fix(sites): 憨憨详情页按真实结构解析豆瓣块与 MediaInfo

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

packages/adapters/src/sites/definitions/hhanclub.json

@@ -264,7 +264,9 @@
       },
       "page": {
         "labelSelectors": ["div.font-bold.leading-6", "span.font-bold"],
-        "fileListPath": "viewfilelist.php?id={id}"
+        "fileListPath": "viewfilelist.php?id={id}",
+        "mediaInfo": "#mediainfo-raw",
+        "descriptionStrip": "div.mt-5, div.flex.flex-row.items-center"
       }
     },
     "vipLevels": ["贵宾", "VIP"],

packages/adapters/src/sites/spider/detail-page.test.ts

@@ -1,4 +1,9 @@
-import { HHANCLUB_DETAIL_PAGE_2026, HHANCLUB_DETAIL_PAGE_FULL, HHANCLUB_FILELIST_FRAGMENT } from "@umt/testkit";
+import {
+  HHANCLUB_DETAIL_PAGE_2026,
+  HHANCLUB_DETAIL_PAGE_FULL,
+  HHANCLUB_DETAIL_PAGE_REAL_2026,
+  HHANCLUB_FILELIST_FRAGMENT,
+} from "@umt/testkit";
 import { describe, expect, it } from "vitest";
 import { hhanclubDefinition } from "../definitions/index.ts";
 import { looksLikeMediaInfo, pageFromBbcode } from "../detail-text.ts";
@@ -109,6 +114,30 @@ Format : AVC[/quote]
   });
 });
 
+describe("憨憨真实详情页(2026-10-02 摘录)", () => {
+  const { detail } = parseDetailPage(HHANCLUB_DETAIL_PAGE_REAL_2026, hhanclubDefinition, "219860", OBSERVED);
+  const page = detail!.page!;
+
+  it("MediaInfo 只取 #mediainfo-raw 原文,不带摘要表", () => {
+    expect(page.mediaInfo!.startsWith("General\nComplete name")).toBe(true);
+    expect(page.mediaInfo).not.toContain("体　积");
+  });
+
+  it("豆瓣块一项一行;去掉分隔线、空值项、标题行与更新提示;链接之间已有逗号不再补分隔", () => {
+    const t = page.descriptionText!;
+    expect(t.split("\n").slice(0, 3)).toEqual([
+      "片名: 斗罗大陆Ⅱ绝世唐门 年番1",
+      "别名: 斗罗大陆2,斗罗大陆2：绝世唐门\u200e,斗罗大陆2绝世唐门,绝世唐门动画版",
+      "年代: 2023",
+    ]);
+    expect(t).toContain("\n导演: 沈乐平 Leping Shen\n");
+    expect(t).toContain("编剧: 沈乐平 Leping Shen,唐家三少 Wei Zhang");
+    expect(t).toContain("主演:\n梁达伟 Dawei Liang / 蒋丽 Li Jiang");
+    expect(t).not.toMatch(/-{5,}|标语|该信息最后更新|\(23691\)/);
+    expect(page.doubanId).toBe("35567787");
+  });
+});
+
 describe("详情页扩展内容(憨憨 2026 真实结构:分节 div,无 #kdescr)", () => {
   const { kind, detail } = parseDetailPage(HHANCLUB_DETAIL_PAGE_2026, hhanclubDefinition, "91003", OBSERVED);
   const page = detail!.page!;

packages/adapters/src/sites/spider/detail-page.ts

@@ -220,16 +220,38 @@ export function extractDetailPage($: CheerioAPI, conf: DetailConf, base: string)
   } else if (infoBlock) {
     const c = infoBlock.clone();
     c.find("img").remove();
+    if (p.descriptionStrip) c.find(p.descriptionStrip).remove();
     // 块里的「标签」span 与值之间补上冒号 / 换行,读起来是「片名:xxx」一行一项
+    // (换行要用 BR 占位:源码里的换行在 blockText 里按空白处理)
+    // 站点排版用的分隔线(憨憨是行内 font):换成换行
+    c.find("font, span, div")
+      .filter((_, x) => /^[-=_—]{5,}$/.test($(x).text().trim()))
+      .replaceWith(BR);
     c.find(p.labelSelectors?.join(", ") || "span.font-bold").each((_, n) => {
       const e = $(n);
       const t = e.text().trim();
+      // 空值的项(憨憨「标语:」)整项去掉
+      const v = e.next();
+      if (v.length && !v.text().trim() && !v.find("img").length) {
+        (e.parent().children().length === 2 ? e.parent() : e).remove();
+        v.remove();
+        return;
+      }
       if (t && t.length <= 16 && !/[:：]$/.test(t)) e.text(`${t}：`);
-      e.before("\n");
+      e.before(BR);
+    });
+    // 并排的演职员链接之间补分隔(之间已有逗号等文字的不补)
+    c.find("a + a").each((_, n) => {
+      const prev = n.prev;
+      if (prev && prev.type === "text" && prev.data.trim()) return;
+      $(n).before(" / ");
     });
     const t = blockText($, c)
+      .replace(/[ \t]{2,}/g, " ")
       .replace(/[ \t]*\n[ \t]*/g, "\n")
-      .replace(/：\s*\n(?=\S)/g, "：");
+      .replace(/\n{3,}/g, "\n\n")
+      .replace(/：\s*\n(?=\S)/g, "：")
+      .trim();
     if (t) page.descriptionText = t.slice(0, 20_000);
   }