Movement by subwords

[JeftavanderHorst]

This PR adds commands that enable movement within subwords that are snake_cased or camelCased, as previously proposed in discussion #5463. Subwords are delimited by underscores and by transitions from lowercase to uppercase.

The commands are named {move,extend}_{next,prev}_sub_word_{start,end}. This PR doesn't add any keybinds.

Example usage (| characters indicate cursor):

• |f|oo bar baz -> move next sub word -> |foo |bar baz
• |f|oo_bar_baz -> move next sub word -> |foo_|bar_baz
• |f|ooBarBaz -> move next sub word -> |foo|BarBaz

Example keybinds:

[keys.normal]
W = 'move_next_sub_word_start'
B = 'move_prev_sub_word_start'
E = 'move_next_sub_word_end'

[keys.select]
W = 'extend_next_sub_word_start'
B = 'extend_prev_sub_word_start'
E = 'extend_next_sub_word_end'

[JeftavanderHorst]

I notice many PRs have been marked with S-waiting-on-review but this one hasn't, is there anything I need to do to request a review?

[shaleh]

Can confirm this works as advertised. Thanks, I was literally making the same change and decided to look for a PR first.

[shaleh]

My only concern is the widespread use of "char_is_word". I wonder in what other ways the definition of "word" is not matching "subword". Many of the places there is little context, just a solitary character.

Initially I thought Grapheme::is_word_boundary might be a good avenue but that is only used in doc formatting for some reason....

[JeftavanderHorst]

It's been a while since I implemented this, so I can't remember whether I looked at char_is_word. There is at least one other definition of "word" that I didn't update, which is in text objects. Meaning miW is still Match inner WORD, instead of Match inner sub word. Sub word text objects are implemented in #8658 though, if you need them.

That being said, I've been using this branch in my daily driver since I created it, and so far everything works as expected, so I don't think char_is_word is too important. If the core devs are interested I can work on this as well, but for now, I tried to keep the diff as small as possible (except for the tests) so that I can easily keep up with master.

[pascalkuthe]

sorry it took a while to get to this. I think this is a pretty good implementation and I think this feature does carry its weight. I don't think we need to use subword in other places, char_is_word is to differentiate whitespace and special chars and usually have nothing to do with word segmentation.

[JeftavanderHorst]

Friendly ping to @the-mikedavis to request a review, since this PR already got one approval and is relatively small (ignoring the tests)