cxa/readable.js

## readable.js
#!/usr/bin/env node

// npm i -g @mozilla/readability puppeteer

const puppeteer = require("puppeteer");
const fs = require("fs");
const readabilityJsStr = fs.readFileSync(
  require.resolve("@mozilla/readability/Readability.js"),
  { encoding: "utf-8" }
);

function executor() {
  return new Readability({}, document).parse();
}

const dump = async (url) => {
  const browser = await puppeteer.launch();
  const page = await browser.newPage();
  await page.goto(url);
  const resultArticle = await page.evaluate(`
    (function(){
      ${readabilityJsStr}
      ${executor}
      return executor();
    }())
  `);

  const content = `
<!doctype html>
<html>
<head>
  <title>${resultArticle.title}</title>
</head>
<body>
  ${resultArticle.content}
</body>
</html>`;
  browser.close();
  return content;
};

const main = async () => {
  try {
    const html = await dump(process.argv[2]);
    console.log(html);
    process.exit(0);
  } catch (e) {
    console.error(e);
    process.exit(1);
  }
};

main();
	#!/usr/bin/env node

	// npm i -g @mozilla/readability puppeteer

	const puppeteer = require("puppeteer");
	const fs = require("fs");
	const readabilityJsStr = fs.readFileSync(
	require.resolve("@mozilla/readability/Readability.js"),
	{ encoding: "utf-8" }
	);

	function executor() {
	return new Readability({}, document).parse();
	}

	const dump = async (url) => {
	const browser = await puppeteer.launch();
	const page = await browser.newPage();
	await page.goto(url);
	const resultArticle = await page.evaluate(`
	(function(){
	${readabilityJsStr}
	${executor}
	return executor();
	}())
	`);

	const content = `
	<!doctype html>
	<html>
	<head>
	<title>${resultArticle.title}</title>
	</head>
	<body>
	${resultArticle.content}
	</body>
	</html>`;
	browser.close();
	return content;
	};

	const main = async () => {
	try {
	const html = await dump(process.argv[2]);
	console.log(html);
	process.exit(0);
	} catch (e) {
	console.error(e);
	process.exit(1);
	}
	};

	main();