bug#23752: [PATCH] grep: try fgrep matcher for case insensitive matching by grep -F in multibyte locale

[Norihiro Tanaka]

On Thu, 1 Sep 2016 09:50:11 -0700
Paul Eggert <address@hidden> wrote:

> Suppose all the multibyte characters in the pattern are non-letters, so that 
> case-folding does not affect them. Could grep -iF be fast in that case?

I prefer DFA matcher to KWset matcher due to low memory.  grep -F is
offen used for big patterns.  DFA has a risk which memory is exhausted
for them.  I thank that so as far as possible grep -F should not use DFA
matcher.

> Is the problem that some encodings allow two different representations for 
> the same character, and we want the pattern to match both representations?

KWset does not support multibyte character, so we must exclude cases
that upper case of single byle character is multi-byte character.