I simplified the structure of Corpus.
Overhaul
Structure
So, according to the discussion here, Corpus has an issue of over-differentiation. The solution proposed there is adopted here, which is to only keep the top-level categories, i.e., the six layers, and implement sub-categories through tags.
corpus.zsh and other scripts are completely rewritten, removing all the logic for handling sub-categories.
A new --insta flag is added for quickly creating and opening a headless nvim instance for editing.
Template
Template system has been modified here, changing the English and Latin prompts to Traditional Chinese to enhance (for my) readability (embarrassed). However, after the hierarchy merge, a single template is inevitably not suitable for all entries under a hierarchy, so the prompt content in the template is simply removed, leaving only the YAML header, and all content is edited by users themselves.
Let's see how it works in practice.