Live Coding 5/24/2017 - recovering blog posts with Cheerio

save-mah-blog.ts

import * as fs from 'fs';
import * as path from 'path';

import * as glob from 'glob';
import * as cheerio from 'cheerio';

const files = glob.sync(path.join(__dirname, '**/*.html'));
const filesAndContent = files.reduce((acc, x) => {
  let $ = cheerio.load(fs.readFileSync(x, 'utf8'));

  let content = $('main .post-content');
  if (content.length < 1) return acc;

  let date = $('time.post-date').attr('datetime');
  let title = $('article h1.post-title').text();

  acc[x] = {
    content: content.html(),
    date, title
  };

  return acc;
}, {});

Object.keys(filesAndContent).forEach(k => {
  fs.writeFileSync(k.replace(/\.html$/i, '.json'), JSON.stringify(filesAndContent[k]));
  fs.writeFileSync(k.replace(/\.html$/i, '-content.html'), filesAndContent[k].content);
});

debugger;