Script to move files depending on size

[moonman89]

---

Hey all,
I want to make a script the will be used in a back up program, i want it to use a if statement to see weather a file is larger than 2gb and move it to a folder so it doesent take up Tape space for the back up...So when the script runs it will go through a folder and move anything that is over 2gb to /backup/exclude/ and all the other files that are smaller than 2gb will be left in the original folder /backup/include

I have no idea where to start, other than the ls -lhsS command
Any help?
Thanks HEAPS
Nathan

[sharonenoch]

I hipe you know bash scripting basics like the syntax of the if command and how to do basic comparison.

Then i can give u a start of how to go through with it

ls -l | awk '{print int($5)}'`

Try executing this command in shell within a directory. You will get the size of each file printed.

You can learn more about the 'awk' command from the net

[moonman89]

Hey, thanks. Yeah i do know about syntax.
But how do i set up the loop? A do loop? Untill the end of the folder?

if ls -l | awk '{print int($5)}'` >2.0gb then
cp ??????? ???????
fi


thanks....
Would it be like VB reading a text file via line?
Thanks..
Nathan

[Cabhan]

for line in `ls -1`
do
    size=`awk '{print int($5)}' $line`
    if [$size > 2097152]
    then
        cp $file 2_gig_folder/
    fi
done

Something like that perhaps? Note that I don't know Bash very well, so the above may well have errors.

Anyway, this uses a foreach loop to take each line from `ls -l` and run that loop on it.


EDIT:

I always forget about the find command... useful bugger, that.

[Dolda2000]

I'd do it like this:

find /backup/include -size +2000000k -exec mv -i {} /backup/exclude \;

That's a one-liner, so you can easily just run it directly from an interactive shell instead of creating a script.

If you really want to do it with a script (or just know how it would be done), there are, naturally, a great number of ways to do it, but I'd do it like this:

#!/bin/sh

find /backup/include -printf '%s %p\n' | while read size name; do
    if [ "$size" -gt 2000000000 ]; then
        mv -i "$name" /backup/exclude
    fi
done

[moonman89]

I want to do it in a script because the backup program will run, then exclude any files larger than the quota, move them to another pleace to exclude them, then tar all the other files to a tape, i have the rest but not the moving part, the one liner works, but not the .sh one...I c/p'd it but no luck, i have talked to me boss to see if he can help, but he is clueless like me...
Help?
Thanks

[sharonenoch]

I also tried the script which dolda has given but i got the output as this for a set of files. I tried for a lesser size just to see . The files were getting copied to the location mentioned but the errors were coming. so i added debug statements to it


find /usr/macubex -printf '%s %p\n' | while read size name; do
echo "Size - "$size
if [ $size -gt 20 ]; then
echo "moving" "$name"
mv -i $name /backup/exclude
fi
done

Otuput
Size - 4096
moving /usr/macubex
Size - 1036
moving /usr/macubex/Test_Binary.o
mv: cannot stat `/usr/macubex/Test_Binary.o': No such file or directory
Size - 6636
moving /usr/macubex/diskconf
mv: cannot stat `/usr/macubex/diskconf': No such file or directory
Size - 4666
moving /usr/macubex/Test_Binary.so
mv: cannot stat `/usr/macubex/Test_Binary.so': No such file or directory
Size - 204
moving /usr/macubex/Test_Binary.c
mv: cannot stat `/usr/macubex/Test_Binary.c': No such file or directory


From this i noticed that the "/usr/macubex" is been copied at the first itself as a directory into the folder /backup/exclude

That is why when it tried for the files after that it says cannot stat the respective file.
I am also not too familiar with the find command

[sharonenoch]

Code:
TEST="/backup/include/file.txt"
SOURCEFOLDER="/backup/include"
DESTINATIONFOLDER="/backup/exclude"
cd $SOURCEFOLDER
echo `ls -l > $TEST`
while read cmd_line
do
size=`echo $cmd_line | awk '{print int($5)}'`
filename=`echo $cmd_line | awk '{print $9}'`
if [ $size -gt 20 ]
then
echo $filename - $size
cp -i $SOURCEFOLDER"/$filename" $DESTINATIONFOLDER
fi
done < $TEST


This is kind of a roundabout method to do it using a for loop. but it will work
I think i have to look into the "find command and see if i am able to do it that way more easily

[sharonenoch]

Instead of "cp" use "mv". Ok

[davebardsley]

Far be it for me to advocate for dolda however his script works great. Not to put words in his mouth, but his script is meant to demonstrate the correct procedure but doesn't provide any error checking (it is up to you to fill in the detals). sharonenoch, you've changed your test file size to so small a size that it is on par with that of the directory. Either provide a copy-if-not-directory test, or use a larger file size and you'll see it (the script as shown) work as intended.

PS: don't forget the quotes around the name variable in the move statement or "funny" file names are going to give you woes.