ns-rss-spider 0.0.2

package/README.md

@@ -0,0 +1,23 @@
+# rss-spider
+
+[![NPM version](https://img.shields.io/npm/v/rss-spider.svg?style=flat)](https://npmjs.org/package/rss-spider)
+[![NPM downloads](http://img.shields.io/npm/dm/rss-spider.svg?style=flat)](https://npmjs.org/package/rss-spider)
+
+## Install
+
+```bash
+$ yarn install
+```
+
+```bash
+$ npm run dev
+$ npm run build
+```
+
+## Options
+
+TODO
+
+## LICENSE
+
+MIT

package/dist/cjs/cli.d.ts

@@ -0,0 +1,2 @@
+#!/usr/bin/env node
+export {};

package/dist/cjs/cli.js

@@ -0,0 +1,61 @@
+#!/usr/bin/env node
+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 __copyProps = (to, from, except, desc) => {
+  if (from && typeof from === "object" || typeof from === "function") {
+    for (let key of __getOwnPropNames(from))
+      if (!__hasOwnProp.call(to, key) && key !== except)
+        __defProp(to, key, { get: () => from[key], enumerable: !(desc = __getOwnPropDesc(from, key)) || desc.enumerable });
+  }
+  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
+));
+
+// src/cli.ts
+var import_promises = require("fs/promises");
+var import_yargs_parser = __toESM(require("yargs-parser"));
+var import_index = require("./index");
+var import_zx = require("zx");
+var import_getData = require("./utils/getData");
+var argv = (0, import_yargs_parser.default)(process.argv);
+if (argv.v || argv.version) {
+  const pkg = require("../../package.json");
+  console.log(pkg.version);
+  process.exit(0);
+}
+async function main() {
+  const data = (0, import_getData.getData)() || {};
+  const { cmd, ...rest } = data;
+  let rst;
+  switch (cmd) {
+    case "parse":
+      rst = await (0, import_index.parseRss)(rest.name, rest.feed);
+      break;
+    default:
+      console.warn("未知命令", cmd);
+      return;
+  }
+  if (process.env.DISP_ARGV) {
+    console.info("写入结果文件");
+    return (0, import_promises.writeFile)("./result.json", JSON.stringify(rst || {}));
+  } else {
+    console.info("输出结果");
+    console.log(rst);
+  }
+}
+main().catch((err) => {
+  console.error(import_zx.chalk.red(err.message));
+  console.error(err);
+  process.exit(1);
+});

package/dist/cjs/index.d.ts

@@ -0,0 +1,3 @@
+export { parseRss } from "./parse";
+export { strategies } from "./strategy";
+export { parseContent } from "./parseContent";

package/dist/cjs/index.js

@@ -0,0 +1,35 @@
+var __defProp = Object.defineProperty;
+var __getOwnPropDesc = Object.getOwnPropertyDescriptor;
+var __getOwnPropNames = Object.getOwnPropertyNames;
+var __hasOwnProp = Object.prototype.hasOwnProperty;
+var __export = (target, all) => {
+  for (var name in all)
+    __defProp(target, name, { get: all[name], enumerable: true });
+};
+var __copyProps = (to, from, except, desc) => {
+  if (from && typeof from === "object" || typeof from === "function") {
+    for (let key of __getOwnPropNames(from))
+      if (!__hasOwnProp.call(to, key) && key !== except)
+        __defProp(to, key, { get: () => from[key], enumerable: !(desc = __getOwnPropDesc(from, key)) || desc.enumerable });
+  }
+  return to;
+};
+var __toCommonJS = (mod) => __copyProps(__defProp({}, "__esModule", { value: true }), mod);
+
+// src/index.ts
+var src_exports = {};
+__export(src_exports, {
+  parseContent: () => import_parseContent.parseContent,
+  parseRss: () => import_parse.parseRss,
+  strategies: () => import_strategy.strategies
+});
+module.exports = __toCommonJS(src_exports);
+var import_parse = require("./parse");
+var import_strategy = require("./strategy");
+var import_parseContent = require("./parseContent");
+// Annotate the CommonJS export names for ESM import in node:
+0 && (module.exports = {
+  parseContent,
+  parseRss,
+  strategies
+});

