buffer-search-forward case sensitive by default?

[italomaia-bk]

Hi. Just noticed buffer-search-forward is case sensitive by default. I think that makes it harder to search a buffer, given the most common case is to do a case-insensitive search then fall back to case-sensitive if too many results appear. Could we have case-insensitive search by default?

[refi64]

I personally like it as-is, but a way to make it insensitive (e.g. ctrl-I) would actually be really nice.

[italomaia-bk]

I think a case-insensitive is the most common case, but I might be wrong. If the shortcuts give priority to behave in the most common useful ways, that would probably be a good thing. More opinions?

[Janfel]

VSCode has a neat feature where the search is case-insensitive as long as there is no uppercase character in the query. I propose a config value for this, that can be set to true, false or "auto", defaulting to "auto", because this is more in line with Howls "Features enabled by default" philosophy.

[shalabhc]

Emacs does case insensitive also until you type an uppercase letter (IIRC, it's been a while)

There's still the corner case of case sensitive matching for all lowercase text. If we do this, I prefer the toggle shortcut (e.g. pressing ctrl_i while the command is active). We have toggle features for some other commands such as open where ctrl_s will switch to recursive directory search.

I think buffer-search-forward has limited applicability because better features exist in Howl (probably just for me :smile: ). I rarely use it, but I used it a lot more in other editors. More often I use buffer-grep which does not only case insensitive but also somewhat of fuzzy matching and does the entire buffer in one shot. The second most common search I use is buffer-search-forward-word - position the cursor at a word and type ctrl_period and voila - it finds the next instance of exactly the word I was on. This is the typical 'find me the next usage of this identifier'.