Interactive script for downloading software from network.pivotal.io using the download API. I have created a newer version at https://gist.github.com/mnemesh-pivotal/940d4358e421fd19caad42d03a1af7c3. It uses jq instead of sed/awk/grep/tac/etc to gather information from the the json files.

get_piv_sw.sh

#!/bin/bash
[ "$1" = "-d" ] && debug=on || debug=off

# Make working directory
export wrkdir=/tmp/wrkdir.$$
mkdir -p ${wrkdir}

# Clean up function
clean_up()
{
	trap ":" 1 2 3 15
	if [ "${SIGSET}" ]
	then
		printf "\n%s aborted...\n"  $(basename $0)
		[ -f "${dldir}/${target_file}" ] && rm -f ${dldir}/${target_file}
	fi
	[ "${wrkdir}" ] && rm -rf ${wrkdir}
}

# Clean up if the script is quit
trap "SIGSET=TRUE;clean_up;exit 1" 1 2 3 15

# Get download directory or hard code and remove the next hash mark
#export dldir=/tmp
while [ -z ${dldir} ]; do
	read -p "Enter the download directory : " dldir
	if [ ! -d "${dldir}" ]
	then
		printf "\n%s does not exist...\n" ${dldir}
		unset dldir
	elif [ ! -w "${dldir}" ]
	then
		printf "\n%s is not writable...\n" ${dldir}
		unset dldir
	fi
done
while true; do
	[ ${debug} = 'on' ] && echo "dldir=\""${dldir}"\""

	# Get api_token or hard code and remove the next hash mark
	#api_token="aAaAaAaAaAaAaAaAaAaA"
	[ -z ${api_token} ] && read -p "Enter your network.pivotal.io API Token: " api_token
	[ ${debug} = 'on' ] && echo "api_token=\""${api_token}"\""

	# Authenticate
	curl --silent -i -H "Accept: application/json" -H "Content-Type: application/json" -H "Authorization: Token ${api_token}" -X GET https://network.pivotal.io/api/v2/authentication > ${wrkdir}/authenticate

	if [ $(grep -c "HTTP/1.1 200 OK" ${wrkdir}/authenticate) -ne 1 ]
	then
		printf "Authentication failed, please check your API Token and try again.  Exiting...\n"
	cat ${wrkdir}/authenticate
		clean_up
		exit 1
	fi

	# Get products list
	curl --silent -H "Accept: application/json" -H "Content-Type: application/json" -H "Authorization: Token ${api_token}" -X GET https://network.pivotal.io/api/v2/products | python -mjson.tool > ${wrkdir}/prod_list
	grep '"name":' ${wrkdir}/prod_list | egrep -vi 'suite|foundry|buildpacks' | cut -d'"' -f4 | sort > ${wrkdir}/prod_name_list

	# Get product name
	printf "\nThese are the available products:\n\n"
	column ${wrkdir}/prod_name_list

	printf "\n"
	while true; do
		read -p "Which product do you want to download? " prod_name
		if [ $(grep -c "^${prod_name}$" ${wrkdir}/prod_name_list) -eq 1 ]
		then
			break
		else
			printf "Sorry can't find that product, please re-enter the product name from the list above\n"
		fi
	done

	# Get slug
	export prod_slug=$(sed -n "/\"${prod_name}\"/,/\"slug\"/p" ${wrkdir}/prod_list | tail -1 | cut -d'"' -f4)
	[ ${debug} = 'on' ] && echo "prod_slug=\""${prod_slug}"\""

	# Get product ID
	export prod_id=$(curl --silent -H "Accept: application/json" -H "Content-Type: application/json" -H "Authorization: Token ${api_token}" -X GET https://network.pivotal.io/api/v2/products | python -mjson.tool | sed -n "/${prod_slug}\/releases/,/\"id\":/p" | tail -1 | tr -dc '[:digit:]')
	[ ${debug} = 'on' ] && echo "prod_id=\""${prod_id}"\""

	# Get release ID
	curl --silent -H "Accept: application/json" -H "Content-Type: application/json" -H "Authorization: Token ${api_token}" -X GET https://network.pivotal.io/api/v2/products/${prod_slug}/releases | python -mjson.tool > ${wrkdir}/prod_releases

	# Get version
	printf "\nThese are the available versions:\n\n"
	grep '"version":' ${wrkdir}/prod_releases | cut -d'"' -f4
	printf "\n"
	while true; do
		read -p "Which version do you want to download? " ans
		prod_version=$(echo ${ans} | sed -e 's/\./\\./g')
		if [ $(grep -c "\"version\": \"${prod_version}\"" ${wrkdir}/prod_releases) -eq 1 ]
		then
			break
		else
			printf "Sorry can't find that version, please re-enter the product version from the list above\n"
		fi
	done
	[ ${debug} = 'on' ] && echo "prod_version=\""${prod_version}"\""

