rssany 0.1.5 → 0.2.0

package/README.md

@@ -31,11 +31,7 @@
 | --- | ------------------------------------------------------------ |
 | 运行时 | Node.js **20–23**（见 `package.json` `engines`）                |
 | 后端  | Hono、`tsx` 开发入口                                              |
-| 数据  | **SQLite**（`better-sqlite3`），默认 **`~/.rssany/data/rssany.db`**（Windows：`%USERPROFILE%\.rssany\data\rssany.db`） |
-| 前端  | `webui/`（SvelteKit + Vite，构建输出由根服务托管）                        |
-
-
-原生模块 `**better-sqlite3**` 安装时会编译；若遇绑定缺失，请确认未禁用构建（仓库 `pnpm-workspace.yaml` 中已允许其 `allowBuilds`）。
+| 数据  | **SQLite**（Node.js 内置 `node:sqlite`，Node.js 20+），默认 **`~/.rssany/data/rssany.db`**（Windows：`%USERPROFILE%\.rssany\data\rssany.db`） |
 
 ---
 

package/app/plugins/builtin/agi-eval-evaluation.rssany.js

@@ -131,7 +131,7 @@ function buildSummary(record) {
   return clampText(detail);
 }
 
-function toFeedItem(record, origin, source) {
+function toFeedItem(record, origin, _source) {
   if (!record || typeof record !== "object") return null;
   const title = normalizeText(record.name);
   if (!title) return null;

package/app/plugins/builtin/brightdata-blog.rssany.js

@@ -160,7 +160,7 @@ async function fetchFeedItems(feedUrl) {
       .map((node) => normalizeText(node.textContent))
       .filter(Boolean);
     const fallbackCategory = extractCategoryFromLink(link);
-    const finalCategories = categories.length > 0
+    const _finalCategories = categories.length > 0
       ? uniqueTexts(categories)
       : (fallbackCategory ? [fallbackCategory] : undefined);
 

package/app/plugins/builtin/five-radar.rssany.js

@@ -266,7 +266,7 @@ function mapRowToFeedItem(row, origin) {
 
   const summary = normalizeText(row?.summary);
   const author = normalizeText(row?.source || row?.author);
-  const category = normalizeText(row?.theme?.title);
+  const _category = normalizeText(row?.theme?.title);
   const idText = String(row?.id ?? "").trim();
 
   return {

package/app/plugins/builtin/google-deepmind-research.rssany.js

@@ -4,7 +4,7 @@ let _deps;
 
 
 
-const DEEPMIND_RESEARCH_URL = "https://deepmind.google/research/";
+const _DEEPMIND_RESEARCH_URL = "https://deepmind.google/research/";
 const DEEPMIND_ORIGIN = "https://deepmind.google";
 const MONTH_TO_INDEX = {
   january: 0,

package/app/plugins/builtin/opendrivelab-publications.rssany.js

@@ -104,7 +104,7 @@ async function fetchItems(sourceId, ctx) {
     if (seen.has(link)) continue;
     seen.add(link);
 
-    const { summary, category, year } = extractContext(anchor);
+    const { summary, category: _category, year } = extractContext(anchor);
     const pubDate = year != null ? new Date(Date.UTC(year, 0, 1)) : new Date();
 
     items.push({

package/app/plugins/builtin/pjlab-adg-publications.rssany.js

@@ -140,7 +140,7 @@ function parseOneEntry(liNode, currentYear, pageUrl) {
   const fallbackYear = parseYear(`${periodical} ${detailNode.textContent}`);
   const finalYear = currentYear ?? fallbackYear;
   const pubDate = finalYear != null ? new Date(Date.UTC(finalYear, 0, 1, 0, 0, 0)) : new Date();
-  const badge = normalizeText((liNode.querySelector(".abbr .badge") ?? liNode.querySelector("abbr"))?.textContent) || undefined;
+  const _badge = normalizeText((liNode.querySelector(".abbr .badge") ?? liNode.querySelector("abbr"))?.textContent) || undefined;
   const link = pickBestLink(detailNode, pageUrl, entryId);
   const guidSeed = entryId || link || `${title}|${author ?? ""}|${finalYear ?? ""}`;
 