package/dist/cjs/parse.d.ts

@@ -0,0 +1 @@
+export declare function parseRss(name: string, feed: string): Promise<(import("./types").SimpleContent | import("./types").RichContent)[]>;

package/dist/cjs/parse.js

@@ -0,0 +1,59 @@
+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)
+    __defProp(target, name, { get: all[name], enumerable: true });
+};
+var __copyProps = (to, from, except, desc) => {
+  if (from && typeof from === "object" || typeof from === "function") {
+    for (let key of __getOwnPropNames(from))
+      if (!__hasOwnProp.call(to, key) && key !== except)
+        __defProp(to, key, { get: () => from[key], enumerable: !(desc = __getOwnPropDesc(from, key)) || desc.enumerable });
+  }
+  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/parse.ts
+var parse_exports = {};
+__export(parse_exports, {
+  parseRss: () => parseRss
+});
+module.exports = __toCommonJS(parse_exports);
+var import_rss_parser = __toESM(require("rss-parser"));
+var import_zx = require("zx");
+var import_strategy = require("./strategy");
+var import_parseContent = require("./parseContent");
+async function parseRss(name, feed) {
+  const parser = new import_rss_parser.default();
+  console.log(import_zx.chalk.blue("正在拉取 rss 列表"));
+  const result = await parser.parseURL(feed);
+  if (!result.items.length) {
+    throw Error("rss no conent");
+  }
+  const contents = [];
+  for (let item of result.items) {
+    console.log(import_zx.chalk.blue(`正在解析文章 【${item.title}】`));
+    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));
+    if (content) {
+      contents.push(content);
+    }
+  }
+  return contents;
+}
+// Annotate the CommonJS export names for ESM import in node:
+0 && (module.exports = {
+  parseRss
+});

package/dist/cjs/parseContent/getArticleHtml.d.ts

@@ -0,0 +1,2 @@
+import { ParseStrategy } from "../types";
+export declare const getArticleHtml: (url: string, fetcher: ParseStrategy["fetcher"]) => Promise<any>;

package/dist/cjs/parseContent/getArticleHtml.js

@@ -0,0 +1,49 @@
+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)
+    __defProp(target, name, { get: all[name], enumerable: true });
+};
+var __copyProps = (to, from, except, desc) => {
+  if (from && typeof from === "object" || typeof from === "function") {
+    for (let key of __getOwnPropNames(from))
+      if (!__hasOwnProp.call(to, key) && key !== except)
+        __defProp(to, key, { get: () => from[key], enumerable: !(desc = __getOwnPropDesc(from, key)) || desc.enumerable });
+  }
+  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/parseContent/getArticleHtml.ts
+var getArticleHtml_exports = {};
+__export(getArticleHtml_exports, {
+  getArticleHtml: () => getArticleHtml
+});
+module.exports = __toCommonJS(getArticleHtml_exports);
+var import_axios = __toESM(require("axios"));
+var getArticleHtml = async (url, fetcher) => {
+  if (fetcher === "playwright") {
+    throw Error("todo");
+  } else {
+    const rst = await import_axios.default.get(url, {
+      responseType: "text"
+    });
+    return rst.data;
+  }
+};
+// Annotate the CommonJS export names for ESM import in node:
+0 && (module.exports = {
+  getArticleHtml
+});

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

@@ -0,0 +1,16 @@
+import { Item as RssItem } from "rss-parser";
+import { Content, ParseStrategy } from "../types";
+/**
+ * 解析文章内容
+ * 1. 获取 html 片段。(rss、http、playwright)
+ * 2. 对 html 处理
+ *  清理无用样式
+ * 3. 从 html 计算 images、videos、parsed(todo)
+ * 4. 获取自定义的额外解析项(可选)
+ * 5. 返回结果
+ *
+ * @param item
+ * @param strategy
+ * @returns
+ */
+export declare function parseContent(item: RssItem, strategy?: ParseStrategy): Promise<Content>;

