Repository navigation
Markdown parsing breaks internal README links #3
Description
Steps to reproduce: try clicking any of these links; the actual anchors in the generated HTML document served by quickdocs are named sequentially, but this renaming does not occur in the link anchors.
Following from this invocation, it appears that this call would rewrite the anchor IDs if suppressAutoLink were false, which is default value; I am tempted to open a PR simply deleting that key from the function call, although I do not currently have quickdocs setup locally for testing this.
Activity
fukamachi commentedon Oct 27, 2022
The README seems to be originally written in org-mode.
Currently, it is rendered in the frontend by org-js.
adlai commentedon Oct 27, 2022
After some sandbox testing of only the parsing:
The fix is probably not so simple, because the internal links do not get recognized by this regex, due to not being complete URIs (they are only anchors). I suspect the proper way to fix this would be mostly within org-js, rather than here.
adlai commentedon Dec 16, 2022
Would it be sane of me, to move the rendering to dist-extractor? It is my inclination, although only due to an aversion towards further work in Javascript.
fukamachi commentedon Dec 21, 2022
It is a possible way. In fact, that is how it was done in Quickdocs before the rewrite.
I'm still not sure whether it should be done in dist-extractor or dist-updater, though.
adlai commentedon Dec 26, 2022
I considered asking your opinion first... I have skimmed both projects now, and my impression is that it could be done in either one, and almost certainly the change should only affect one of these two projects. I'm not certain why I considered it obvious when writing my previous comment.
I skimmed the archived code (thank you for leaving this online!), and it appears that the rendering was originally done in the updater component, using the external tool pandoc. Is reintroducing the dependency acceptable?