	# Get release ID
	export rel_id=$(tac ${wrkdir}/prod_releases | sed -n "/\"version\": \"${prod_version}/,/\"id\":/p"|tail -1|tr -dc '[:digit:]')
	[ ${debug} = 'on' ] && echo "rel_id=\""${rel_id}"\""

	# Get file ID
	curl --silent -H "Accept: application/json" -H "Content-Type: application/json" -H "Authorization: Token ${api_token}" -X GET https://network.pivotal.io/api/v2/products/${prod_slug}/releases/${rel_id} | python -mjson.tool > ${wrkdir}/prod_fileid

	# Get file to download
	printf "\nThese are the available files:\n\n"
	sed -n "/\"download\":/,/\"name\":/p" ${wrkdir}/prod_fileid | grep name | cut -d'"' -f4
	printf "\n"
	while true; do
		read -p "Which do you want to download? " prod_file
			if [ $(grep -c "\"name\": \"${prod_file}\"" ${wrkdir}/prod_fileid) -eq 1 ]; then
				echo -e "INFO - Setting up request to download the file: " $prod_file
				prod_file=$(sed 's/[\/]/\\\//g' <<< $prod_file)  #This is to escape the file name that has / on the name ( like PL/PERL ) , since the sed assumes that it is its command
				export download_url=$(tac ${wrkdir}/prod_fileid | sed -n "/\"name\": \"${prod_file}\"/,/\/download\"/p"|tail -1|cut -d'"' -f4)
				[ ${debug} = 'on' ] && echo "download_url=\""${download_url}"\""
				export target_file=$(basename $(tac ${wrkdir}/prod_fileid | sed -n "/\"name\": \"${prod_file}\"/,/aws_object_key/p"|tail -1|cut -d'"' -f4))
				[ ${debug} = 'on' ] && echo "target_file=\""${target_file}"\""
				break
			elif [ $(grep -c "\"name\": \"${prod_file}\"" ${wrkdir}/prod_fileid) -gt 1 ]; then
				echo -e "\nINFO - Found multiple version for file:" $prod_file
				prod_file=$(sed 's/[\/]/\\\//g' <<< $prod_file)
				tac ${wrkdir}/prod_fileid | sed -n "/\"name\": \"${prod_file}\"/,/\/download/p" > ${wrkdir}/prod_fileid_version
				printf "\nThese are the available version for the file:\n\n"
				tac ${wrkdir}/prod_fileid | sed -n "/\"${prod_file}\"/,/file_version/p" | grep file_version | cut -d '"' -f 4
				printf "\n"
				while true; do
					read -p "Which version of the file \" ${prod_file} \" do you want to download? " prod_file_version
					if [ $(grep -c "\"file_version\": \"${prod_file_version}\"" ${wrkdir}/prod_fileid_version) -eq 1 ]; then
						export download_url=$(cat ${wrkdir}/prod_fileid_version | sed -n "/\"file_version\": \"${prod_file_version}\",/,/\/download\"/p"|tail -1|cut -d'"' -f4)
						[ ${debug} = 'on' ] && echo "download_url=\""${download_url}"\""
						export target_file=$(basename $(cat ${wrkdir}/prod_fileid_version | sed -n "/\"file_version\": \"${prod_file_version}\",/,/aws_object_key/p"|tail -1|cut -d'"' -f4))
						[ ${debug} = 'on' ] && echo "target_file=\""${target_file}"\""
						break
					else
						printf "sorry can't find the version, please re-enter the version again from the list above \n"
					fi
				done
				break
			else
				printf "Sorry can't find that file, please re-enter the product file from the list above\n"
			fi
	done

	export download_url=$(tac ${wrkdir}/prod_fileid | sed -n "/\"name\": \"${prod_file}\"/,/\/download\"/p"|tail -1|cut -d'"' -f4)
	[ ${debug} = 'on' ] && echo "download_url=\""${download_url}"\""

	export target_file=$(basename $(tac ${wrkdir}/prod_fileid | sed -n "/\"name\": \"${prod_file}\"/,/aws_object_key/p"|tail -1|cut -d'"' -f4))
	[ ${debug} = 'on' ] && echo "target_file=\""${target_file}"\""

