@aiquants/html-to-markdown 0.1.0

package/dist/index.cjs

@@ -0,0 +1 @@
+"use strict";Object.defineProperty(exports,Symbol.toStringTag,{value:"Module"});const e=require("./core-D1CksXT5.cjs");exports.getMessage=e.getMessage,exports.htmlToMarkdown=e.htmlToMarkdown;

package/dist/index.js

@@ -0,0 +1,5 @@
+import { g, h } from "./core-C45e4gDg.js";
+export {
+  g as getMessage,
+  h as htmlToMarkdown
+};

package/dist/main.cjs

@@ -0,0 +1,2 @@
+#!/usr/bin/env node
+"use strict";Object.create,Object.defineProperty,Object.getOwnPropertyDescriptor,Object.getOwnPropertyNames,Object.getPrototypeOf,Object.prototype.hasOwnProperty;const e=require("./core-D1CksXT5.cjs");var t="undefined"!=typeof document?document.currentScript:null;(("undefined"==typeof document?require("url").pathToFileURL(__filename).href:t&&"SCRIPT"===t.tagName.toUpperCase()&&t.src||new URL("main.cjs",document.baseURI).href).endsWith(process.argv[1])||process.argv[1]&&process.argv[1].includes("vite-node"))&&(async()=>{const t=(await import("yargs-parser")).default,r=await import("node:fs"),c=await import("node:path"),o=t(process.argv.slice(2),{string:["locale"],default:{locale:"en-US"}}),s=o._[0],a=o.locale;s||process.exit(1);try{const t=await e.htmlToMarkdown(s,{locale:a}),o=`${s.replace(/https?:\/\//,"").replace(/[^a-zA-Z0-9]/g,"_")}_${Date.now()}.md`,n=c.join(process.cwd(),".outputs","raw");r.existsSync(n)||r.mkdirSync(n,{recursive:!0});const i=c.join(n,o);r.writeFileSync(i,t)}catch(n){process.exit(1)}})();

package/dist/main.d.ts

@@ -0,0 +1,6 @@
+#!/usr/bin/env node
+/**
+ * Entry point for htmlToMarkdown CLI.
+ * htmlToMarkdown CLI のエントリーポイント
+ */
+export {};

package/dist/main.js

@@ -0,0 +1,53 @@
+#!/usr/bin/env node
+import { g as getMessage, h as htmlToMarkdown } from "./core-C45e4gDg.js";
+const runCli = async () => {
+  const yargsParser = (await import("yargs-parser")).default;
+  const fs = await import("node:fs");
+  const path = await import("node:path");
+  const argv = yargsParser(process.argv.slice(2), {
+    string: ["locale"],
+    default: {
+      locale: "en-US"
+      // デフォルトのロケール
+    }
+  });
+  const targetUrl = argv._[0];
+  const locale = argv.locale;
+  if (!targetUrl) {
+    console.error(getMessage(locale, "error_url_required"));
+    console.info(getMessage(locale, "usage"));
+    console.info(getMessage(locale, "example"));
+    process.exit(1);
+  }
+  try {
+    console.info(`
+${getMessage(locale, "cli_start")}`);
+    console.info(getMessage(locale, "cli_url", { url: targetUrl }));
+    console.info(getMessage(locale, "cli_locale", { locale }));
+    const markdown = await htmlToMarkdown(targetUrl, { locale });
+    const safeFileName = targetUrl.replace(/https?:\/\//, "").replace(/[^a-zA-Z0-9]/g, "_");
+    const timestamp = Date.now();
+    const fileName = `${safeFileName}_${timestamp}.md`;
+    const outputDir = path.join(process.cwd(), ".outputs", "raw");
+    if (!fs.existsSync(outputDir)) {
+      fs.mkdirSync(outputDir, { recursive: true });
+    }
+    const filePath = path.join(outputDir, fileName);
+    fs.writeFileSync(filePath, markdown);
+    console.info(`
+${getMessage(locale, "success_save")}`);
+    console.info(getMessage(locale, "success_path", { path: filePath }));
+  } catch (error) {
+    console.error(`
+${getMessage(locale, "error_app")}`, error);
+    process.exit(1);
+  }
+};
+const isDirectRun = (
+  // `node dist/main.js` や `npx` 経由での実行を判定
+  import.meta.url.endsWith(process.argv[1]) || // `vite-node main.ts` 経由での実行を判定
+  process.argv[1] && process.argv[1].includes("vite-node")
+);
+if (isDirectRun) {
+  runCli();
+}

package/dist/src/browser.d.ts

@@ -0,0 +1,11 @@
+/**
+ * Browser automation module using Playwright.
+ * Playwright を使用したブラウザ自動化モジュール
+ */
+/**
+ * Playwrightを使用してURLから動的にレンダリングされたHTMLを取得する
+ * @param url 取得対象のURL
+ * @param locale ブラウザのロケール
+ * @returns HTMLコンテンツ
+ */
+export declare const getHtmlWithPlaywright: (url: string, locale: string) => Promise<string>;

