Re: Generating a listing of all symbols (16K+) and labeling subsets

[Hans BKK <hansbkk@gmail.com>]

On Friday, April 18, 2014 3:19:40 AM UTC-4, Thien-Thi Nguyen wrote:
>    I'd specifically like to ask about the logic of the
>    categorizing breakdown and for better taxonomy terms.
> 
> 
> 
> It is a category error to categorize Emacs Lisp symbols into
> 
> (single) buckets, unless you wish to defer understanding.  :-D

I realize there is overlap, hence the duplication of tests for primitive vs lisp etc within multiple buckets. And within a given cond level, order is by (my judgement) of priority - commands first, ID as a functions before values, user options before setq ones, etc.

This logic flow is exactly what I was looking for feedback on, as well as how to identify more specifically those currently falling into the "t" or "other" buckets - currently over half the symbols are neither functions nor variables nor faces. Too bad apparently predicates don't exist to cover all possible symbol types.

Once I have the "skeleton" to my satisfaction, I plan to output the relevant docstrings as well, so when I'm looking at a package's "diff report" everything's in one place.