UNIX Desktop Questions & Answers

I have loads of files in different directories which I do miner changed on them and till now I have to create totally new directories for them, specially as I deal with a whole directory at once
it creates such as mess really
e.g.

Code:

FILES="F1/*"
for X in $FILES
do
	name=$(basename $X) 
	sort  $X | uniq  | awk '{ $1=""}1' | awk '{ $NF=""}1'  > F2/${name} 
done

can it be written on the same files? on F1
can some one please let me know how I can write on output on the same exact file?

cheers

Any time you have a pipeline that reads from a file that is overwritten by a redirection at the end of the pipeline, you will lose data. The shell's will wipe out the current data in the file specified by the redirection before any command in the pipeline has a chance to read from it.

However, you could add:

Code:

        mv F2/$name /F1

after the pipeline starting with sort in your for loop as long as:If 1 or 3 above is not true, your script already has problems. If there are hard links to files in F1 that you need to preserve, your could replace the command added above with:

Code:

        cp F2/$name F1/$name && rm F2/$name

to preserve the links.

You could also simplify the and speed up your script if you replace:

Code:

sort  $X | uniq  | awk '{ $1=""}1' | awk '{ $NF=""}1'  > F2/${name}

in your loop with:

Code:

sort -u $X | awk '{ $NF="";$1=""}1' > F2/${name}

very helpful... thanks a lot