package/dist/src/cli.d.ts

@@ -0,0 +1,9 @@
+/**
+ * Command Line Interface module.
+ * コマンドラインインターフェースモジュール
+ */
+/**
+ * CLI main function.
+ * CLI のメイン関数
+ */
+export declare const runCli: () => Promise<void>;

package/dist/src/converter.d.ts

@@ -0,0 +1,12 @@
+/**
+ * HTML to Markdown conversion module.
+ * HTML から Markdown への変換モジュール
+ */
+/**
+ * HTML文字列をMarkdownに変換する
+ * @param html HTMLコンテンツ
+ * @param baseUrl リンクを絶対パスに変換するためのベースURL
+ * @param locale ロケール
+ * @returns Markdownテキスト
+ */
+export declare const convertHtmlToMarkdown: (html: string, baseUrl: string, locale: string) => Promise<string>;

package/dist/src/core.d.ts

@@ -0,0 +1,10 @@
+import { HtmlToMarkdownOptions } from './types.js';
+/**
+ * Converts the HTML content of a given URL to Markdown.
+ * 指定されたURLのHTMLコンテンツをMarkdownに変換します。
+ *
+ * @param url The URL of the web page to convert. / 変換対象のWebページのURL。
+ * @param options Options for the conversion process. / 変換プロセスのオプション。
+ * @returns A promise that resolves to the Markdown string. / Markdown文字列を解決するPromise。
+ */
+export declare const htmlToMarkdown: (url: string, options?: HtmlToMarkdownOptions) => Promise<string>;

package/dist/src/i18n.d.ts

@@ -0,0 +1,12 @@
+import { MessageKey } from './types.js';
+/**
+ * Gets a localized message.
+ * ローカライズされたメッセージを取得します。
+ * @param locale The locale to use.
+ * @param key The message key.
+ * @param replacements Optional replacements for placeholders.
+ * @returns The localized message string.
+ */
+export declare const getMessage: (locale: string, key: MessageKey, replacements?: {
+    [key: string]: string;
+}) => string;

package/dist/src/index.d.ts

@@ -0,0 +1,7 @@
+/**
+ * Main export module for htmlToMarkdown library.
+ * htmlToMarkdown ライブラリのメインエクスポートモジュール
+ */
+export { htmlToMarkdown } from './core.js';
+export type { HtmlToMarkdownOptions } from './types.js';
+export { getMessage } from './i18n.js';

package/dist/src/plugins/index.d.ts

@@ -0,0 +1,7 @@
+/**
+ * Rehype plugins index.
+ * Rehype プラグインのインデックス
+ */
+export { rehypeSanitizeHtml } from './rehype-sanitize-html.js';
+export { rehypeAbsoluteLinks } from './rehype-absolute-links.js';
+export { rehypeWikipediaFootnotes } from './rehype-wikipedia-footnotes.js';

package/dist/src/plugins/rehype-absolute-links.d.ts

@@ -0,0 +1,9 @@
+import { Root } from 'hast';
+/**
+ * A custom rehype plugin to make relative links absolute for various tags.
+ * @param options - The base URL to prepend to relative links.
+ * @returns A unified transformer function.
+ */
+export declare const rehypeAbsoluteLinks: (options: {
+    baseUrl: string;
+}) => (tree: Root) => void;

package/dist/src/plugins/rehype-sanitize-html.d.ts

@@ -0,0 +1,7 @@
+import { Root } from 'hast';
+/**
+ * A custom rehype plugin to remove empty comments and clean up whitespace.
+ * - 空のコメントノードを削除します (`<!-- -->`)
+ * - 隣接するテキストノードを結合し、連続する空白を単一のスペースに正規化します。
+ */
+export declare const rehypeSanitizeHtml: () => (tree: Root) => void;

package/dist/src/plugins/rehype-wikipedia-footnotes.d.ts

@@ -0,0 +1,6 @@
+import { Element } from 'hast';
+/**
+ * A custom rehype plugin to transform Wikipedia-style footnotes into standard Markdown footnotes.
+ * @returns A unified transformer function.
+ */
+export declare const rehypeWikipediaFootnotes: () => (tree: Element) => void;

package/dist/src/types.d.ts

@@ -0,0 +1,25 @@
+/**
+ * Type definitions for htmlToMarkdown module.
+ * htmlToMarkdown モジュールの型定義
+ */
+/**
+ * Options for the htmlToMarkdown function.
+ * htmlToMarkdown 関数のオプション
+ */
+export interface HtmlToMarkdownOptions {
+    /**
+     * The locale to use for the browser context. Defaults to 'en-US'.
+     * ブラウザコンテキストで使用するロケール。デフォルトは 'en-US' です。
+     */
+    locale?: string;
+}
+/**
+ * Message keys for i18n.
+ * i18n 用のメッセージキー
+ */
+export type MessageKey = "error_url_required" | "usage" | "example" | "cli_start" | "cli_url" | "cli_locale" | "success_save" | "success_path" | "error_app" | "playwright_launch" | "playwright_goto" | "playwright_get_content" | "playwright_close" | "playwright_error" | "convert_start" | "convert_success";
+/**
+ * Supported locales.
+ * サポートされているロケール
+ */
+export type SupportedLocale = "ja-JP" | "en-US";

