When passing a filename that contains spaces, docx2gfm shows a warning and creates an empty markdown file.

For example

$ docx2gfm -f my_\ file\ with\ spaces.docx
pandoc: my_: openBinaryFile: does not exist (No such file or directory)

---
layout: post
title: "YOUR POST TITLE"
comments: true
categories: [tag1, tag2, multi word tag]
author: <a href="link to twitter, personal blog, linkedin, etc">YOUR NAME</a>
image: "/images/own/post_directory/logo_image_NOT_SVG_FORMAT.png"
---

This can likely be fixed by making sure that the filename is quoted correctly when passing it along to pandoc.

pandoc comes with an option --extract-media=DIR that allows to extract the images from the docx to a local folder. Could this be used to make it easier to create the final markdown with the images in the correct folder?

Some markdown special characters get escaped in the step convert_to_reference_style_links(content).

For example:
Goodbye\!

Currently we are performing a manual cleanup of those but that is quite error-prone.

I filed an issue with commonmarker to see if this behavior is actually correct:
gjtorikian/commonmarker#91

As a user I want to use this tool without cloning the repo. Publishing this tool as a gem (on https://rubygems.org/) would achieve this.

However publishing this tool as a gem is pretty much the same as open sourcing it, as the source inside a gem is publicly available. Hence I should first make it public, then gemify it.

I think there might be a bug in the link extraction when the link contains a closing bracket at the end.
e.g. https://en.wikipedia.org/wiki/Beirut_(disambiguation)

The link extraction that we do seem so swallow the last ).
Need to reproduce this in a test.

See assets/sample.docx

When running ruby docx2gfm.rb assets/sample.docx the following happens:

• the title of the doc is removed (this happens for any docx I think)
• the first paragraph and image is removed (this only happens for some docx files)

Is this a bug in pandoc?

Instead of downloading the .docx and then converting it, would be great to download the online gdoc and convert it directly. Saves steps, and hence time.

https://developers.google.com/drive/api/v3/reference/files

How to give the tool permission to read gDrive files?
https://www.krautcomputing.com/blog/2013/12/17/how-to-access-your-google-drive-files-with-ruby/

Can I move the cleanup steps that are regexps into pandoc itself?
See function cleanup_content(content).

This would assume that pandoc is making mistakes in the docx to gfm conversation.

docx2gfm currently generates a jekyll YAML front matter and puts it at the very beginning of the markdown output.

We could create an option that allows to toggle this behavior on/off for people that really just need gfm and not jekyll output? That behavior also seems more consistent with the name of this tool, as it is docx2gmf and not docx2jekyll :)

Pandoc adds a funny marker sometimes.
<!-- end list -->

We are removing it manually right now, but then it gets re-introduced by commonmarker. So I suspect it is part of the markdown specs?

No tests so far. Yeah :)

Things to test:

• the extra cleanup that I am performing to the pandoc output
• behavior of the CLI options

The sample document should test all possibles formatting options, available in GFM:
https://guides.github.com/features/mastering-markdown/

I need to test:

• tables
• quotes
• code highlighting
• task lists
• ...

Single quotation marks are escaped with a backslash.

i.e. the text Let's link turns into Let\'s link.

This is not really a problem, as the backslash should not get rendered by markdown, as it just escapes the next character. I don't know why single quotation marks need escaping.

Underlining of text is not even an option in markdown, is it? Funny that I never caught that before.
https://help.github.com/articles/basic-writing-and-formatting-syntax/ /

https://softwareengineering.stackexchange.com/questions/207727/why-there-is-no-markdown-for-underline

So maybe we should make docx2gfm print a WARNING when it finds underlined text?