for loop through find results in sh script

[dirkraft]

I've spent days on this now trying all kinds of things to get this for loop to work. My script is a bit more involved than this, but all my problems boil down to this.


dirkraft@ubuntu:~/Desktop/local$ find test/
test/
test/fileone
test/filetwo
dirkraft@ubuntu:~/Desktop/local$ cat printfiles.sh
#!/bin/sh

IFS=$'\n'
for x in `find test/`
do
echo "FILE"
echo $x
done

dirkraft@ubuntu:~/Desktop/local$ ./printfiles.sh
FILE
test/
test/fileo
FILE
e
test/filetwo
dirkraft@ubuntu:~/Desktop/local$


Output should have been very simple:


FILE
test/
FILE
test/fileone
FILE
test/filetwo


How can I loop over the results of a 'find' command?!

Thanks.

[falconindy]

while read line; do
#do whatever with each line of the results
done < <(find test)

[gmargo]

It's apparently to do with the IFS setting. While the $'\n' seems to work in bash, it does not seem to work in dash (aka sh). Change IFS to the following and your script works:


IFS='\

'

Could also do with this way (I've just always used the above):


IFS='
'

[dirkraft]

Thanks to both for your replies.

I just tried this and can't believe that this is the only way to set IFS to a newline in a script.

Thanks again.

[geirha]

Iterating find output with for is error-prone and bad practice. See http://mywiki.wooledge.org/BashFAQ/020 and http://mywiki.wooledge.org/UsingFind

In bash4 you can use globstar instead



shopt -s globstar
for file in test/**; do
echo "FILE: $file"
done