package/app/plugins/builtin/theinformation-briefings.rssany.js

@@ -0,0 +1,150 @@
+let _deps;
+
+// The Information — AI Agenda 和 Briefings 列表页
+// 当前结构：.article.feed-item，标题 h3.title a，分类 .category-content a，作者 .authors，摘要 .recent-excerpt .long-excerpt
+
+const ORIGIN = "https://www.theinformation.com";
+const LIST_URL_RE =
+  /^https?:\/\/(www\.)?theinformation\.com\/(briefings|features\/[^/]+)\/?(\?.*)?$/i;
+
+
+function normalizeText(text) {
+  return (text ?? "").replace(/\s+/g, " ").trim();
+}
+
+
+function hashGuid(input) {
+  return _deps.createHash("sha256").update(input).digest("hex");
+}
+
+
+function toAbsoluteHttpUrl(rawHref, baseUrl) {
+  if (!rawHref) return null;
+  const href = rawHref.trim();
+  if (!href || href.startsWith("#") || href.startsWith("javascript:")) return null;
+  try {
+    const url = new URL(href, baseUrl);
+    if (!/^https?:$/i.test(url.protocol)) return null;
+    return url.href;
+  } catch {
+    return null;
+  }
+}
+
+
+function pad2(n) {
+  return String(n).padStart(2, "0");
+}
+
+
+/** .authors 文本：By Author · Apr 14, 2026 · 7:52am PDT */
+function parseAuthorsDate(raw) {
+  let t = normalizeText(raw);
+  t = t.replace(/\s*·\s*\d+\s+comments?\s*$/i, "").trim();
+
+  const m = t.match(
+    /^By\s+(.+?)\s*·\s*(.+?\d{4})\s*·\s*(\d{1,2}:\d{2}\s*(?:am|pm))\s*(PDT|PST|PT)\s*$/i
+  );
+  if (m) {
+    const author = m[1].trim();
+    const datePart = m[2].trim();
+    const timePart = m[3].trim();
+    const tz = m[4].toUpperCase();
+    const offset = tz === "PDT" ? "-07:00" : "-08:00";
+
+    const hm = timePart.match(/(\d{1,2}):(\d{2})\s*(am|pm)/i);
+    const d0 = new Date(datePart);
+    if (hm && !Number.isNaN(d0.getTime())) {
+      let h = Number(hm[1]);
+      const min = Number(hm[2]);
+      const ap = hm[3].toLowerCase();
+      if (ap === "pm" && h < 12) h += 12;
+      if (ap === "am" && h === 12) h = 0;
+      const y = d0.getFullYear();
+      const mo = d0.getMonth() + 1;
+      const da = d0.getDate();
+      const iso = `${y}-${pad2(mo)}-${pad2(da)}T${pad2(h)}:${pad2(min)}:00${offset}`;
+      const pubDate = new Date(iso);
+      if (!Number.isNaN(pubDate.getTime())) return { author, pubDate };
+    }
+  }
+
+  const authorMatch = t.match(/^By\s+(.+?)\s*·/i);
+  const author = authorMatch ? authorMatch[1].trim() : undefined;
+  const dateStr = t.replace(/^By\s+.*?\s*·\s*/, "").trim();
+  const pubDate = new Date(dateStr);
+  return { author, pubDate: Number.isNaN(pubDate.getTime()) ? new Date() : pubDate };
+}
+
+
+function parseFeedItems(html, pageUrl) {
+  const root = _deps.parseHtml(html);
+  const items = [];
+  const seen = new Set();
+
+  for (const node of root.querySelectorAll(".article.feed-item")) {
+    const linkEl = node.querySelector("h3.title a[href]");
+    if (!linkEl) continue;
+
+    const title = normalizeText(linkEl.textContent);
+    const link = toAbsoluteHttpUrl(linkEl.getAttribute("href"), pageUrl);
+    if (!title || !link || seen.has(link)) continue;
+    seen.add(link);
+
+    const authorsText = normalizeText(node.querySelector(".authors")?.textContent ?? "");
+    const { author, pubDate } = parseAuthorsDate(authorsText);
+
+    const summary = normalizeText(
+      node.querySelector(".recent-excerpt .long-excerpt")?.textContent ??
+      node.querySelector(".recent-excerpt")?.textContent ??
+      node.querySelector(".short-excerpt")?.textContent ??
+      ""
+    ) || undefined;
+
+    const categoryEl = node.querySelector(".category-content a");
+    const category = categoryEl ? normalizeText(categoryEl.textContent) : undefined;
+
+    items.push({
+      guid: hashGuid(link),
+      title,
+      link,
+      pubDate,
+      author,
+      summary,
+      categories: category ? [category] : undefined,
+    });
+  }
+
+  return items;
+}
+
+
+async function fetchItems(sourceId, ctx) {
+  _deps = ctx.deps;
+  const { html, finalUrl, status } = await ctx.fetchHtml(sourceId, {
+    waitMs: 5000,
+    waitForSelector: ".article.feed-item",
+    waitForSelectorTimeoutMs: 25_000,
+  });
+
+  const pageUrl = finalUrl || sourceId || ORIGIN;
+  const items = parseFeedItems(html, pageUrl);
+
+  if (items.length === 0) {
+    const hint = status && status >= 400 ? ` HTTP ${status}` : "";
+    throw new Error(
+      `[theinformation] 未解析到条目，页面结构可能已变化或需登录后抓取。${hint}`
+    );
+  }
+
+  items.sort((a, b) => b.pubDate.getTime() - a.pubDate.getTime());
+  return items;
+}
+
+
+export default {
+  id: "theinformation",
+  listUrlPattern: LIST_URL_RE,
+  refreshInterval: "1h",
+  fetchItems,
+};

