Open ee7 opened 2 years ago
FYI The additional rendering options that we support are footnotes and allowing unsafe HTML: https://github.com/gjtorikian/commonmarker#render-options
Thanks. Do we support tables and strikethrough too?
website/app/commands/markdown/parse_doc.rb#L10
website/test/commands/markdown/parse_test.rb#L70-L94
website/test/commands/markdown/parse_test.rb#L144-L146
What's the rationale behind using UNSAFE
? And how does it interact with tagfilter
?
To others, links about footnotes:
Do we support tables and strikethrough too?
Yes.
What's the rationale behind using UNSAFE? And how does it interact with tagfilter?
@iHiD?
No idea. I guess we trust maintainers' code and allow them to do whatever?
🤷
I'd probably lean towards:
<iframe>
and <script>
?markdownlint
workflow in org-wide-files
(tracked in https://github.com/exercism/org-wide-files/issues/214) eventually anyway.markdownlint
error could be disabled on a repo-by-repo basis, with its configuration requiring @exercism/maintainers-admin
approval.
I believe that the website uses
commonmarker
(see here), which is the wrapper forlibcmark-gfm
. And therefore the website supports rendering a markdown file that uses GitHub Flavored Markdown features.However, the Exercism Markdown spec does not state whether a Markdown file can use GFM features. We should state that explicitly, so that e.g. a track knows whether they can use a tables, task list items, or strikethroughs.
Configlet will soon learn to parse and render Markdown, but this will at least initially only support the CommonMark 0.30 spec (not GFM). If necessary, it should be straightforward for it to support GFM (by wrapping
libcmark-gfm
, rather thanlibcmark
).Somewhat related: GitHub is thinking about adding a new admonition syntax. See https://talk.commonmark.org/t/4173
But even if the GFM spec is changed, I would suggest that we keep the current admonition syntax: