manage_dropbox_images.sh: The goal of this script is to reduce the size of the dropbox folder where are automatically uploaded images taken with a smartphone, by linking to the same images found in other folders (for example if all images of your smartphone are regularly uploaded to a specific folder)

manage_dropbox_images.sh

#!/bin/bash

# The goal of this script is to reduce the size of the dropbox folder
# where are automatically uploaded images taken with a smartphone, by
# linking to the same images found in other folders (for example if all
# images of your smartphone are regularly uploaded to a specific folder)


DIR_CHARG=~/Dropbox/Chargements\ appareil\ photo/
DIR_IMG=~/Dropbox/Photos/Images/

LANG=C

# set -x

cd "$DIR_CHARG"

# ignore images that are already links
IMAGES=`find ./ -name "*.jpg" -type f | sort`

# we'll split found images by line, ignoring spaces in names
IFS="
"

# ok loop on all found images
for IMG_CHARG in $IMAGES
do

    # keep only the filename
    IMG_NAME=`basename "$IMG_CHARG"`

    # get the exif date of the image
    EXIV_CHARG=`exiv2 pr "$IMG_CHARG" | grep 'Image timestamp'| cut -c19-`

    # easily find other images, ie IMG_20140920_121314 will be a match for 2014-09-20-12-13-14
    IMG_RE=`echo "$IMG_NAME" | sed -e 's/[^0-9]\+/\\*/g'`

    # get the first image with a matching name
    IMG_ORIG=`find "$DIR_IMG" -name "*$IMG_RE.jpg" | head -1`

    # if an image is found, we'll ensure the equlity of the exif dates
    if [ "$IMG_ORIG" ]
    then
        # get the exif date of the found image
        EXIV_ORIG=`exiv2 pr "$IMG_ORIG" | grep 'Image timestamp'| cut -c19-`
        if [ "$EXIV_CHARG" != "$EXIV_ORIG" ]
        then
            # if the exif dates don't match, do as we didn't have found this image
            IMG_ORIG=""
        fi
    fi

    # we didn't found any matching image, so we'll check by exif date
    if [ ! "$IMG_ORIG" ]
    then
        # but we can't if the original image didn't have an exif date
        if [ "$EXIV_CHARG" ]
        then

            # check for images in the same minutes decade as the exif date
            DATE_RE=`echo "$EXIV_CHARG" | cut -c1-15 | sed -e 's/[^0-9]\+/\\*/g'`
            MATCH_IMGS=`find "$DIR_IMG" -name "*$DATE_RE[0-9]*[0-9][0-9]*.jpg"`

            # if not found, check for the same hour
            if [ ! "$MATCH_IMGS" ]
            then
                DATE_RE=`echo "$EXIV_CHARG" | cut -c1-13 | sed -e 's/[^0-9]\+/\\*/g'`
                MATCH_IMGS=`find "$DIR_IMG" -name "*$DATE_RE*[0-9][0-9]*[0-9][0-9]*.jpg"`
            fi

            # if not found, check for the same day
            if [ ! "$MATCH_IMGS" ]
            then
                DATE_RE=`echo "$EXIV_CHARG" | cut -c1-10 | sed -e 's/[^0-9]\+/\\*/g'`
                MATCH_IMGS=`find "$DIR_IMG" -name "*$DATE_RE*[0-9][0-9]*[0-9][0-9]*[0-9][0-9]*.jpg"`
            fi

            # check for a match in exif dates in the found images
            for MATCH_IMG in $MATCH_IMGS
            do
                # get the exif date of the found image
                EXIV_MATCH=`exiv2 pr "$MATCH_IMG" | grep 'Image timestamp'| cut -c19-`
                if [ "$EXIV_CHARG" == "$EXIV_MATCH" ]
                then
                    # we have a match, keep the image
                    IMG_ORIG=$MATCH_IMG
                    break;
                fi
            done

            # no image found with the correct exif date, stop here
            if [ ! "$IMG_ORIG" ]
            then
                echo "$IMG_CHARG => [FAILURE] NO IMG FOUND for exif date '$EXIV_CHARG'"
            fi

        # no exif date found in the original file, stop here
        else
            echo "$IMG_CHARG => [FAILURE] NO EXIF DATE FOUND"
        fi
    fi

    # an image is found, link to it
    if [ "$IMG_ORIG" ]
    then
        # get the relative path to make the link
        IMG_ORIG=`python -c "import os.path; print os.path.relpath('$IMG_ORIG')"`

        echo "$IMG_NAME => $IMG_ORIG"

        # and make the final link
        ln -f -s "$IMG_ORIG" "$IMG_NAME"
    fi
done