capture exit code of function inside for loop

[Paolo Gilli]

hi,
I have this problem running a for loop:
-----------------------------------------------
for MONTHTOCHECK in `find /pippo -path "/pippo/*" -prune -name "DC*.DAT*" -type f -mtime +45 |sed s/.*DC/DC/| sort -d | awk '{ printf( substr($1,3,6)"\n" ); }' | uniq`
do
echo ciao
done

find: cannot stat /pippo
$ echo $?
0
-----------------------------------------------
how cah I cpture the exit code of find or any other function inside the for loop?

Thankyou
Paolo

[RAC_1]

for MONTHTOCHECK in `find /pippo -path "/pippo/*" -prune -name "DC*.DAT*" -type f -mtime +45 |sed s/.*DC/DC/| sort -d | awk '{ printf( substr($1,3,6)"\n" ); }' | uniq`
export exit_code=$?
do
echo ciao
done

echo ${exit_code)

Anil

[Ranjith_5]

HI Paolo,

echo $?

the above will give the status value of last command executed.

See an example.

SVR#-/crash>ls
bounds crash.0 iostat lost+found web.log.0827_1400hrs.Z
SVR#-/crash>echo $?
0
SVR#-/crash>xyz
ksh: xyz: not found
SVR#-/crash>echo $?
1

This means, for a command which is executed succesfully, the value of $? will be zero. If the last command is not successfull it will give a value which is non zero. This function is the common method which is used to check the exit code, and widely used in shell scripts.

Regards,
Syam

[Fred Ruffet]

Something strange with your find command :
If you use -path instead of -name, path will be used, but leading / removed (see manpage for details). So, you're looking for all files under pippo dir in /pippo (i.e. /pippo/pippo/*). Then you apply -name... very strange.

If I really understand, you want all files under /pippo named DC*.DAT* that have been modified under 45 days, without going into subdirs ? here it is :
find /pippo/* -prune -type f -name "DC*.DAT*" -mtime +45

Regards,

Fred

[Muthukumar_5 · Accepted solution]

One was as,

for file in `find . -type f;echo $?`; do
echo $file
done

Last file data contains $? (return value of find command)

Else try as,

find /pippo -path "/pippo/*" -prune -name "DC*.DAT*" -type f -mtime +45 |sed s/.*DC/DC/| sort -d | awk '{ printf( substr($1,3,6)"\n" ); }' | uniq > /tmp/testlog

if [[ $? -eq 0 ]]
then
while read line; do
echo $line
done < /tmp/testlog
fi

HTH.

[Paolo Gilli]

Thank to all,
What I need is know if it's possible capture the error that can occour in any function inside the for loop.
So if find,sort,awk or uniq fail, I can capture the exit code !0 and stop the proceure

Anil, your suggestion don't work

[Peter Godron]

Paolo,
I suggest you break your oneliner into seperate steps and check the error code at each step. (find/sed/sort/awk)
The end result would then be a temporary file with all your data, which you can then process within your loop.
Regards

[RAC_1]

There is a typo in my post.
echo ${exit_code) needs to as follows.
echo ${exit_code}

[Steve Steel]

Hi

do the find twice is the best way

example

param=$1
find $param > /dev/null 2>&1
if [ "$?" = "0" ]
then
for a in $(find $param -name *.log)
do
echo $a
done
else
echo bad
fi


Steve Steel

[Paolo Gilli]

you wrote:
for MONTHTOCHECK in `find /pippo -path "/pippo/*" -prune -name "DC*.DAT*" -type f -mtime +45 |sed s/.*DC/DC/| sort -d | awk '{ printf( substr($1,3,6)"\n" ); }' | uniq`
export exit_code=$?
do
echo ciao
done

echo ${exit_code)

but "export exit_code=$?" is before "do" so the loop don't work