The discussion revolves around the effectiveness of Discourse’s Google Docs to Markdown conversion feature. mbbroberg appreciates the smooth transition and automatic table formatting, wondering about the underlying code. codinghorror credits @vinothkannans for the hard work, who reveals that a custom HTML to markdown JS module was created in /app/assets/javascripts/discourse/app/lib/to-markdown.js.
Other users, such as travelvc, share their positive experiences with the feature. qaisjp provides a tangential tip on exporting Google Docs as .docx for better conversion results using pandoc, but also plans to try the ‘Discourse converter’ in the future.
I frequently need to convert Google Docs to Markdown for work. Instead of installing a slightly-sketchy extension or downloading at .odt then running pandoc to convert to .md, I frequently copy/paste right into a Discourse discussion and viola.
Every single time it’s the smoothest Markdown transition. The same happens for tables when I copy table-structured information and it automatically builds a Markdown table around it.
How does it do that?
I started poking around the source code but the JS libraries are a bit out of my skillset. I ask because if it’s something I could use elsewhere in open source (first up: Joplin).
Any tips or advice on walking the code is appreciated.
Slightly tangential to the rest of your post… I read somewhere that you can get better results if you export as .docx and then pandoc that to Markdown. And, for me, it’s worked fine (but I haven’t tried exporting as .odt and converting that to Markdown).
But the next time I need to convert from Google Docs to Markdown, I’ll definitely give the ‘Discourse converter’ a go