package/app/plugins/builtin/x.rssany.js

@@ -34,6 +34,153 @@ function statusPathFromHref(href) {
 }
 
 
+/** 非推文正文链接：头像区、分析页、单张图/视频子路径等，避免误当作主帖 ID */
+function isAuxStatusSubpath(href) {
+  return /\/status\/\d+\/(photo|video|analytics|likes|retweets|quotes)\b/i.test(href || "");
+}
+
+
+/**
+ * 主帖路径：优先取时间戳旁 permalink（与 UI 一致），避免首条任意 /status/ 链到引用帖或图集子链。
+ */
+function extractPrimaryStatusPath(article) {
+  const timeEl = article.querySelector("time[datetime]");
+  if (timeEl) {
+    const a = timeEl.closest("a[href*='/status/']");
+    if (a) {
+      const href = a.getAttribute("href") || "";
+      if (!isAuxStatusSubpath(href)) {
+        const p = statusPathFromHref(href);
+        if (p) return p;
+      }
+    }
+  }
+  for (const a of article.querySelectorAll('a[href*="/status/"]')) {
+    const href = a.getAttribute("href") || "";
+    if (isAuxStatusSubpath(href)) continue;
+    const p = statusPathFromHref(href);
+    if (p) return p;
+  }
+  return null;
+}
+
+
+/** 引用/转推卡片内层 article 再解析会重复 guid，只处理时间轴最外层 tweet */
+function isNestedTweetArticle(article) {
+  let p = article.parentElement;
+  while (p) {
+    if (p.matches?.("article[data-testid='tweet']")) return true;
+    p = p.parentElement;
+  }
+  return false;
+}
+
+
+function extractSocialContext(article) {
+  const el = article.querySelector('[data-testid="socialContext"]');
+  return normalizeText(el?.textContent);
+}
+
+
+/** 时间轴上「转推/转发」帖：socialContext 含 Repost/Retweet 或中文 */
+function isRepostArticle(article) {
+  const ctx = extractSocialContext(article);
+  if (!ctx) return false;
+  if (/reposted?|retweet/i.test(ctx)) return true;
+  if (/转推|转发/.test(ctx)) return true;
+  return false;
+}
+
+
+function normalizeTwimgUrl(src) {
+  if (!src || typeof src !== "string") return src;
+  try {
+    const u = new URL(src, X_ORIGIN);
+    if (u.hostname.includes("twimg.com") && u.pathname.includes("/media/")) {
+      u.searchParams.set("format", "jpg");
+      u.searchParams.set("name", "small");
+    }
+    return u.href;
+  } catch {
+    return src;
+  }
+}
+
+
+/** 链接预览卡大图（card_img），与推文配图 /media/ 不同 */
+function normalizeCardImgUrl(src) {
+  if (!src || typeof src !== "string") return src;
+  try {
+    const u = new URL(src, X_ORIGIN);
+    if (u.hostname.includes("twimg.com") && /\/card_img\//.test(u.pathname)) {
+      if (!u.searchParams.has("format")) u.searchParams.set("format", "jpg");
+      if (!u.searchParams.has("name")) u.searchParams.set("name", "small");
+    }
+    return u.href;
+  } catch {
+    return src;
+  }
+}
+
+
+/**
+ * 无 tweetText 时：从链接卡 aria-label（"domain.com 标题…"）或 small 卡正文取一行摘要
+ */
+function extractLinkCardSummary(article) {
+  const a = article.querySelector('[data-testid="card.wrapper"] a[aria-label]');
+  if (a) {
+    const label = a.getAttribute("aria-label") || "";
+    const idx = label.indexOf(" ");
+    if (idx > 0) {
+      const rest = normalizeText(label.slice(idx + 1));
