Re: title-case function

["Paul W. Rankin" <hello@paulwrankin.com>]

On Sun, Apr 21 2019, Emanuel Berg wrote:
> Did you look into this [2] ?
>     
> BTW make a search for "Emanuel Berg" :)

Ah right, no I didn't find that one. Nice.

Given that there are no definitive rules regarding title case minor 
words, I think I prefer using a defcustom for these (i.e. I wouldn't 
consider "over" or "past" as a minor words because they're four 
letters).

I've found thing-at-point to be quite resource hungry so I try to avoid 
using it unless absolutely necessary.

I am curious about the comparative performance of using (member WORDS) 
vs (looking-at (regexp-opt WORDS)).

>        (if (looking-at "[:\x2013\x2014]")
>            (capitalize-word 1)
>          (skip-syntax-forward "-." last-word)
>          (if (looking-at (concat "\\b" (regexp-opt 
>          title-case-minor-words)
>                                  "\\b") )
>              (downcase-word 1)
>            (capitalize-word 1)))
>
> It seems to work alright, but I don't
> understand it? You don't need `progn' anymore?

The regexp "[:\x2013\x2014]" just looks for a colon, an en dash or an em 
dash and makes sure to capitalize a word following one of those. I used 
the character codes for the dashes because otherwise they're difficult 
to discern from hyphens in monospace.

Do you mean using progn for the ELSE in the if expression? From the 
documentation on `if':

> (if COND THEN ELSE...)
>
> THEN must be one expression, but ELSE... can be zero or more 
> expressions.

So yeah, progn is unnecessary.

Thanks for the trance, but I'm already spending my rainy Easter day 
listening to this: 
https://www.discogs.com/Various-Dark-Was-The-Night/release/1655481

-- 
https://www.paulwrankin.com