Quick question about batch renaming...

[Pierre-Alexandre]

I have to create a script which takes 3 command line args in order to rename all .ppm files in a folder. The first arg is the name of the files. The second is the base and the third is the increment. So let's say I enter Duplq.sh Hello 1 3, the script would rename all .ppm files to Hello1.ppm, Hello4.ppm, Hello7.ppm, etc.

The script must check if a file with the new name already exists. If it does, it copies it to a temp file and brings it back on the following iteration.

This, however, doesnt work at all... It's totaly broken. And I really dont know why!

Here is the code so far:



#!/bin/bash
###note: Mon systeme utilise le shell Dash par défaut, c'est pour ca que j'ai du mettre #!/bin/bash au lieu de sh.
if [ $# -ne 3 ]; then
echo "ERREUR: Ce script prend 3 arguments passés a la ligne de commande. USAGE: Duplq.sh NOM DEPART INCREMENT. Ex: Dupliq Allo 0 3"
exit 1
elif [ ! $(echo "$2" | grep -E "^[0-9]+$") ]; then
echo "ERREUR: $2 n'est pas un entier"
exit 1
elif [ ! $(echo "$3" | grep -E "^[0-9]+$") ]; then
echo "ERREUR: $3 n'est pas un entier"
exit 1
fi
ETAPE=$2
REFAIRE=false
for FICHIER in *.ppm
do
NFICHIER="${1}${ETAPE}.ppm"
if [ $REFAIRE == true ]
then
mv $TEMP $NFICHIER
ETAPE=$(($ETAPE+$3))
REFAIRE=false
else

if [ -e $NFICHIER ]
then
TEMP="${NFICHIER}.BAK"
cp $NFICHIER $TEMP
mv $FICHIER $NFICHIER
REFAIRE=true
else
mv $FICHIER $NFICHIER
ETAPE=$(($ETAPE+$3))
fi
fi
done

ETAPE = the current file were at.
NFICHIER = New file to be created
FICHIER = File in the for loop


Thanks!!


edit:

I must add that there is no verbrose output. Files simply disapear when I do another pass on them!

[knattlhuber]

To get more insight into what's going on, you could run your script with the bash debugging option (-x) on:


bash -x Duplq.sh hello 1 3

[lloyd_b]

Could you please explain exactly how you expect the $REFAIRE check to operate. It appears that if, for example, "FILE3.ppm" exists, it will be backed up, and then moved back to overwrite the "FILE3.ppm" that you just created in the previous iteration. I don't see the point.

In addition, each time this occurs, the for loop reads a new $FICHIER, which is then discarded. So on a rerun (with the same parameters) it will be discarding every other file...

You really need to rewrite this, so that ALL of the operations required for a given $FICHIER are performed without iterating the loop:
for FICHIER in *.ppm; do
REFAIRE=false
NFICHIER="${1}${ETAPE}.ppm"

# Check to see if destination file exists
if [ -e "$NFICHIER" ]; then
REFAIRE=true
TEMP="$NFICHIER.BAK"
mv "$NFICHIER" "$TEMP"
fi

# Move target to destination and increment file counter
mv "$FICHIER" "$NFICHIER"
ETAPE=$(($ETAPE+$3))

# If we created a backup file earlier, deal with it here
if [ $REFAIRE == true ]; then
mv "$TEMP" "$NFICHIER"
REFAIRE=false
fi
done

Note: I have NOT tested this code - it's just provided for illustration. Also note that I've put quotation marks around all filename variables - this will save you if any of them happen to have spaces in the name (though spaces in filenames will also cause trouble in the "for" loop).

For future reference - you may want to include some debugging "echo" commands (for instance, 'echo "Creating Backup of $NFICHIER"'). This is helpful in sorting out what's going on with the script. Of course, you can do the "bash -x ..." recommended by a previous poster as well.