package/dist/cjs/parseContent/index.js

@@ -0,0 +1,80 @@
+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)
+    __defProp(target, name, { get: all[name], enumerable: true });
+};
+var __copyProps = (to, from, except, desc) => {
+  if (from && typeof from === "object" || typeof from === "function") {
+    for (let key of __getOwnPropNames(from))
+      if (!__hasOwnProp.call(to, key) && key !== except)
+        __defProp(to, key, { get: () => from[key], enumerable: !(desc = __getOwnPropDesc(from, key)) || desc.enumerable });
+  }
+  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/parseContent/index.ts
+var parseContent_exports = {};
+__export(parseContent_exports, {
+  parseContent: () => parseContent
+});
+module.exports = __toCommonJS(parseContent_exports);
+var import_getArticleHtml = require("./getArticleHtml");
+var cheerio = __toESM(require("cheerio"));
+var import_stripeHtml = require("./stripeHtml");
+var import_parseContent = require("./parseContent");
+var import_zx = require("zx");
+async function parseContent(item, strategy) {
+  var _a, _b;
+  let element;
+  let $;
+  if (!strategy || !strategy.fetcher) {
+    $ = cheerio.load(item.content || "", {}, false);
+    element = $.root();
+  } else {
+    console.log(import_zx.chalk.blue("正在拉取文章内容"), item.link);
+    const article = await (0, import_getArticleHtml.getArticleHtml)(item.link, strategy.fetcher);
+    $ = cheerio.load(article);
+    element = ((_a = strategy.getContentHtmlFromArticle) == null ? void 0 : _a.call(strategy, $)) || $.root();
+  }
+  console.log(import_zx.chalk.blue("正在预处理 html"));
+  (0, import_stripeHtml.stripeHtml)($, element);
+  if (!(strategy == null ? void 0 : strategy.parse)) {
+    return {
+      type: "simple",
+      title: item.title,
+      description: ((_b = element.html()) == null ? void 0 : _b.trim()) || "",
+      pubDate: new Date(item.pubDate)
+    };
+  }
+  console.log(import_zx.chalk.blue("正在解析文章内容"));
+  const rich = await (0, import_parseContent.parseContent)($, element, item, strategy);
+  let extra = {};
+  if (strategy.getExtraItems) {
+    console.log(import_zx.chalk.blue("正在执行 getExtraItems"));
+    extra = await strategy.getExtraItems($, rich, item).catch((e) => {
+      console.error(import_zx.chalk.red("getExtraItems 识别"), e);
+    }) || {};
+  }
+  return {
+    ...rich,
+    ...extra
+  };
+}
+// Annotate the CommonJS export names for ESM import in node:
+0 && (module.exports = {
+  parseContent
+});

package/dist/cjs/parseContent/parseContent.d.ts

@@ -0,0 +1,4 @@
+import { Item as RssItem } from "rss-parser";
+import type { ParseStrategy, RichContent } from "../types";
+import { AnyNode, Cheerio, CheerioAPI } from "cheerio";
+export declare function parseContent($: CheerioAPI, $element: Cheerio<AnyNode>, item: RssItem, strategy?: ParseStrategy): Promise<RichContent>;