package/dist/tests/100-percent-coverage.test.d.ts

@@ -0,0 +1 @@
+export {};

package/dist/tests/100-percent-final.test.d.ts

@@ -0,0 +1 @@
+export {};

package/dist/tests/absolute-final-coverage.test.d.ts

@@ -0,0 +1 @@
+export {};

package/dist/tests/browser.test.d.ts

@@ -0,0 +1 @@
+export {};

package/dist/tests/cli.test.d.ts

@@ -0,0 +1 @@
+export {};

package/dist/tests/converter.test.d.ts

@@ -0,0 +1 @@
+export {};

package/dist/tests/core.test.d.ts

@@ -0,0 +1 @@
+export {};

package/dist/tests/coverage-edge-cases.test.d.ts

@@ -0,0 +1 @@
+export {};

package/dist/tests/direct-coverage.test.d.ts

@@ -0,0 +1 @@
+export {};

package/dist/tests/final-100-percent.test.d.ts

@@ -0,0 +1 @@
+export {};

package/dist/tests/final-coverage.test.d.ts

@@ -0,0 +1 @@
+export {};

package/dist/tests/i18n.test.d.ts

@@ -0,0 +1 @@
+export {};

package/dist/tests/index.test.d.ts

@@ -0,0 +1 @@
+export {};

package/dist/tests/line-70-coverage.test.d.ts

@@ -0,0 +1 @@
+export {};

package/dist/tests/main.test.d.ts

@@ -0,0 +1 @@
+export {};

package/dist/tests/plugins.test.d.ts

@@ -0,0 +1 @@
+export {};

package/dist/tests/rehype-absolute-links.test.d.ts

@@ -0,0 +1 @@
+export {};

package/dist/tests/rehype-sanitize-html.test.d.ts

@@ -0,0 +1 @@
+export {};

package/dist/tests/rehype-wikipedia-footnotes.test.d.ts

@@ -0,0 +1 @@
+export {};

package/dist/tests/types.test.d.ts

@@ -0,0 +1 @@
+export {};

package/dist/tests/ultimate-coverage.test.d.ts

@@ -0,0 +1 @@
+export {};

package/dist/vite.config.d.ts

@@ -0,0 +1,2 @@
+declare const _default: import('vite').UserConfig;
+export default _default;

package/dist/vitest.config.d.ts

@@ -0,0 +1,2 @@
+declare const _default: import('vite').UserConfig;
+export default _default;

package/package.json

@@ -0,0 +1,74 @@
+{
+  "name": "@aiquants/html-to-markdown",
+  "version": "0.1.0",
+  "description": "HTML to Markdown converter",
+  "type": "module",
+  "main": "./dist/index.js",
+  "module": "./dist/index.js",
+  "types": "./dist/index.d.ts",
+  "exports": {
+    ".": {
+      "import": "./dist/index.js",
+      "require": "./dist/index.cjs"
+    }
+  },
+  "bin": "./dist/main.js",
+  "files": [
+    "dist",
+    "package.json",
+    "README.md",
+    "LICENSE"
+  ],
+  "scripts": {
+    "start": "vite-node main.ts",
+    "build": "vite build",
+    "lint": "biome lint app/",
+    "format": "biome format --write",
+    "check": "biome check --fix src/",
+    "test": "vitest run",
+    "coverage": "vitest run --coverage"
+  },
+  "keywords": [
+    "html",
+    "markdown",
+    "converter"
+  ],
+  "author": {
+    "name": "fehde-k",
+    "url": "https://github.com/fehde-k"
+  },
+  "license": "MIT",
+  "dependencies": {
+    "github-slugger": "^2.0.0",
+    "hast": "^1.0.0",
+    "hast-util-to-html": "^9.0.5",
+    "playwright": "^1.54.1",
+    "rehype-parse": "^9.0.1",
+    "rehype-raw": "^7.0.0",
+    "rehype-remark": "^10.0.1",
+    "rehype-slug": "^6.0.0",
+    "remark-gfm": "^4.0.1",
+    "remark-stringify": "^11.0.0",
+    "unified": "^11.0.5",
+    "unist-util-visit": "^5.0.0",
+    "yargs-parser": "^22.0.0"
+  },
+  "devDependencies": {
+    "@biomejs/biome": "^2.1.3",
+    "@types/github-slugger": "^2.0.0",
+    "@types/hast": "^3.0.4",
+    "@types/mdast": "^4.0.4",
+    "@types/node": "^22.16.5",
+    "@types/yargs-parser": "^21.0.3",
+    "@vitest/coverage-v8": "^3.2.4",
+    "happy-dom": "^18.0.1",
+    "rehype-stringify": "^10.0.1",
+    "terser": "^5.43.1",
+    "typescript": "^5.8.3",
+    "vite": "^7.0.6",
+    "vite-node": "^3.2.4",
+    "vite-plugin-dts": "^4.5.4",
+    "vitest": "^3.2.4",
+    "vitest-tsconfig-paths": "^3.4.1"
+  }
+}