gi-get.sh

#!/bin/bash

#
# As-is, 100% free, late-night software.
#
# DO NOT RUN AS ROOT!
#
# Description: By default, tries to download 300 google
# image search results for search string into folder named 'img'.
#
# Requires: tidy and wget
# apt-get install wget tidy (Ubuntu/Debian)
# or yum install wget tidy (RedHat/CentOS/Fedora)
#

# max downloads
maxsearches=300
# download path
downloadpath="img"

# tmp dir (gets deleted)
tmpdir=$(mktemp -d)

# read your search term
read -p "enter search term: " search
search=$(echo $search | tr ' ' '+')

# make folder for downloads
mkdir -p $downloadpath || exit 1

# useragent (google search does not like wget usergent)
useragent='Mozilla/5.0 (Windows; U; Windows NT 5.1; en-GB; rv:1.8.1.6) Gecko/20070725 Firefox/2.0.0.6'

pushd $downloadpath &>/dev/null || exit 1
num=0
while [[ $num -lt $maxsearches ]]; do
        prev=$num
        let num=num+20

    # poor man's progress bar
        echo -en "${prev} through ${num}..."

    # google url for search
        googleurl="https://www.google.com/search?q=${search}&um=1&hl=en&safe=off&start=${num}&sa=N&tbm=isch&sout=1"

    # grab google results
        wget --quiet -U "$useragent" -O ${tmpdir}/google "$googleurl"

    # clean up google output
        tidy -q ${tmpdir}/google 2>/dev/null > ${tmpdir}/google.tidy

    # grab urls
        grep http ${tmpdir}/google.tidy | sed -e 's:amp:\n:g' | grep imgurl | sed -e 's:".*imgurl=::g' -e 's:\&$::g' > ${tmpdir}/urls

    # download all urls
        for url in $(cat ${tmpdir}/urls); do
                wget --timeout=2 --dns-timeout=3 --connect-timeout=2 --read-timeout=3 \
                     --quiet -U "$useragent" -A jpg,png,gif "$url"
        # progress bar
                echo -en "."
        done

    # clean up
        echo "done."
done

rm -rf ${tmpdir}
popd &>/dev/null

exit 0