Re: commit-msg hook

[Paul Eggert]

On 04/13/2015 08:48 AM, Eli Zaretskii wrote:
> why do we need to rely on system
> libraries to implement UTF-8 and [:print:] correctly?
Some POSIXish environments do not support unibyte locales; even the C 
locale is multibyte (it uses UTF-8).  In such environments, I expect 
that a regular expression like /\342\202\254/ won't necessarily match 
the string "€" (U+20AC, equivalent to the byte string "\342\202\254") in 
any locale, just as in Emacs the call (re-search-forward "\342\202\254") 
won't find the string "€" in a UTF-8 file, regardless of language settings.

I suppose the script could detect whether we're in such an environment, 
but I dunno, it sounds like more trouble than it's worth.