ns-rss-spider 0.0.19 → 0.0.21

package/dist/cjs/parse.d.ts

@@ -1,5 +1,6 @@
 import { ServerInfo } from "./upload";
-export declare function parseRss(name: string, feed: string, server?: ServerInfo): Promise<import("./types").Article[] | {
+import { Article } from "./types";
+export declare function parseRss(name: string, feed: string, server?: ServerInfo): Promise<Article[] | {
     storePath: any;
     type: "simple" | "rich";
     guid: string;
@@ -10,4 +11,5 @@ export declare function parseRss(name: string, feed: string, server?: ServerInfo
     author?: string | undefined;
     source?: string | undefined;
     categories?: string | undefined;
+    thumb?: string | undefined;
 }[]>;

package/dist/cjs/parse.js

@@ -65,6 +65,7 @@ async function parseRss(name, feed, server) {
       item.guid = item.link;
     }
   });
+  const strategy = import_strategy.strategies[name];
   const contents = [];
   for (let item of result.items) {
     console.log(import_zx.chalk.green(`正在解析文章 【${item.title}】`));
@@ -72,11 +73,23 @@ async function parseRss(name, feed, server) {
       console.log(JSON.stringify(item));
       throw Error(`item has no guid`);
     }
-    const content = await (0, import_parseContent.parseContent)(item, import_strategy.strategies[name]).catch((e) => console.error(import_zx.chalk.red("文章解析失败"), e.message));
+    const content = await (0, import_parseContent.parseContent)(item, strategy).catch((e) => console.error(import_zx.chalk.red("文章解析失败"), e));
     if (content) {
       contents.push(content);
     }
   }
+  if (strategy == null ? void 0 : strategy.getThumbs) {
+    console.log(import_zx.chalk.green(`正在获取缩略图数据`));
+    const thumbKv = await (strategy == null ? void 0 : strategy.getThumbs(contents));
+    if (thumbKv && Object.keys(thumbKv).length) {
+      console.log(import_zx.chalk.green(`正在更新缩略图`));
+      contents.forEach((content) => {
+        if (thumbKv[content.guid]) {
+          content.thumb = thumbKv[content.guid];
+        }
+      });
+    }
+  }
   if (server) {
     const r = await (0, import_upload.uploadContent)({
       app: name,

package/dist/cjs/parseContent/parseContent.js

@@ -55,6 +55,13 @@ async function parseContent($, $element, item, strategy) {
     }
   });
   for (let item2 of srcs) {
+    if (strategy == null ? void 0 : strategy.ignoreProbeImage) {
+      images.push({
+        url: item2.src,
+        title: item2.title
+      });
+      continue;
+    }
     console.log(import_zx.chalk.green("正在解析图片", item2.src));
     const result = await (0, import_probe_image_size.default)(item2.src, {
       rejectUnauthorized: false,
@@ -99,7 +106,7 @@ async function parseContent($, $element, item, strategy) {
   return {
     type: "rich",
     ...(0, import_getBasicFromItem.getBasicFromItem)(item),
-    content: (0, import_html_entities.decode)($element.html().trim()),
+    content,
     ssrContent: (0, import_generateSsrContent.generateSsrContent)(content, images),
     images
   };

package/dist/cjs/parsers/cnbeta.js

@@ -1,6 +1,8 @@
+var __create = Object.create;
 var __defProp = Object.defineProperty;
 var __getOwnPropDesc = Object.getOwnPropertyDescriptor;
 var __getOwnPropNames = Object.getOwnPropertyNames;
+var __getProtoOf = Object.getPrototypeOf;
 var __hasOwnProp = Object.prototype.hasOwnProperty;
 var __export = (target, all) => {
   for (var name in all)
@@ -14,6 +16,14 @@ var __copyProps = (to, from, except, desc) => {
   }
   return to;
 };
+var __toESM = (mod, isNodeMode, target) => (target = mod != null ? __create(__getProtoOf(mod)) : {}, __copyProps(
+  // If the importer is in node compatibility mode or this is not an ESM
+  // file that has been converted to a CommonJS file using a Babel-
+  // compatible transform (i.e. "__esModule" has not been set), then set
+  // "default" to the CommonJS "module.exports" for node compatibility.
+  isNodeMode || !mod || !mod.__esModule ? __defProp(target, "default", { value: mod, enumerable: true }) : target,
+  mod
+));
 var __toCommonJS = (mod) => __copyProps(__defProp({}, "__esModule", { value: true }), mod);
 
 // src/parsers/cnbeta.ts
@@ -22,6 +32,9 @@ __export(cnbeta_exports, {
   cnbeta: () => cnbeta
 });
 module.exports = __toCommonJS(cnbeta_exports);
+var import_axios = __toESM(require("axios"));
+var import_zx = require("zx");
+var import_cheerio = require("cheerio");
 var cnbeta = {
   parse: true,
   fetcher: "http",
@@ -42,6 +55,31 @@ var cnbeta = {
       description: rich.description.replace(`
  阅读全文`, "")
     };
+  },
+  ignoreProbeImage: true,
+  getThumbs: async (articles) => {
+    const res = await import_axios.default.get("https://m.cnbeta.com.tw/", {
+      responseType: "text"
+    }).catch((e) => {
+      console.error(import_zx.chalk.red("获取 m.cnbeta.com.tw 错误"), e);
+    });
+    if (!res)
+      return void 0;
+    const html = res.data;
+    const $ = (0, import_cheerio.load)(html, null, false);
+    const kv = {};
+    $("#it_tech .txt_thumb > a").each((_, $a) => {
+      var _a, _b;
+      const thumb = $($a).find("img").attr("src");
+      const href = $($a).attr("href") || "";
+      const reg = /([\d]+).htm/;
+      const articleId = (_a = reg.exec(href)) == null ? void 0 : _a[1];
+      const guid = (_b = articles.find((a) => a.guid.includes(`/${articleId}.htm`))) == null ? void 0 : _b.guid;
+      if (guid && thumb) {
+        kv[guid] = thumb;
+      }
+    });
+    return kv;
   }
 };
 // Annotate the CommonJS export names for ESM import in node:

package/dist/cjs/types.d.ts

@@ -10,6 +10,8 @@ export interface ParseStrategy {
     getContentElementFromArticle?: (aritcle: CheerioAPI) => Cheerio<AnyNode>;
     getContentFromHtml?: ($: CheerioAPI, node: Cheerio<AnyNode>) => string;
     getExtraItems?: ($: CheerioAPI, current: RichArticle, item: Item) => Promise<Record<string, any>>;
+    ignoreProbeImage?: boolean;
+    getThumbs?: (articles: Article[]) => Promise<Record<string, string> | undefined>;
 }
 export interface SimpleArticle {
     type: 'simple';
@@ -22,6 +24,7 @@ export interface SimpleArticle {
     author?: string;
     source?: string;
     categories?: string;
+    thumb?: string;
 }
 export interface RichArticleImage {
     url: string;

package/dist/cjs/upload/index.d.ts

@@ -21,4 +21,5 @@ export declare function uploadContent({ app, getUploadApi, items, }: {
     author?: string | undefined;
     source?: string | undefined;
     categories?: string | undefined;
+    thumb?: string | undefined;
 }[]>;

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "ns-rss-spider",
-  "version": "0.0.19",
+  "version": "0.0.21",
   "description": "",
   "main": "dist/cjs/index.js",
   "types": "dist/cjs/index.d.ts",