+      if (rest) return rest;
+    }
+  }
+  const detail = article.querySelector('[data-testid="card.layoutSmall.detail"]');
+  if (detail) {
+    for (const el of detail.querySelectorAll(':scope > div[dir="auto"]')) {
+      const t = normalizeText(el.textContent);
+      if (t && t.length > 8 && !/^[a-z0-9.-]+\.[a-z]{2,}$/i.test(t)) return t;
+    }
+  }
+  return "";
+}
+
+
+/**
+ * 首图：推文配图 > 视频 poster（blob 源无法持久化，仅 poster 可作缩略图）
+ */
+function extractMediaUrl(article) {
+  for (const img of article.querySelectorAll('[data-testid="tweetPhoto"] img[src]')) {
+    const src = img.getAttribute("src");
+    if (!src || /profile_images/i.test(src)) continue;
+    if (/pbs\.twimg\.com\/media/i.test(src) || /twimg\.com\/media/i.test(src)) {
+      return normalizeTwimgUrl(src);
+    }
+  }
+  const video = article.querySelector("video[poster]");
+  if (video) {
+    const poster = video.getAttribute("poster");
+    if (poster && /^https?:\/\//i.test(poster)) return poster;
+  }
+  for (const img of article.querySelectorAll(
+    '[data-testid="card.wrapper"] img[src*="twimg.com/card_img"], [data-testid="card.wrapper"] img[src*="pbs.twimg.com/card_img"]',
+  )) {
+    const src = img.getAttribute("src");
+    if (src && /^https?:\/\//i.test(src) && !/profile_images/i.test(src)) {
+      return normalizeCardImgUrl(src);
+    }
+  }
+  for (const img of article.querySelectorAll('img[src*="pbs.twimg.com/media"], img[src*="twimg.com/media"]')) {
+    const src = img.getAttribute("src");
+    if (src && !/profile_images/i.test(src) && !img.closest('[data-testid="User-Name"]')) {
+      return normalizeTwimgUrl(src);
+    }
+  }
+  return undefined;
+}
+
+
 function extractAuthor(article, statusPath) {
   const nameBlock = article.querySelector('[data-testid="User-Name"]');
   if (nameBlock) {
@@ -55,35 +202,49 @@ function extractAuthor(article, statusPath) {
 
 
 function extractTweetText(article) {
-  const textNode = article.querySelector('[data-testid="tweetText"]') ?? article.querySelector('[lang]');
-  const text = normalizeText(textNode?.textContent);
+  const nodes = article.querySelectorAll('[data-testid="tweetText"]');
+  const parts = [];
+  for (const node of nodes) {
+    const t = normalizeText(node.textContent);
+    if (t) parts.push(t);
+  }
+  let body = parts.length ? parts.join("\n\n") : "";
   const hasShowMore = !!article.querySelector('[data-testid="tweet-text-show-more-link"]');
-  if (!text) return hasShowMore ? "推文内容较长，请打开原文查看" : "";
-  return hasShowMore ? `${text} ...` : text;
+  if (!body) {
+    body = extractLinkCardSummary(article);
+  }
+  if (!body) {
+    body = hasShowMore ? "推文内容较长，请打开原文查看" : "";
+  } else if (hasShowMore) {
+    body = `${body} ...`;
+  }
+  if (!body) {
+    const fallback = article.querySelector("[lang]");
+    body = normalizeText(fallback?.textContent) || "";
+  }
+  return normalizeText(body);
 }
 
