A bash function to scrape the wallpaper site ``simpledesktops.com"

simpledesks.sh

function simpledesks() {
    useragent='Mozilla/5.0 (Windows NT 6.3; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/33.0.1750.117 Safari/537.36' # Mimic Google Chrome 33.0.1750.117 on 64-bit Windows 8.1

    if [ $# != 1 ]
    then
        echo 'Enter the number of pages you want to scrape as the only argument. Each page has (give or take) 28 images. The total number of pages at this time is 46.'
        return 1
    elif [ $1 -gt 46 ]
    then
        limit=46
    else
        limit=$1
    fi

    limit=$1
    counter=1
    
    while [ $counter -le $limit ]
    do
        for url in $(curl -s -A "$useragent" http://simpledesktops.com/browse/$counter/ | grep '<h2><a href="/browse/desktops' | sed 's/\t*<h2><a href="\(.*\)">.*<\/a><\/h2>$/http:\/\/simpledesktops.com\1/')
        do
            name=$(sed 's/^.*\/[0-9][0-9]\/\(.*\)\/$/\1/' <<< $url)
            echo -n Downloading $name...
            if [ $(ls -1 | grep ^$name\....) ]  # If we already have the file, no need to re-download
            then
                :
            else
                curl -s -L -o temp -A "$useragent" $(curl -s $url | grep '<h2><a href="/download' | sed 's/^\t*<h2><a href="\(.*\)">.*<\/a><\/h2>$/http:\/\/simpledesktops.com\1/')
                mv temp $name.$(file -b temp | cut -f1 -d' ' | tr [A-Z] [a-z] | sed 's/jpeg/jpg/')    # I know this causes more overhead, but I've noticed a few files aren't PNG's
            fi
            echo ' Done'
        done
        counter=$(($counter + 1))
    done
}