ls command

[uguess]

Hi Guys

I have a directory structure like below
ac123
ac123456
ac123456
ac134
ac134455
ac194
ac1946778

I want to rename all the directories and its files starting with ac into ab. How can i do this.

I tries outputting ls *.* > list
then renaming them. However the ls produces output that is not in full path format. Is there a better and faster way of renaming all these files.


Thanks

 

[columb]

for i in ac*
do
  if [ -d $i ]
  then
    mv $i $(echo $i | tr 'c' 'b')
    for j in $(echo $i | tr 'c' 'b')/ac*
    do
      mv $j $(echo $j | tr 'c' 'b')
  done
done

This assumes your directory structure is exactly as described - i.e only one layer deep.

Columb Healy

 

[Annihilannic]

# make destination dirs
find ac* -depth -type d | while read NAME
do
        NEWNAME=$(echo ${NAME} | sed 's/\<ac/ab/g')
        [[ ! -d ${NEWNAME} ]] && mkdir -p ${NEWNAME}
done
# move and rename files
find ac* -depth -type f | while read NAME
do
        NEWNAME=$(echo ${NAME} | sed 's/\<ac/ab/g')
        mv ${NAME} ${NEWNAME}
done
# remove source dirs
rm -r ac*

This should work for any depth of directory structure, and will only change 'ac' if it occurs at the beginning of the file or directory name (the '\<' in the sed search expression ties it to the beginning of a word).

Annihilannic.

 

[Annihilannic]

A couple of points I thought of after my previous post. It is a Korn shell script, so should have #!/bin/ksh on the first line.

Also, you don't want to do that last step if the directories contain other files that aren't called ac*.

Another, perhaps safer, recursive solution:

#!/bin/ksh

recursive_rename() {
        ls -d $* | while read NAME
        do
                [[ -d ${NAME} ]] && (cd ${NAME} ; recursive_rename ac*)
                mv ${NAME} $(echo ${NAME} | sed 's/\<ac/ab/g')
        done
}

recursive_rename ac*

Annihilannic.