grep/echo with accented/special characters like é , å , ă, ß, ü

[80aless]

Hi,
I have a script
1)Taking the titles from .cue files with grep. When a title contains a special character such as "é , å , ă, ü" or others like ß, grep cannot encode it correctly and I see this symbol: �
G�recki - String Quartet No. 1
and not
Górecki - String Quartet No. 1
2)Renaming each mp3 file I have with the title using echo. As a result I get bad titles.
How do I solve this?
Thanks!
alex

[aram535]

There are a million ways of doing this, in perl its rather easy. If you have to use shell script, you can use awk/sed/tr, you can even use iconv if the character set is always the same.

Something as simple as:

# echo fóó | tr 'ó' 'o'
foo

would work

[80aless]

Thanks for you answer, but is there a way to keep those characters ? with awk/tr etc I substitute them. Now I study how iconv works.

[aram535]

Oh I see, sorry I misunderstood your question.

That is strange because grep shouldn't care about the unicode nature of the character.

Can you try using 'grep -b' or 'grep -U'? Does that work?

If not, provide a sample file and sample command that should work. We'll see what it looks like

[80aless]

No, -b messes it up even more and -U does nothing. My script opens the .cue file, greps TITLE and stores in in a file,then reads it and changes the names of the .mp3 files. Now I think the problem is not in the grep, but in a sub-script that I use to read a file: make a test.cue file with inside one title like:
TITLE "Ouverture sur des thèmes Hébreux Op 34b"
and read it with this function read_file.lib :

counter=0
while IFS='\012' read -r LINE
do
title[counter]="$LINE";
let counter=$counter+1 #counter
done < $1
export title
echo "${title[1]}"

I get:
read_file.lib test.cue
TITLE "Ouverture sur des th�mes H�breux Op 34b"

thanks

[aram535]

I didn't actually try your script, but from what I can tell this seems like a lot of work to rewrite `head`.

title =`head -1 $1`

And head will keep your characters intact.

[80aless]

Right, but if my .cue file has more lines I cannot put them in separate variables with head. In my previous script I put each line of the file to be read in the array "titles". And actually head has the same problem:

$ head -1 test.cue
TITLE "Ouverture sur des th�mes H�breux Op 34b"

[Cabhan]

I'm curious: do you see the special characters normally in your terminal ever? I wonder if maybe your terminal doesn't support Unicode. Because head and grep should definitely support Unicode.

[80aless]

indeed I cannot see it with less or head:
TITLE "Ouverture sur des th<E8>mes H<E9>breux Op 34b"
but today,while I was tinkering with the script I managed somehow to visualize the characters in the terminal.