Automatically converting DOC to MD

change-docx.md

First convert DOC to DOCX using LibreOffice:

/Applications/LibreOffice.app/Contents/MacOS/soffice --invisible --convert-to docx file.doc

Then convert to Markdown using pandoc:

pandoc file.docx -f docx -t markdown -o file.md

Convert to Markdown, extracting the image files:

pandoc file.docx --extract-media=/path/to/folder -f docx -t markdown -o file.md

Automatically convert a folder of docx files, extracting the images to named folders with the convert.js

convert.js

var fs = require('fs')
var path = require('path')
var mkdirp = require('mkdirp')
var exec = require('child_process').exec
var BagPipe = require('bagpipe')
var bag = new BagPipe(5)

// '/Users/saibotsivad/Development/thinking/kayser-commentary'
var input_folder = process.argv[2]
// '/Users/saibotsivad/Development/thinking/kayser-commentary-markdown'
var output_folder = input_folder + '-markdown'

var temp_folder = '/tmp/markdown_conversion'

processDirectory(input_folder, '.')

function processDirectory(parentPath, childPath) {
	fs.readdir(path.join(parentPath, childPath), function(err, files) {
		if (err) {
			console.log(err)
		} else {
			files.forEach(checkIfFile.bind(null, parentPath, childPath))
		}
	})
}

function checkIfFile(parentPath, childPath, file) {
	fs.stat(path.join(parentPath, childPath, file), function(err, stats) {
		if (stats.isDirectory()) {
			processDirectory(parentPath, path.join(childPath, file))
		} else {
			beginConversion(parentPath, childPath, file)
		}
	})
}

function uuidv4() {
	return 'xxxxxxxxxxxx4xxxyxxxxxxxxxxxxxxx'.replace(/[xy]/g, function(c) {
		var r = Math.random() * 16 | 0, v = c == 'x' ? r : (r & 0x3 | 0x8)
		return v.toString(16)
	})
}

function beginConversion(parentPath, childPath, file) {
	if (/\.doc$/i.test(file)) {
		bag.push(convertDocToDocxThenMarkdown, parentPath, childPath, file, function(){
			console.log('Success converting to DOCX: ' + file)
		})
	} else if (/\.docx$/i.test(file)) {
		convertDocxToMarkdown(parentPath, childPath, file, path.join(parentPath, childPath))
	}
}

function regexFilename(file) {
	var regex = /(.+)\.[^.]+$/.exec(file)
	if (!regex) {
		console.log('Error regexing filename', file)
	} else {
		return regex[1]
	}
}

function convertDocToDocxThenMarkdown(parentPath, childPath, file, cb) {
	console.log('Starting LibreOffice conversion...', path.join(childPath, file))
	var outputPath = path.join(temp_folder, uuidv4())
	exec('/Applications/LibreOffice.app/Contents/MacOS/soffice --invisible --convert-to docx "' + path.join(parentPath, childPath, file) + '" --outdir "' + outputPath + '"', {
		cwd: input_folder
	}, function(err, stdout, stderr) {
		if (err) {
			console.log('Error converting to DOCX', err, stderr)
		} else {
			console.log(stderr, stdout)
			cb()
			convertDocxToMarkdown(parentPath, childPath, regexFilename(file) + '.docx', outputPath)
		}
	})
}

function convertDocxToMarkdown(parentPath, childPath, filenameWithExtension, inputFilePath) {
	var filenameWithoutExtension = regexFilename(filenameWithExtension)
	var childPathWithFile = path.join(childPath, filenameWithoutExtension)

	mkdirp(path.join(output_folder, childPath), function() {

		exec('/usr/local/bin/pandoc "' + filenameWithExtension + '" --extract-media="./' + filenameWithoutExtension + '" -f docx -t markdown -o "' + path.join(output_folder, childPath, filenameWithoutExtension) + '.md"', {
			cwd: inputFilePath
		}, function(err, stdout, stderr) {
			if (err) {
				console.log('Error converting to markdown', err, stderr)
			} else {
				console.log('Success converting to markdown: ' + childPathWithFile)
				fs.exists(path.join(inputFilePath, filenameWithoutExtension), function(exists) {
					if (exists) {
						console.log('Moving image directory:', childPathWithFile)
						fs.rename(path.join(inputFilePath, filenameWithoutExtension), path.join(output_folder, childPath, filenameWithoutExtension), function(err) {
							if (err) {
								console.log('Error moving image directory:', childPathWithFile)
							} else {
								console.log('Success moving image directory:', childPathWithFile)
							}
						})
					} else {
						console.log('No image directory found:', childPathWithFile)
					}
				})
			}
		})

	})
}

convert.sh

#!/bin/bash

/usr/local/bin/node /Users/saibotsivad/Development/thinking/kayser-commentary/convert.js "$1"