@bytesnz

Jack Farley, Computer Engineer

To handle the manage the amounts of images we seem to be producing on our trips, I created a few simple scripts to do a few tasks that I found myself doing constantly.

Renaming

As Canons files are name IMGXXXX.JPG, where XXXX is a one up number, and everytime you start a new folder (good for a new place) on the 450D, it starts from 1, you end up with quite a few duplicate names. To get around this, I add the folder identifier to the name using this simple bash script. The script can also make web-ready versions of the image, but I now use the next script for that.

#!/bin/bash
# Goes through folders and renames images depending on folder name and
# makes a web ready version of the image

rF=
num=1
chars=XX
cD=`pwd`

##find . -type f -name '*.*' -print0 | while IFS= read -r -d '' file; do
echo "$@" | while IFS= read -r -d '~' d; do
#for d in "$@"; do
        if [ -f "$d" ]; then
                dN=`basename "$d"`
                convert -auto-orient -resize 1100x1100 -quality 80 "$d" "$rF/$dN"
                echo resized $d - $rF/$dN
        elif [ -d "$d" ]; then
                p=''
                dN=`basename "$d"`
                if (echo $dN | grep -q '..._....'); then
                        p=`echo $dN | sed -re 's/(...)_....\/?/\1/'`
                        p="G$p"
                        echo Folder $d from G15 - p = $p
                elif (echo $dN | grep -q '...CANON'); then
                        p=`echo $dN | sed -re 's/(...)CANON\/?/\1/'`
                        p="4$p"
                        echo Folder $d from 450D - p = $p
                elif (echo $dN | grep -q '....CANON'); then
                        p=`echo $dN | sed -re 's/(....)CANON\/?/\1/'`
                        echo Folder $d from a labelled Canon device - p = $p
                else
                        p=$chars$((num++))
                        echo Unknown folder - p = $p
                fi

                if [ "$p" ]; then
                        # Move into folder
                        cd "$d"
                        # Rename files
                        for f in `ls ???_????.??? 2>/dev/null`; do
                                n=`echo $f | sed -re "s/(...)(_....\....)/\1_$p\2/"`
                                mv $f $n
                        done

                        if [ "$rF" ]; then
                                # Make sure resized folder exists
                                if [ ! -d $rF ]; then
                                        mkdir $rF
                                fi

                                # Make web version of image
                                for f in *.JPG; do
                                        convert -auto-orient -resize 1100x1100 -quality 80 "$f" "$rF/$f"
                                        echo resized "$f"
                                done
                        fi

                        # Move back into original folder
                        cd "$cD"
                        echo finished folder $d
                else
                        echo Uknown type of folder - ignoring
                fi
        else
                echo Not sure what to do with $d, ignoring.
        fi

done

Resizing

I currently keep a directory of copied web-ready images so that it can synced with the server holding my blog whenever I get a chance. This script will create web-ready images for any new images it finds. One thing I do need to do to it is make it so it will update the photo if it has been modified.

dir=/home/photos/web-ready

find /home/photos/large -iname '*.jpg' | while read file; do

        path=`dirname "$file"`
        name=`basename "$file"`

        if [ ! -d "$dir/$path" ]; then
                mkdir -p "$dir/$path"
        fi

        if [ ! -f "$dir/$path/$name" ]; then
                convert -auto-orient -resize 1100x1100 -quality 80 "$file" "$dir/$path/$name"

                echo Done $file
        else
                echo Skipping $file
        fi
done