bug#21604: grep doesn't match diacritical chars in ISO-8859 files

[Paul Eggert]

On 10/02/2015 02:43 AM, Santiago Ruano Rincón wrote:
> grep doesn't match characters with diacritical
> marks in ISO-8859 files, inside a Unicode enviroment

That is normal and expected behavior.  In a UTF-8 locale, "á" is 
represented by the two bytes 0xC3 and 0xA1.  In an ISO-8859 file, the 
same character is represented by the single byte 0xE1.  The UTF-8 
pattern won't match the ISO-8859 representation.

To avoid this problem, switch to an ISO-8859 locale before using grep to 
read ISO-8859 text files.  This is true for pretty much any standard 
utility, not just grep.  Alternatively, you can translate the text files 
from ISO-8859 to UTF-8, before giving the resulting text to grep or to 
other utilities.