using "for' in script

[sebastien_7]

Hello,

I have a file which contains a list of backup names, some names have spaces some don't, when using the following it appears that the names with space are treated as two.

for i in `cat $file1 | awk -F"\t" '{print $1}'`;do
echo $i
done

file1 contains things like:
backup1 offline
backup2_online
etc...

I am sure there is a way around this but cannot find it.

Please help
Sebastien

[john korterman]

Hi,
try with qoutes
echo "$i"

regards,
John K.

[Slawomir Gora · Accepted solution]

Hi,

try "while" insted of "for":

cat $file1 | while read
do
echo ${REPLY}
done

[sebastien_7]

thanks guys,
That was pretty fast, Slawomir's suggestion worked just fine using this:
cat $file1|awk -F"\t" '{print $1}'|while read
do
echo ${REPLY}
done

thanks again
sebastien

[Tim D Fulford]

Hi

The problem is you are in effect writing

for i in backup1 offline backup2_online
do
...
done

So the for loop cannot differentiate (even with quotes). You need to join them together. Perl would probably be the tool of choice, but as you've chosed awk I'll do it like so...Try

for i in $(cat file | awk -F"\t" '{print $1}' | sed -e "s/\ /:/g")
do
j=$(echo $i | sed "s/:/\ /g")
etc...
done

Tim

[Lisa Sorbo]

Hope I can tag on this thread. I have the same issue with spaces in file names - the listed solutions work for the echo stmt but I need to work with these files using a cp or mv or rm command - and I find that the same variable that is fine in the echo statement goes to the cp stmt in pieces again.

For example, if I get the file name into variable j, the stmt #cp /here/$j /there/$j will parse out the filename up to the first space.

any suggestions ?

Lisa

[A. Clay Stephenson]

cp /here/$j /there/$j

cp "/here/${j}" "/there/${j}"

Disciplined shell programmers get into the habit of encloses all variables with {}'s and quoting as well because whitespace has always been legal (if dumb) in UNIX pathnames.

[Muthukumar_5]

For loop will handle based on delimiters space or tab or new line also. You have to use while loop to handle "\n" as Record seperator. It is handled by IFS variable.

while read line;
do
echo $line
done <

It is more efficient than cat file | while.

===

lisa,

You can use methods as,

touch "hi bye"
here hi bye is a single file name.
var="hi bye"

You can move as,

# mv "$var" hi

or

# mv ${var} hi

hth.