package/dist/cjs/parseContent/parseContent.js

@@ -0,0 +1,108 @@
+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)
+    __defProp(target, name, { get: all[name], enumerable: true });
+};
+var __copyProps = (to, from, except, desc) => {
+  if (from && typeof from === "object" || typeof from === "function") {
+    for (let key of __getOwnPropNames(from))
+      if (!__hasOwnProp.call(to, key) && key !== except)
+        __defProp(to, key, { get: () => from[key], enumerable: !(desc = __getOwnPropDesc(from, key)) || desc.enumerable });
+  }
+  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/parseContent/parseContent.ts
+var parseContent_exports = {};
+__export(parseContent_exports, {
+  parseContent: () => parseContent
+});
+module.exports = __toCommonJS(parseContent_exports);
+var import_utils = require("./utils");
+var import_probe_image_size = __toESM(require("probe-image-size"));
+var import_zx = require("zx");
+var import_html_entities = require("html-entities");
+async function parseContent($, $element, item, strategy) {
+  var _a;
+  const srcs = [];
+  const images = [];
+  (0, import_utils.walk_the_DOM)($, $element, (node) => {
+    if (node.is("img")) {
+      const src = node.attr("src");
+      const title = node.attr("title") || void 0;
+      if (src) {
+        srcs.push({
+          src,
+          title
+        });
+      }
+    }
+  });
+  for (let item2 of srcs) {
+    console.log(import_zx.chalk.blue("正在解析图片", item2.src));
+    const result = await (0, import_probe_image_size.default)(item2.src, {
+      rejectUnauthorized: false,
+      open_timeout: 1e3 * 5
+    }).catch((e) => {
+      console.error("解析图片失败", e.message);
+      return Error((e == null ? void 0 : e.message) || e || "unknown error");
+    });
+    if (result instanceof Error) {
+      images.push({
+        url: item2.src,
+        title: item2.title
+      });
+    } else {
+      images.push({
+        url: item2.src,
+        title: item2.title,
+        width: result.width,
+        height: result.height,
+        type: result.mime,
+        widthUnit: result.wUnits !== "px" ? result.wUnits : void 0,
+        heightUnit: result.hUnits !== "px" ? result.hUnits : void 0,
+        realUrl: result.url !== item2.src ? result.url : void 0
+      });
+    }
+  }
+  $($element).find("img").each((_, img) => {
+    const src = $(img).attr("src");
+    if (src) {
+      $(img).removeAttr("src");
+      $(img).attr("data-src", src);
+      const meta = images.find((i) => i.url === src);
+      if (meta == null ? void 0 : meta.width) {
+        $(img).attr("data-w", `${meta.width}`);
+      }
+      if (meta == null ? void 0 : meta.height) {
+        $(img).attr("data-h", `${meta.height}`);
+      }
+    }
+  });
+  return {
+    type: "rich",
+    title: item.title,
+    description: ((_a = item.contentSnippet) == null ? void 0 : _a.trim()) || "",
+    content: (0, import_html_entities.decode)($element.html().trim()),
+    pubDate: new Date(item.pubDate),
+    images
+  };
+}
+// Annotate the CommonJS export names for ESM import in node:
+0 && (module.exports = {
+  parseContent
+});

package/dist/cjs/parseContent/stripeHtml.d.ts

@@ -0,0 +1,2 @@
+import type { Cheerio, AnyNode, CheerioAPI } from "cheerio";
+export declare function stripeHtml($: CheerioAPI, element: Cheerio<AnyNode>): void;

package/dist/cjs/parseContent/stripeHtml.js

@@ -0,0 +1,46 @@
+var __defProp = Object.defineProperty;
+var __getOwnPropDesc = Object.getOwnPropertyDescriptor;
+var __getOwnPropNames = Object.getOwnPropertyNames;
+var __hasOwnProp = Object.prototype.hasOwnProperty;
+var __export = (target, all) => {
+  for (var name in all)
+    __defProp(target, name, { get: all[name], enumerable: true });
+};
+var __copyProps = (to, from, except, desc) => {
+  if (from && typeof from === "object" || typeof from === "function") {
+    for (let key of __getOwnPropNames(from))
+      if (!__hasOwnProp.call(to, key) && key !== except)
+        __defProp(to, key, { get: () => from[key], enumerable: !(desc = __getOwnPropDesc(from, key)) || desc.enumerable });
+  }
+  return to;
+};
+var __toCommonJS = (mod) => __copyProps(__defProp({}, "__esModule", { value: true }), mod);
+
+// src/parseContent/stripeHtml.ts
+var stripeHtml_exports = {};
+__export(stripeHtml_exports, {
+  stripeHtml: () => stripeHtml
+});
+module.exports = __toCommonJS(stripeHtml_exports);
+var import_utils = require("./utils");
+function stripeHtml($, element) {
+  $(element).find("script").remove();
+  (0, import_utils.walk_the_DOM)($, element, (el) => {
+    for (let attr of [
+      "id",
+      "style",
+      "class",
+      "width",
+      "height"
+    ]) {
+      $(el).removeAttr(attr);
+    }
+    if ($(el).is("a")) {
+      $(el).removeAttr("target");
+    }
+  });
+}
+// Annotate the CommonJS export names for ESM import in node:
+0 && (module.exports = {
+  stripeHtml
+});

