Shell Programming and Scripting

oops the last file posted was dos2unixed here is the original

When I open the original file in a UNIX text editor I see:

Code:

Distributeur
Alliance Vivafilm
Origine
...

Don't edit text files for UNIX in notepad. If you'd been editing them in UNIX you'd have seen those characters; they're what Notepad inserts at the beginning of Unicode text files to warn itself that they're unicode(UTF8 in this case), and that's what's messing up your comparison.

Here's an improvement on dos2unix for that, stripping out those faulty characters and windows' carriage returns at the same time:

Code:

tr -d '[\r\357\273\277]' < wingarbage.txt > actualtext.txt

try to use :

Code:

if [ "$LINE" != "$Distor" ]; then

i did before posting...the test is OK... the problem is not there...
I extract those from a file i download from a website, when downloaded the file is ISO-8859-1 i use:

Code:

vim +"set bomb | set fileencoding=utf-8 | wq" $(find . -type f -name info.txt)

to change to UTF8

That has done more than changing the character set seemingly -- it also put the identifier in the front.

THANKX !!!

Code:

$ ./test.sh 
Distributeur
Distributeur

123-Distributeur-123
123-Distributeur-123
Distributeur
Alliance Vivafilm

what would be the best way to change encoding w/o using vim ?

Sort the format out on the Microsoft platform.
Imho there is too much effort spent in the unix world sorting out Microsoft file formats.