Python script for style quality control

I’ve written a Python script to perform some checks that can’t easily be
performed via RELAX NG schema validation:

It helped uncover quite a few problems with styles in the style repository ( Most these errors are now fixed (I just
didn’t replace the underscores with hyphens for
mhra_note_without_bibliography.csl because it ships with Zotero). In
addition, the script can report the number of dependent styles per
independent style:

I’ll run the script manually every now and then, although I guess something
like this could be combined with the schema validation that the Zotero Style
Repository already performs.