package/dist/cjs/parseContent/utils.d.ts

@@ -0,0 +1,2 @@
+import { AnyNode, Cheerio, CheerioAPI } from "cheerio";
+export declare function walk_the_DOM($: CheerioAPI, n: Cheerio<AnyNode>, func: (el: Cheerio<AnyNode>) => void): void;

package/dist/cjs/parseContent/utils.js

@@ -0,0 +1,39 @@
+var __defProp = Object.defineProperty;
+var __getOwnPropDesc = Object.getOwnPropertyDescriptor;
+var __getOwnPropNames = Object.getOwnPropertyNames;
+var __hasOwnProp = Object.prototype.hasOwnProperty;
+var __export = (target, all) => {
+  for (var name in all)
+    __defProp(target, name, { get: all[name], enumerable: true });
+};
+var __copyProps = (to, from, except, desc) => {
+  if (from && typeof from === "object" || typeof from === "function") {
+    for (let key of __getOwnPropNames(from))
+      if (!__hasOwnProp.call(to, key) && key !== except)
+        __defProp(to, key, { get: () => from[key], enumerable: !(desc = __getOwnPropDesc(from, key)) || desc.enumerable });
+  }
+  return to;
+};
+var __toCommonJS = (mod) => __copyProps(__defProp({}, "__esModule", { value: true }), mod);
+
+// src/parseContent/utils.ts
+var utils_exports = {};
+__export(utils_exports, {
+  walk_the_DOM: () => walk_the_DOM
+});
+module.exports = __toCommonJS(utils_exports);
+function walk_the_DOM($, n, func) {
+  let node = n;
+  if (node.length === 0) {
+    return;
+  }
+  func(node);
+  const children = node.children();
+  $(children).each((_, c) => {
+    walk_the_DOM($, $(c), func);
+  });
+}
+// Annotate the CommonJS export names for ESM import in node:
+0 && (module.exports = {
+  walk_the_DOM
+});

package/dist/cjs/parsers/cnbeta.d.ts

@@ -0,0 +1,2 @@
+import { ParseStrategy } from "../types";
+export declare const cnbeta: ParseStrategy;

package/dist/cjs/parsers/cnbeta.js

@@ -0,0 +1,44 @@
+var __defProp = Object.defineProperty;
+var __getOwnPropDesc = Object.getOwnPropertyDescriptor;
+var __getOwnPropNames = Object.getOwnPropertyNames;
+var __hasOwnProp = Object.prototype.hasOwnProperty;
+var __export = (target, all) => {
+  for (var name in all)
+    __defProp(target, name, { get: all[name], enumerable: true });
+};
+var __copyProps = (to, from, except, desc) => {
+  if (from && typeof from === "object" || typeof from === "function") {
+    for (let key of __getOwnPropNames(from))
+      if (!__hasOwnProp.call(to, key) && key !== except)
+        __defProp(to, key, { get: () => from[key], enumerable: !(desc = __getOwnPropDesc(from, key)) || desc.enumerable });
+  }
+  return to;
+};
+var __toCommonJS = (mod) => __copyProps(__defProp({}, "__esModule", { value: true }), mod);
+
+// src/parsers/cnbeta.ts
+var cnbeta_exports = {};
+__export(cnbeta_exports, {
+  cnbeta: () => cnbeta
+});
+module.exports = __toCommonJS(cnbeta_exports);
+var cnbeta = {
+  parse: true,
+  fetcher: "http",
+  getContentHtmlFromArticle: ($) => {
+    const el = $(".article-content");
+    el.remove(".tac");
+    return el;
+  },
+  getExtraItems: async ($, rich) => {
+    return {
+      description: rich.description.replace(`
+ 阅读全文`, ""),
+      pageFrom: $(".cnbeta-article .title .source a").text().toLowerCase()
+    };
+  }
+};
+// Annotate the CommonJS export names for ESM import in node:
+0 && (module.exports = {
+  cnbeta
+});

package/dist/cjs/parsers/ifanr.d.ts

@@ -0,0 +1,2 @@
+import { ParseStrategy } from "../types";
+export declare const ifanr: ParseStrategy;

