shell script programming problem
 

hi guys,

i'm new to linux and i am facing a problem with filenames that have spaces in them.

find $folder -type f -printf "%f/n" > file.tmp

I use the above line to copy filenames in a folder to a temp file, but if the filename contain spaces it will be written as 2 separate lines.

example: the filename "kenny rogers" will be displayed as

kenny
rogers

what should i do to?

thanks
Ken

Your command works as expected, but your script doesn't. We'll need to see the script to give an accurate solution.

Im trying to find files in a directory and sort them out according to highest frequency, then alphabetical order.

My code:

#!/bin/sh

start_folder=""

if [ "$1"="" ]; then
start_folder=.
else
start_folder=$1
fi

find $start_folder -type f -printf "%f\n" > file1.tmp
cat file1.tmp|uniq > file2.tmp

for line in `cat file2.tmp` # problem starts here as the for loop treats the space as a delimiter
do
count=`grep -x $line file1.tmp|wc -l`
echo $line $count >> file3.tmp
done

sort -k 2nr -k 1 file3.tmp | cut -d " " -f1 | head

rm *.tmp

Ok, I was worried you might be doing something non-standard. This is my normal solution to this kind of problem:

As a very quick explanation, the IFS variable defines what characters act as delimiters. So the lines in red simply store the normal value, replace that list with the newline as the only delimiter, do your processing, and restore the old value.

Also, the reason you need to surround $line in double quotes is for a similar reason. If your files contain spaces, grep can't know when your expression ends and filename(s) begin.

Thanks a lot!!! Dark.
You're a life saver.
By the way what commands can i use to test for empty files or directories?
I'm still new to linux and not used to the syntax.

The normal method is to use something like this:
The [ ] is a shorthand for the test command. You can snoop around what tests are available by reading the man pages: man test