ack could be the only file-grep tool which could support unicode normalization.
Even the coreutil grep with multibyte support does not find "Café" in a file with "Café", the first using the decomposed "e\x301", the second using the composed "\e9" for the last small e with grave.
See e.g. http://perl11.org/blog/foldcase.html
This could be done by using fc with -i, and not just $str = "(?i)$str",
=> $str = "(?iu)$str" on perl versions, which do support case-folding with /u and unicode patterns. perldoc perlre /u.
/l could also be supported for the current local encoding.
ack could be the only file-grep tool which could support unicode normalization. Even the coreutil grep with multibyte support does not find "Café" in a file with "Café", the first using the decomposed "e\x301", the second using the composed "\e9" for the last small e with grave. See e.g. http://perl11.org/blog/foldcase.html
This could be done by using fc with -i, and not just
$str = "(?i)$str", =>$str = "(?iu)$str"on perl versions, which do support case-folding with /u and unicode patterns.perldoc perlre /u./lcould also be supported for the current local encoding.