On Thu, Dec 10, 2009 at 10:01 AM, stan <[email protected]> wrote:
> BTW I copied the logfile.txt to one of my *NIX machines, and I was ging to
> send t, but it seems to not be a text file. What;s the storry here?
Windows "text" = UTF-16 with some kind of non-unicode encoding. It's
an ASCII superset, to it's usually reasonably processable by something
that can convert UTF-16 to UTF-8.
Here's the script I use to read ZWC logfiles from their CABs (although
I complain so loudly while doing so that I'm rarely asked to look at
such problems anymore .. sometimes being the squeaky wheel gets you
left alone!). Note that not all of the utf-16->utf-8 conversions
work, due to .. some windows stupidity I'm sure.
echo "** Extracting CAB"
cabextract -d "$dirname" "$cabfile" || exit 1
echo "** Fixing text files"
for i in $(find . -name '*.txt'); do
if ! iconv -f utf-16 -t utf-8 $i > $i.utf8; then
echo "skipping $i"
rm -f $i.utf8
continue
fi
mv $i.utf8 $i || exit 1
dos2unix $i
done
Dustin
--
Open Source Storage Engineer
http://www.zmanda.com
