freesound.org sound-samples downloader

freesound_downloader.sh

#!/bin/bash

#############
# CONFIG
redisDB=0
redisClient="/usr/local/bin/redis-cli"
redisServer="/usr/local/bin/redis-server"

diff(){
  awk 'BEGIN{RS=ORS=" "}
       {NR==FNR?a[$0]++:a[$0]--}
       END{for(k in a)if(a[k])print k}' <(echo -n "${!1}") <(echo -n "${!2}")
}

# Welcome:
clear
echo "-------------------------------------------------------------------------------------------------"
echo "| Warning: This script WILL NOT check if your disk has enough free space to store the downloads. |"
echo "-------------------------------------------------------------------------------------------------"
#filesize prefetch via curl-header

# Starting redis server
coproc $redisServer

# Ask for download-folder
echo -n -e "download folder (`pwd`/freesound): "
read path
if [ -z "$path" ]
then
	path="`pwd`/freesound"
fi
echo $path 


# Make sure the download-folder exists; if not, we'll try to create it
if [ -d "$path" ]
then
	cd $path
else
	mkdir $path
	cd $path
fi

# Fetch all collection-ids
ids=(`curl http://www.freesound.org/packsView.php -s | egrep -o 'packsViewSingle.php\?id=[0-9]*' | egrep -o '[0-9]*'`)

# Before we start, tell how many file we'll download
echo "${#ids[*]} sample-collections found on freesound.org"

# Check for filenames, already in redis
rediskeys=(`$redisClient -n $redisDB KEYS \*`)
newIds=($(diff ids[@] rediskeys[@]))
diff=$((${#ids[@]} - ${#newIds[@]}))
echo "$diff already in redis"
echo "${#newIds[@]} filenames remaining"

# Get the filename of each remaining collection
files=( )
echo "Grabbing the archive-filenames of ${#newIds[@]} collections. This could take a while."
i=0
for id in ${ids}
do
	((i++))
	echo -n "$i of ${#ids}: "
	filepath=`curl http://www.freesound.org/packsViewSingle.php?id=$id -s | egrep -o 'http://www.freesound.org\/downloadpack\/.*\.zip'`
	filename=`echo $filepath | egrep -o '[0-9]*_.*\.zip'`	
	
	$redisClient -n $redisDB SET $id $filename > /dev/null
	echo "OK"
	
	echo $filepath
	`curl $filepath -H`
	if [ $i -eq 2 ]
	then
		break
	fi
done

#$redisClient -n $redisDB KEYS \*
#save and shutdown the server (which we opened via 'coproc')
$redisClient shutdown