Re: Character folding in the pretest

[Paul Eggert]

On 02/04/2016 08:54 AM, Eli Zaretskii wrote:
> Emacs is a multilingual environment, so any assumption that
> the main language in every buffer, or even in most buffers, is likely
> to be the locale's language will misfire.

True, but although Emacs is designed to be language-agnostic when 
handling buffer text, that doesn't mean it should be designed to be 
language-agnostic when handling user input. If Emacs starts up in a 
language-X locale, its user probably will be more comfortable using 
language-X rules for searching, even if the main language in a buffer is 
language Y. As an English-speaker when I search Swedish texts by hand, I 
normally want to use English-like rules because English is what I know 
and I can't really read the Swedish anyway. In English we tend to 
consider accents unimportant when searching, and because we treat 
“naïve” like “naive” we also treat “Ångström” like “Angstrom” even 
though the latter is not correct in Swedish.