fwui.log error while running this script

Question

I run this bash script in order to put all the files with the following format dated today in the same folder. for example here I want 2012-12-17_083213 to be my folder and then inside it I have the .log and .log_stats and etc with this name inside it?

But I am not able to get the script working. For example, if I try everything up until dir=$(echo $i | awk '{split($1,a,"_"); print a[1]}');
And then to test it, I did echo $dir, and that displays the following:
fwui.log
I searched and understood its about locks in the db but how to solve it?

This is the script:

#!/bin/bash  
tosend=""  
for i in *.log; do     
    dir=$(echo $i | awk '{split($1,a,"_"); print a[1]}');  
    if [ ! -d $dir ]; then  
        mkdir $dir;  
    fi;  
    cp $i $dir;  
    tosend="$tosend\n$dir"  
done  
for i in $(echo -e $tosend | uniq);do  
    echo Compressing $i  
    tar -zcvf $i.tar.gz $i  
    scp $i.tar.gz user@server  
done  

and this is the file formats:

2012-12-17_083213_1.log  
2012-12-17_083213_1.log_stats  
2012-12-17_083213_1.logaccount_ptr    
2012-12-17_083213_1.loginitial_ptr  
2012-12-17_083213_1.logptr  
2012-12-17_093049_2.log  
2012-12-17_093049_2.log_stats  
2012-12-17_093049_2.logaccount_ptr  
2012-12-17_093049_2.loginitial_ptr  
2012-12-17_093049_2.logptr  

Answer 1

first change that 1st awk line with:

dir="$(echo "$i" | awk '{split($0,a,"_"); printf "%s_%s",a[1],a[2]}')"

so that, in case they contain whitespaces, it will still work.

Then replace the "for i in" line with:

echo -e "$tosend" | grep -v '^$' | sort | uniq | while IFS= read i ;do

and add some "" around relevant places:

#!/bin/bash  
tosend=""  
for i in 20*.log*; do     
    dir="$(echo "$i" | awk '{split($1,a,"_"); printf "%s_%s",a[1],a[2]}')";  
    if [ ! -d "$dir" ]; then  
        mkdir "$dir" || { echo "ERROR during: mkdir '$dir'" ; exit 1 ; };  
    fi;  
    cp "$i" "$dir" || { echo "ERROR during: cp '$i' '$dir'" ; exit 2 ; };  
    tosend="${tosend}\n${dir}"  
done  
echo -e "$tosend" | grep -v '^$' | sort | uniq| while IFS= read i ;do  
    echo "Compressing $i"  
    echo REMOVE THIS ECHO ONCE TESTED tar -zcvf "${i}.tar.gz" "$i" \
    && { echo REMOVE THIS ECHO ONCE TESTED scp "${i}.tar.gz" user@server:/tmp/somedir/ \
         || { echo "ERROR during: scp '${i}.tar.gz' user@server:/tmp/somedir/" ; exit 4 ; };
       }
    || { echo "ERROR during: tar -zcvf '${i}.tar.gz' '$i'" ; exit 3 ; }; 
done  

you may want to take out the "v" part in the tar, to save some display (once you know it works)

another way for the above:

replace my

for i in 20*.log*; do 

with:

ls -1d *|grep '^[0-9]*-[0-9]*-[0-9]*_[0-9]*_[0-9]*\.log.*'| while IFS= read i ; do

Or simplify the grep :

ls -1d *|grep '^[0-9_-]*\.log' | while IFS= read i ; do

which should be enough (but could match a "./-_-1234.log.hahaha" file ...)

Answer 2

You should start with for i in *.log*; do instead of for i in *.log; do; otherwise you only get files that ends in .log. Beside the fact that it may possibile be rewritten simpler, I cannot see any other problem in you script.