Shell Programming and Scripting

Hello all,
I read somewher that regular expressions work with ASCII table so when i type

Code:

grep "[a-z][a-z]*" file_name

it uses values from ACII dec97(a) to dec122(z), right ?
But if I have file containing diacritics, lets say (ordinary Slovak language characters):

Code:

marek@cepi:~$ cat diakritika 
áôúéťľúľščťžýáíéôäúú
ÁôúÉŤĽÚĽŠČŤŽÝÁÍÉôäÚÚ

marek@cepi:~$ grep -o "[a-z][a-z]*" diakritika 
áôúéťľúľščť
ýáíéôäúú
ôú
ôä

Why this regexp know diacritics? And why know only lower case and not "ž" ??? This is strange for me. Friend told me it could be something with $LANG. So my $LANG is:

Code:

marek@cepi:~$ echo $LANG
en_US.UTF-8

Also I would ask if I want uppercase file with diacritic i type:

Code:

marek@cepi:~$ cat diakritika | tr "[:lower:]" "[:upper:]"
áôúéťľúľščťžýáíéôäúú
ÁôúÉŤĽÚĽŠČŤŽÝÁÍÉôäÚÚ

why it not change lower to upper ?
Thanks a lot for reply
PS: I hope that characters display properly

maybe i know why is "ž" different it is "behind" z so regex [a-z] did not match the "ž" but still many thinngs are unclear

If you want uppercase as well you have to specify

Code:

grep "[a-zA-Z][a-zA-Z]*" file_name

What you are discusiing is called a collating sequence. Do a web search for "POSIX collating sequence" for further information.

To be language-neutral, your example would be written as

Code:

grep "[[:alpha:]][[:alpha:]]*"  filename

or if you only want lowercase characters

Code:

grep "[[:lower:]][[:lower:]]*"  filename

ž comes after z, so it is not in the range you gave.

Probably because those characters are not part of the en_US.UTF-8 definition of [:lower:] and [:upper:].

Thank you for reply. Is there an option how to convert lowercase diacritics to uppercase ?