"Smart" case fold?

[AmaiKinono]

In isearch, the default behavior is do case insensitive search when the user input is all lower case, and do case sensitive search otherwise. This is neat.

I once thought this behavior is controled by case-fold-search, but then I found it controls case sensitivity directly. If there's no built-in mechanism for this, we may need to invent our own.

[raxod502]

That seems fine as an optional configuration. We can just use isearch-no-upper-case-p; it's not going anywhere.

[raxod502]

Case folding is now guessed based on the current search query by default, and you can toggle it persistently (for the search session) using M-s c, the same binding as in Isearch.

[AmaiKinono]

This works well! But I don't know if the keybinding is the best. M-s calls next-matching-history-item in minibuffer by default, so I have to bind ctrlf-toggle-case-fold-search to another key to use it.

[raxod502]

I noticed that. Any suggestions on a better binding? I was unable to find anything standard that wasn't a bad idea in one way or another.

[AmaiKinono]

How about M-c? That seems to do the same thing in isearch.

[raxod502]

But as I mention in #29, that conflicts with capitalize-dwim, which to me is even worse than conflicting with next-matching-history-item as it is a better-known command.

[AmaiKinono]

Ah yes... I think bind M-s to nil is not a bad idea. Or if we have to invent it, I think we could have a prefix key (like C-o in ivy. I personally have M-l (means "local") in various temporary state / special major mode to execute their commands), and put c under it to toggle case fold. In the future we may have other switches, and commands like "recursive edit" during replacing, we could put those under that prefix too. Actually I think it's a good idea to also have search style switches under it.

[raxod502]

M-l is bad because it conflicts with downcase-dwim, but C-o is a fair choice because wanting to use open-line in the middle of a search would be an odd use case. Changed.