ -414,6 +466,70 @@ export class SiteSpider {
     }
   }
 
+  /** MP `get_subtitle_info`:缺标题或下载链接的行丢弃;任一字段异常整条返回 {}。 */
+  getSubtitleInfo(row: Sel): SpiderTorrent {
+    this.info = {};
+    try {
+      this.getTitle(row);
+      this.getDescription(row);
+      this.getDetail(row);
+      this.getDownload(row);
+      this.getSize(row);
+      this.getPubdate(row);
+      this.getDateElapsed(row);
+      this.getGrabs(row);
+      this.getLinkField(row, "language_icon", "language_icon");
+      this.getLinkField(row, "report", "report_url");
+      for (const name of ["language", "uploader", "torrent_id", "subtitle_id", "file_name"] as const) {
+        this.getSubtitleField(row, name);
+      }
+      this.fillSubtitleIds();
+      if (!this.info.title || !this.info.enclosure) return {};
+      return { ...this.info };
+    } catch {
+      return {};
+    }
+  }
+
+  /** MP `__get_language_icon` / `__get_report_url`:相对链接补全为绝对链接 */
+  private getLinkField(row: Sel, name: string, key: "language_icon" | "report_url") {
+    if (!(name in this.fields)) return;
+    const selector = this.fields[name] ?? {};
+    const link = this.filterText(this.safeQuery(row, selector), selector.filters);
+    if (link) this.info[key] = this.normalizeLink(String(link));
+  }
+
+  /** MP `__get_subtitle_field` */
+  private getSubtitleField(row: Sel, name: "language" | "uploader" | "torrent_id" | "subtitle_id" | "file_name") {
+    const selector = this.fields[name];
+    if (!selector || Object.keys(selector).length === 0) return;
+    const v = this.filterText(this.safeQuery(row, selector), selector.filters);
+    if (v != null) this.info[name] = v instanceof ParsedDateTime ? v.text : String(v);
+  }
+
+  /** MP `__fill_subtitle_ids`:从下载链接 query 补种子 ID(torrentid/torrent_id)与字幕 ID(subid/subtitle) */
+  private fillSubtitleIds() {
+    const enclosure = this.info.enclosure;
+    if (!enclosure) return;
+    const q = enclosure.includes("?") ? enclosure.slice(enclosure.indexOf("?") + 1).split("#")[0]! : "";
+    const params = new URLSearchParams(q);
+    const first = (...keys: string[]) => {
+      for (const k of keys) {
+        const v = params.getAll(k).find((x) => x !== "");
+        if (v !== undefined) return v;
+      }
+      return undefined;
+    };
+    if (!this.info.torrent_id) {
+      const v = first("torrentid", "torrent_id");
+      if (v !== undefined) this.info.torrent_id = v;
+    }
+    if (!this.info.subtitle_id) {
+      const v = first("subid", "subtitle");
+      if (v !== undefined) this.info.subtitle_id = v;
+    }
+  }
+
   // ---- 字段 ----
 
   private getTitle(row: Sel) {
@@ -532,7 +648,7 @@ export class SiteSpider {
     this.info[key] = 1;
     if (selector.case) {
       for (const [css, value] of Object.entries(selector.case)) {
-        if (row.find(normalizeSelector(css)).length > 0) {
+        if (this.rowScope(row, css).length > 0) {
           this.info[key] = value;
           break;
         }
@@ -606,7 +722,7 @@ export class SiteSpider {
       this.info.labels = [];
       return;
     }
-    const q = row.find(normalizeSelector(selector.selector)).clone();
+    const q = this.rowScope(row, selector.selector).clone();
     this.remove(q, selector);
     this.info.labels = this.attributeOrText(q, selector).filter((x): x is string => Boolean(x));
   }
@@ -614,7 +730,7 @@ export class SiteSpider {
   private getHitAndRun(row: Sel) {
     if (!("hr" in this.fields)) return;
     const selector = this.fields.hr ?? {};
-    this.info.hit_and_run = selector.selector ? row.find(normalizeSelector(selector.selector)).length > 0 : false;
+    this.info.hit_and_run = selector.selector ? this.rowScope(row, selector.selector).length > 0 : false;
   }
 
   private getCategory(row: Sel) {
@@ -629,13 +745,21 @@ export class SiteSpider {
     const selector = this.fields.category ?? {};
     const value = this.filterText(this.safeQuery(row, selector), selector.filters);
     if (value != null && value !== "") this.info.category_id = String(value);
-    let resolved = resolveCategoryMediaType(value, this.def.category);
+    let resolved = resolveCategoryMediaType(value, this.categoryCfg);
     if (resolved === "未知" && this.siteMediaType) resolved = MEDIA_VALUE[this.siteMediaType];
     this.info.category = resolved;
   }
 
   // ---- 通用查询 ----
 
+  /** 行级标记查询:行自身 + 后代(PyQuery descendant-or-self,见文件头差异 4);`:scope` 选择器只查后代 */
+  private rowScope(row: Sel, selector: string): Sel {
+    const css = normalizeSelector(selector);
+    const desc = row.find(css);
+    if (css.includes(":scope")) return desc;
+    return row.filter(css).add(desc);
+  }
+
   /** MP `_safe_query` */
   safeQuery(row: Sel, config: SpiderFieldSelector | undefined): string | null | undefined {
     if (!config?.selector) return null;