package/dist/cjs/parsers/ifanr.js

@@ -0,0 +1,36 @@
+var __defProp = Object.defineProperty;
+var __getOwnPropDesc = Object.getOwnPropertyDescriptor;
+var __getOwnPropNames = Object.getOwnPropertyNames;
+var __hasOwnProp = Object.prototype.hasOwnProperty;
+var __export = (target, all) => {
+  for (var name in all)
+    __defProp(target, name, { get: all[name], enumerable: true });
+};
+var __copyProps = (to, from, except, desc) => {
+  if (from && typeof from === "object" || typeof from === "function") {
+    for (let key of __getOwnPropNames(from))
+      if (!__hasOwnProp.call(to, key) && key !== except)
+        __defProp(to, key, { get: () => from[key], enumerable: !(desc = __getOwnPropDesc(from, key)) || desc.enumerable });
+  }
+  return to;
+};
+var __toCommonJS = (mod) => __copyProps(__defProp({}, "__esModule", { value: true }), mod);
+
+// src/parsers/ifanr.ts
+var ifanr_exports = {};
+__export(ifanr_exports, {
+  ifanr: () => ifanr
+});
+module.exports = __toCommonJS(ifanr_exports);
+var ifanr = {
+  parse: true,
+  fetcher: "http",
+  getContentHtmlFromArticle: ($) => {
+    const el = $("#article-content-wrapper article");
+    return el;
+  }
+};
+// Annotate the CommonJS export names for ESM import in node:
+0 && (module.exports = {
+  ifanr
+});

package/dist/cjs/parsers/ithome.d.ts

@@ -0,0 +1,2 @@
+import { ParseStrategy } from "../types";
+export declare const ithome: ParseStrategy;

package/dist/cjs/parsers/ithome.js

@@ -0,0 +1,53 @@
+var __defProp = Object.defineProperty;
+var __getOwnPropDesc = Object.getOwnPropertyDescriptor;
+var __getOwnPropNames = Object.getOwnPropertyNames;
+var __hasOwnProp = Object.prototype.hasOwnProperty;
+var __export = (target, all) => {
+  for (var name in all)
+    __defProp(target, name, { get: all[name], enumerable: true });
+};
+var __copyProps = (to, from, except, desc) => {
+  if (from && typeof from === "object" || typeof from === "function") {
+    for (let key of __getOwnPropNames(from))
+      if (!__hasOwnProp.call(to, key) && key !== except)
+        __defProp(to, key, { get: () => from[key], enumerable: !(desc = __getOwnPropDesc(from, key)) || desc.enumerable });
+  }
+  return to;
+};
+var __toCommonJS = (mod) => __copyProps(__defProp({}, "__esModule", { value: true }), mod);
+
+// src/parsers/ithome.ts
+var ithome_exports = {};
+__export(ithome_exports, {
+  ithome: () => ithome
+});
+module.exports = __toCommonJS(ithome_exports);
+var import_utils = require("../parseContent/utils");
+var ithome = {
+  parse: true,
+  fetcher: "http",
+  getContentHtmlFromArticle: ($) => {
+    const el = $("#paragraph");
+    $(el).find("img").each((_, img) => {
+      const original = $(img).attr("data-original");
+      if (original) {
+        $(img).attr("src", original);
+        $(img).removeAttr("data-original");
+      }
+    });
+    [".dy-live-bar", ".tougao-user", ".ad-tips", "dir"].forEach((sel) => {
+      $(el).find(sel).remove();
+    });
+    (0, import_utils.walk_the_DOM)($, $(el), (c) => {
+      $(c).removeAttr("data-vmark");
+    });
+    return $(el);
+  },
+  getExtraItems: async ($, rich, item) => {
+    return {};
+  }
+};
+// Annotate the CommonJS export names for ESM import in node:
+0 && (module.exports = {
+  ithome
+});

package/dist/cjs/parsers/theverge.d.ts

@@ -0,0 +1,2 @@
+import { ParseStrategy } from "../types";
+export declare const theverge: ParseStrategy;

