ntotten/html_convert.js

## html_convert.js
var lsr = require('lsr');
var path = require('path');
var fs = require('fs');
var toMarkdown = require('to-markdown');
var striptags = require('striptags');
var mkdirp = require('mkdirp');
var matter = require('gray-matter');
var toTitleCase = require('titlecase');
var beautifyHtml = require('js-beautify').html;

var files = [];

const outputFolder = 'output'

function tryGetTitle(content) {
  var regex = /\#{1}[^\n\#]+/g;
  var match = content.match(regex);

  if (match && match.length) match = match[0].slice(1).trim();

  return match;
}

var deleteFolderRecursive = function(path) {
  if( fs.existsSync(path) ) {
    fs.readdirSync(path).forEach(function(file,index){
      var curPath = path + "/" + file;
      if(fs.lstatSync(curPath).isDirectory()) { // recurse
        deleteFolderRecursive(curPath);
      } else { // delete file
        fs.unlinkSync(curPath);
      }
    });
    fs.rmdirSync(path);
  }
};

deleteFolderRecursive(path.join(__dirname, outputFolder));

function loadPath(loadPath) {
  var resolvedPath = path.resolve(__dirname, loadPath);

  lsr
    .sync(resolvedPath)
    .forEach(function(fileStat) {
      var filename = fileStat.path;
      var filepath = path.resolve(resolvedPath, filename);

      if (/\.c\.html$/.test(filename)) {
        return;
      }

      if (!/\.html$/.test(filename)) return;

      var text = fs.readFileSync(filepath, 'utf8');
      files.push({
        filename: path.join('/', loadPath, filename),
        text: text
      });
    });
}

[
  './www.digitechbranding.com/pages',
  './www.digitechbranding.com/instructions',
  './www.dynasend.com/pages'
].map(loadPath);

files.forEach(function(file) {

  var x = file.text.indexOf('<!--START PAGE CONTENT-->');
  var y = file.text.indexOf('<!--END PAGE CONTENT-->');
  var z = 25;

  if (x < 0 || y < 0) {
    x = file.text.indexOf('<body>');
    y = file.text.indexOf('</body>');
    z = 6;
  }

  var text;
  if (x > -1 && y > -1) {
    text = file.text.substring(x + z, y).trim();
  } else {
    text = file.text;
  }

  if (!text) return;

  var md = striptags(toMarkdown(text));
  md = md.replace(/\n\s*\n\s*\n/g, '\n\n'); // strip extra line breaks
  md = md.replace(/(\d)\\\./g, '$1.'); // fix lists

  var title = tryGetTitle(md);
  if (!title) {
    title = path.basename(file.filename, '.html').replace(/\-/g, ' ');
  }
  title = toTitleCase(title);

  var parts = file.filename.split('/');
  parts.shift();
  parts[0] = '';
  var redirectPath = parts.join('/');

  var options = {
    layout: 'page',
    title: title,
    redirect_from: [
      redirectPath
    ]
  }

  var outputPath = path.join(__dirname, outputFolder, file.filename.toLowerCase());

  mkdirp.sync(path.dirname(outputPath));


  var html = beautifyHtml(text);
  html = html.replace(/\<p\>\&nbsp\;\<\/p\>/g, ''); // strip empty paragraphs
  html = html.replace(/\n\s*\n\s*\n/g, '\n\n'); // strip extra line breaks
  fs.writeFileSync(outputPath, matter.stringify(html, options));
  if (md) {
    fs.writeFileSync(outputPath.replace('.html', '.md'), matter.stringify(md, options));
  }
});
	var lsr = require('lsr');
	var path = require('path');
	var fs = require('fs');
	var toMarkdown = require('to-markdown');
	var striptags = require('striptags');
	var mkdirp = require('mkdirp');
	var matter = require('gray-matter');
	var toTitleCase = require('titlecase');
	var beautifyHtml = require('js-beautify').html;

	var files = [];

	const outputFolder = 'output'

	function tryGetTitle(content) {
	var regex = /\#{1}[^\n\#]+/g;
	var match = content.match(regex);

	if (match && match.length) match = match[0].slice(1).trim();

	return match;
	}

	var deleteFolderRecursive = function(path) {
	if( fs.existsSync(path) ) {
	fs.readdirSync(path).forEach(function(file,index){
	var curPath = path + "/" + file;
	if(fs.lstatSync(curPath).isDirectory()) { // recurse
	deleteFolderRecursive(curPath);
	} else { // delete file
	fs.unlinkSync(curPath);
	}
	});
	fs.rmdirSync(path);
	}
	};

	deleteFolderRecursive(path.join(__dirname, outputFolder));

	function loadPath(loadPath) {
	var resolvedPath = path.resolve(__dirname, loadPath);

	lsr
	.sync(resolvedPath)
	.forEach(function(fileStat) {
	var filename = fileStat.path;
	var filepath = path.resolve(resolvedPath, filename);

	if (/\.c\.html$/.test(filename)) {
	return;
	}

	if (!/\.html$/.test(filename)) return;

	var text = fs.readFileSync(filepath, 'utf8');
	files.push({
	filename: path.join('/', loadPath, filename),
	text: text
	});
	});
	}

	[
	'./www.digitechbranding.com/pages',
	'./www.digitechbranding.com/instructions',
	'./www.dynasend.com/pages'
	].map(loadPath);

	files.forEach(function(file) {

	var x = file.text.indexOf('<!--START PAGE CONTENT-->');
	var y = file.text.indexOf('<!--END PAGE CONTENT-->');
	var z = 25;

	if (x < 0 \|\| y < 0) {
	x = file.text.indexOf('<body>');
	y = file.text.indexOf('</body>');
	z = 6;
	}

	var text;
	if (x > -1 && y > -1) {
	text = file.text.substring(x + z, y).trim();
	} else {
	text = file.text;
	}

	if (!text) return;

	var md = striptags(toMarkdown(text));
	md = md.replace(/\n\s\n\s\n/g, '\n\n'); // strip extra line breaks
	md = md.replace(/(\d)\\\./g, '$1.'); // fix lists

	var title = tryGetTitle(md);
	if (!title) {
	title = path.basename(file.filename, '.html').replace(/\-/g, ' ');
	}
	title = toTitleCase(title);

	var parts = file.filename.split('/');
	parts.shift();
	parts[0] = '';
	var redirectPath = parts.join('/');

	var options = {
	layout: 'page',
	title: title,
	redirect_from: [
	redirectPath
	]
	}

	var outputPath = path.join(__dirname, outputFolder, file.filename.toLowerCase());

	mkdirp.sync(path.dirname(outputPath));


	var html = beautifyHtml(text);
	html = html.replace(/\<p\>\&nbsp\;\<\/p\>/g, ''); // strip empty paragraphs
	html = html.replace(/\n\s\n\s\n/g, '\n\n'); // strip extra line breaks
	fs.writeFileSync(outputPath, matter.stringify(html, options));
	if (md) {
	fs.writeFileSync(outputPath.replace('.html', '.md'), matter.stringify(md, options));
	}
	});