Octopress uses a non-conformant Markdown processor as standard
I have just discovered, while adding an about page to this blog, that the standard Markdown processor used by Octopress (called RDiscount) doesn’t conform to the Markdown standard. I mean in particular the reference-style links, which I happen to be quite fond of. You should be able to include
> characters around URLs like so:
[eg]: <http://www.example.com> "The canonical example page"
Instead of stripping out these optional characters RDiscount passes them on and they get translated into HTML entities (
<) which causes further steps in the chain to choke. Ultimately you end up with a very broken web page.
An [example link] [eg] turns into the following mess — oh dear.
<p>An <a href="<http://www.example.com>” title=”The canonical example page”>example link</a>.</p>
Notice how the opening quote never gets closed? The rest of that paragraph is the
href for that link! So I will have to learn to omit the angle brackets:
[eg]: http://www.example.com "The canonical example page"
PS. While writing this post I discovered that Pygments.rb (which does syntax highlighting for Octopress) doesn’t support Markdown as a highlightable language (fair enough, can’t cover everything) — but also that it crashes if you name a language it doesn’t know. Not having a good day.
``` markdown I will kill your highlighter [eg]: <http://www.example.com> "The canonical example page" ```