 
 function parseArticles(root, origin) {
   const entries = [];
   const seen = new Set();
-  const articles = root.querySelectorAll('article[data-testid="tweet"], article[role="article"]');
+  let articles = root.querySelectorAll('article[data-testid="tweet"]');
+  if (articles.length === 0) {
+    articles = root.querySelectorAll('article[role="article"]');
+  }
   for (const article of articles) {
-    const links = article.querySelectorAll('a[href*="/status/"]');
-    let statusPath = null;
-    for (const a of links) {
-      const p = statusPathFromHref(a.getAttribute("href"));
-      if (p) {
-        statusPath = p;
-        break;
-      }
-    }
+    if (isNestedTweetArticle(article)) continue;
+    const statusPath = extractPrimaryStatusPath(article);
     if (!statusPath || seen.has(statusPath)) continue;
     seen.add(statusPath);
     const link = new URL(statusPath, origin).href;
     const text = extractTweetText(article);
     const author = extractAuthor(article, statusPath);
     const pubDate = article.querySelector("time[datetime]")?.getAttribute("datetime") || undefined;
-    entries.push({ link, text, author, pubDate });
+    const imageUrl = extractMediaUrl(article);
+    const isRepost = isRepostArticle(article);
+    entries.push({ link, text, author, pubDate, imageUrl, isRepost });
   }
   return entries;
 }
@@ -108,14 +269,22 @@ function extractEntriesFromJson(data, origin) {
 
 
 function entriesToFeedItems(entries) {
-  return entries.map(({ link, text, author, pubDate }) => ({
-    guid: _deps.createHash("sha256").update(link).digest("hex"),
-    title: text || undefined,
-    link,
-    pubDate: pubDate ? new Date(pubDate) : new Date(),
-    author,
-    summary: text || undefined,
-  }));
+  return entries.map(({ link, text, author, pubDate, imageUrl, isRepost }) => {
+    const item = {
+      guid: _deps.createHash("sha256").update(link).digest("hex"),
+      /** 转发帖显示标题 Repost；其余不展示标题 */
+      title: isRepost ? "Repost" : "",
+      link,
+      pubDate: pubDate ? new Date(pubDate) : new Date(),
+      author,
+      summary: text || undefined,
+    };
+    if (imageUrl) {
+      item.imageUrl = imageUrl;
+      item.cover_img = imageUrl;
+    }
+    return item;
+  });
 }
 
 

package/app/plugins/builtin/zhipu-research.rssany.js

@@ -113,7 +113,7 @@ function buildItemsFromBlogsItems(blogsItems) {
     const summary = normalizeText(blog.resume_zh ?? blog.resume_en ?? "");
     const createdAt = String(blog.createAt ?? "").trim();
     const pubDate = createdAt ? new Date(createdAt) : new Date();
-    const category = normalizeText(blog.tag_zh ?? blog.tag_en ?? "");
+    const _category = normalizeText(blog.tag_zh ?? blog.tag_en ?? "");
     items.push({
       guid: hashGuid(link),
       title,
@@ -263,7 +263,7 @@ function buildItemsFromLeafSequence(html, titleIdMap) {
   for (let i = 0; i < leafTexts.length; i += 1) {
     const dateText = leafTexts[i];
     if (!isDateText(dateText)) continue;
-    const category = i > 0 && RESEARCH_TAGS.has(leafTexts[i - 1]) ? leafTexts[i - 1] : undefined;
+    const _category = i > 0 && RESEARCH_TAGS.has(leafTexts[i - 1]) ? leafTexts[i - 1] : undefined;
 
     let title = "";
     let summary;

package/app/plugins/site.rssany.js

@@ -7,6 +7,7 @@
 
 export default {
   id: "__PLUGIN_ID__",
+  // eslint-disable-next-line no-undef
   listUrlPattern: __LIST_URL_PATTERN__,
   refreshInterval: "1day",