With the release of pandoc 1.8, it occurs to me that a super-simple
way to get a basic static repo running could be as simple as a little
script that take two arguments: csldir (for the git repo) and outdir
(where to dump the output), and then …
- updates the git styles (git pull)
- checks out the style files (and runs the output filter to add
updated date-time stamp) and copies them to outdir - runs pandoc on a standard markdown file and a custom template to
add the style metadata and preview to the HTML output
So at this point we just have a big directory like:
apa.csl
apa.html
chicago.csl
chicago.html
We then need …
- an .htaccess file to do simple content-negotiation (to serve styles
as text/x-csl) - generate the index as a single file, using a sortable and
filterable table (and maybe alternately as a json file?)
The result should be simple, low-resource, and fast.
We could then later replace it with a dynamic app (using, say,
citeproc-node for previewing) if we need more features or scale?
Bruce