package/dist/cjs/parsers/theverge.js

@@ -0,0 +1,31 @@
+var __defProp = Object.defineProperty;
+var __getOwnPropDesc = Object.getOwnPropertyDescriptor;
+var __getOwnPropNames = Object.getOwnPropertyNames;
+var __hasOwnProp = Object.prototype.hasOwnProperty;
+var __export = (target, all) => {
+  for (var name in all)
+    __defProp(target, name, { get: all[name], enumerable: true });
+};
+var __copyProps = (to, from, except, desc) => {
+  if (from && typeof from === "object" || typeof from === "function") {
+    for (let key of __getOwnPropNames(from))
+      if (!__hasOwnProp.call(to, key) && key !== except)
+        __defProp(to, key, { get: () => from[key], enumerable: !(desc = __getOwnPropDesc(from, key)) || desc.enumerable });
+  }
+  return to;
+};
+var __toCommonJS = (mod) => __copyProps(__defProp({}, "__esModule", { value: true }), mod);
+
+// src/parsers/theverge.ts
+var theverge_exports = {};
+__export(theverge_exports, {
+  theverge: () => theverge
+});
+module.exports = __toCommonJS(theverge_exports);
+var theverge = {
+  parse: true
+};
+// Annotate the CommonJS export names for ESM import in node:
+0 && (module.exports = {
+  theverge
+});

package/dist/cjs/strategy.d.ts

@@ -0,0 +1,2 @@
+import { ParseStrategy } from "./types";
+export declare const strategies: Record<string, ParseStrategy>;

package/dist/cjs/strategy.js

@@ -0,0 +1,38 @@
+var __defProp = Object.defineProperty;
+var __getOwnPropDesc = Object.getOwnPropertyDescriptor;
+var __getOwnPropNames = Object.getOwnPropertyNames;
+var __hasOwnProp = Object.prototype.hasOwnProperty;
+var __export = (target, all) => {
+  for (var name in all)
+    __defProp(target, name, { get: all[name], enumerable: true });
+};
+var __copyProps = (to, from, except, desc) => {
+  if (from && typeof from === "object" || typeof from === "function") {
+    for (let key of __getOwnPropNames(from))
+      if (!__hasOwnProp.call(to, key) && key !== except)
+        __defProp(to, key, { get: () => from[key], enumerable: !(desc = __getOwnPropDesc(from, key)) || desc.enumerable });
+  }
+  return to;
+};
+var __toCommonJS = (mod) => __copyProps(__defProp({}, "__esModule", { value: true }), mod);
+
+// src/strategy.ts
+var strategy_exports = {};
+__export(strategy_exports, {
+  strategies: () => strategies
+});
+module.exports = __toCommonJS(strategy_exports);
+var import_cnbeta = require("./parsers/cnbeta");
+var import_ifanr = require("./parsers/ifanr");
+var import_theverge = require("./parsers/theverge");
+var import_ithome = require("./parsers/ithome");
+var strategies = {
+  cnbeta: import_cnbeta.cnbeta,
+  ifanr: import_ifanr.ifanr,
+  theverge: import_theverge.theverge,
+  ithome: import_ithome.ithome
+};
+// Annotate the CommonJS export names for ESM import in node:
+0 && (module.exports = {
+  strategies
+});

package/dist/cjs/types.d.ts

@@ -0,0 +1,30 @@
+import type { AnyNode, Cheerio, CheerioAPI } from 'cheerio';
+import { Item } from 'rss-parser';
+export interface ParseStrategy {
+    parse: boolean;
+    fetcher?: 'http' | 'playwright';
+    getContentHtmlFromArticle?: (aritcle: CheerioAPI) => Cheerio<AnyNode>;
+    getExtraItems?: (html: CheerioAPI, current: RichContent, item: Item) => Promise<Record<string, any>>;
+}
+export interface SimpleContent {
+    type: 'simple';
+    title: string;
+    description: string;
+    pubDate: Date;
+}
+export interface RichContentImage {
+    url: string;
+    type?: string;
+    width?: number;
+    height?: number;
+    realUrl?: string;
+    widthUnit?: string;
+    heightUnit?: string;
+    title?: string;
+}
+export interface RichContent extends Omit<SimpleContent, 'type'> {
+    type: 'rich';
+    content: string;
+    images: RichContentImage[];
+}
+export type Content = SimpleContent | RichContent;

