The script, "style-metadata-export.py", uses lxml for XML parsing
(http://lxml.de/), and outputs to "style-metadata.json". You can see
the results in "style-metadata.html". I committed a small sample JSON
file so you can play with the table without running the script (the
JSON file for the entire style repo is 1.1 MB). The current webpage
loads for me in Safari and Firefox, but not in Chrome due to some
permission issue with loading the locale JSON file via AJAX.
I already found the first errors to fix after a brief look through the
data: the "Durchstarten" styles all have a trailing double quote mark
in the title (https://zotero.org/styles?q=durchstarten), and
"Universität Bremen - Institut für Politikwissenschaft" doesn't
specify the style's default-locale in the title, even though it's set
to "de-DE" (https://zotero.org/styles?q=bremen). So I imagine it can
help us maintain the styles repo.
DataTables seems quite snappy, and it allows for server-side
processing too. Dan, perhaps it could be used to speed up
https://zotero.org/styles as well?