package/dist/cjs/types.js

@@ -0,0 +1,17 @@
+var __defProp = Object.defineProperty;
+var __getOwnPropDesc = Object.getOwnPropertyDescriptor;
+var __getOwnPropNames = Object.getOwnPropertyNames;
+var __hasOwnProp = Object.prototype.hasOwnProperty;
+var __copyProps = (to, from, except, desc) => {
+  if (from && typeof from === "object" || typeof from === "function") {
+    for (let key of __getOwnPropNames(from))
+      if (!__hasOwnProp.call(to, key) && key !== except)
+        __defProp(to, key, { get: () => from[key], enumerable: !(desc = __getOwnPropDesc(from, key)) || desc.enumerable });
+  }
+  return to;
+};
+var __toCommonJS = (mod) => __copyProps(__defProp({}, "__esModule", { value: true }), mod);
+
+// src/types.ts
+var types_exports = {};
+module.exports = __toCommonJS(types_exports);

package/dist/cjs/utils/getData.d.ts

@@ -0,0 +1 @@
+export declare const getData: (str?: string) => any;

package/dist/cjs/utils/getData.js

@@ -0,0 +1,63 @@
+"use strict";
+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)
+    __defProp(target, name, { get: all[name], enumerable: true });
+};
+var __copyProps = (to, from, except, desc) => {
+  if (from && typeof from === "object" || typeof from === "function") {
+    for (let key of __getOwnPropNames(from))
+      if (!__hasOwnProp.call(to, key) && key !== except)
+        __defProp(to, key, { get: () => from[key], enumerable: !(desc = __getOwnPropDesc(from, key)) || desc.enumerable });
+  }
+  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/utils/getData.ts
+var getData_exports = {};
+__export(getData_exports, {
+  getData: () => getData
+});
+module.exports = __toCommonJS(getData_exports);
+var import_yargs_parser = __toESM(require("yargs-parser"));
+var getData = function(str) {
+  let argv_str = "";
+  if (str) {
+    argv_str = str;
+  } else if (process.env.DISP_ARGV) {
+    argv_str = `--data=${process.env.DISP_ARGV}`;
+  } else if (process.argv) {
+    argv_str = process.argv;
+  }
+  ;
+  const argv = (0, import_yargs_parser.default)(argv_str);
+  if (argv.data && typeof argv.data == "string") {
+    try {
+      const data = JSON.parse(decodeURIComponent(Buffer.from(argv.data, "base64").toString()));
+      return data;
+    } catch (e) {
+      return null;
+    }
+    ;
+  }
+  ;
+  return null;
+};
+// Annotate the CommonJS export names for ESM import in node:
+0 && (module.exports = {
+  getData
+});

package/package.json

@@ -0,0 +1,43 @@
+{
+  "name": "ns-rss-spider",
+  "version": "0.0.2",
+  "description": "",
+  "main": "dist/cjs/index.js",
+  "types": "dist/cjs/index.d.ts",
+  "scripts": {
+    "dev": "father dev",
+    "test": "vitest run",
+    "build": "father build",
+    "build:deps": "father prebundle",
+    "check": "father doctor",
+    "prepublishOnly": "father doctor && npm run build",
+    "pub": "npm run check && npm version patch && npm run build && npm publish --registry=https://registry.npmjs.org && git push origin master --tags && npm run sync"
+  },
+  "bin": {
+    "ns-rss-spider": "./dist/cjs/cli.js"
+  },
+  "keywords": [],
+  "authors": [],
+  "license": "MIT",
+  "files": [
+    "dist",
+    "compiled"
+  ],
+  "publishConfig": {
+    "access": "public"
+  },
+  "devDependencies": {
+    "@types/probe-image-size": "^7.2.4",
+    "father": "^4.4.0",
+    "vitest": "^1.2.2"
+  },
+  "dependencies": {
+    "axios": "^1.6.7",
+    "cheerio": "^1.0.0-rc.12",
+    "html-entities": "^2.4.0",
+    "probe-image-size": "^7.2.3",
+    "rss-parser": "^3.13.0",
+    "yargs-parser": "^21.1.1",
+    "zx": "4.x"
+  }
+}