	# Accept EULA
	curl --silent -H "Accept: application/json" -H "Content-Type: application/json" -H "Content-Length: 0" -H "Authorization: Token ${api_token}" -X POST https://network.pivotal.io/api/v2/products/${prod_slug}/releases/${rel_id}/eula_acceptance | python -mjson.tool > ${wrkdir}/eula_acceptance

	if [ $(grep -c "accepted_at" ${wrkdir}/eula_acceptance) -ne 1 ]
	then
		if [ $(grep -c "\"status\": 401" ${wrkdir}/eula_acceptance) -eq 1 ]
		then
			printf "EULA acceptance failed, user could not be authenticated.  Exiting...\n"
		elif [ $(grep -c "\"status\": 404" ${wrkdir}/eula_acceptance) -eq 1 ]
		then
			printf "EULA acceptance failed, product or release cannot be found.  Exiting...\n"
		else
			printf "EULA acceptance failed, command reults were:\n"
			cat ${wrkdir}/eula_acceptance
			printf "Exiting...\n"
		fi
		clean_up
		exit 1
	fi

	# Show environment variables if debug mode is on
	if [ ${debug} = 'on' ]
	then
		echo
		echo "Variables to export to match the executed environment"
		echo "export dldir=\""${dldir}"\""
		echo "export api_token=\""${api_token}"\""
		echo "export prod_name=\""${prod_name}"\""
		echo "export prod_slug=\""${prod_slug}"\""
		echo "export prod_version=\""${prod_version}"\""
		echo "export prod_id=\""${prod_id}"\""
		echo "export rel_id=\""${rel_id}"\""
		echo "export prod_file=\""${prod_file}"\""
		echo "export download_url=\""${download_url}"\""
		echo "export target_file=\""${target_file}"\""
	fi

	# Download product file
	wget --output-document="${dldir}/${target_file}" --post-data="" --header="Authorization: Token ${api_token}" ${download_url}

	unset ans
	while [ -z "${ans}" ]; do
		read -p "Would you like to download another product? (Y/N) " ans
		yorn=$(echo ${ans} | tr [:upper:] [:lower:])
		case ${yorn} in
			y|n)
				break
				;;
			*)
				unset ans
				;;
		esac
	done
	if [ ${yorn} = "n" ]; then break; fi
done
# Clean up on exit
clean_up
exit 0

Hello,

Thanks for the script , very useful , but i found two bugs in the script.

1. If the file name contains "/" then the script breaks ( eg.s PL/PERL etc )
2. There are some places where there is multiple product file name , like for eg.s on the version of GPDB 4.3.4.0 you will find multiple file with the name "Greenplum Database 4.3 - PL/R Extension for RHEL" , the script breaks since its only accepts one entry for it.

To fix both the above bug , i have modified the code to the one below ..

while true; do
    read -p "Which do you want to download? " prod_file
        if [ $(grep -c "\"name\": \"${prod_file}\"" ${wrkdir}/prod_fileid) -eq 1 ]; then
            echo -e "INFO - Setting up request to download the file: " $prod_file
            prod_file=$(sed 's/[\/]/\\\//g' <<< $prod_file)  #This is to escape the file name that has / on the name ( like PL/PERL ) , since the sed assumes that it is its command
            export download_url=$(tac ${wrkdir}/prod_fileid | sed -n "/\"name\": \"${prod_file}\"/,/\/download\"/p"|tail -1|cut -d'"' -f4)
            [ ${debug} = 'on' ] && echo "download_url=\""${download_url}"\""
            export target_file=$(basename $(tac ${wrkdir}/prod_fileid | sed -n "/\"name\": \"${prod_file}\"/,/aws_object_key/p"|tail -1|cut -d'"' -f4))
            [ ${debug} = 'on' ] && echo "target_file=\""${target_file}"\""
            break
        elif [ $(grep -c "\"name\": \"${prod_file}\"" ${wrkdir}/prod_fileid) -gt 1 ]; then
            echo -e "\nINFO - Found multiple version for file:" $prod_file
            prod_file=$(sed 's/[\/]/\\\//g' <<< $prod_file)
            tac ${wrkdir}/prod_fileid | sed -n "/\"name\": \"${prod_file}\"/,/\/download/p" > ${wrkdir}/prod_fileid_version
            printf "\nThese are the available version for the file:\n\n"
            tac ${wrkdir}/prod_fileid | sed -n "/\"${prod_file}\"/,/file_version/p" | grep file_version | cut -d '"' -f 4
            printf "\n"
            while true; do
                read -p "Which version of the file \" ${prod_file} \" do you want to download? " prod_file_version
                if [ $(grep -c "\"file_version\": \"${prod_file_version}\"" ${wrkdir}/prod_fileid_version) -eq 1 ]; then
                    export download_url=$(cat ${wrkdir}/prod_fileid_version | sed -n "/\"file_version\": \"${prod_file_version}\",/,/\/download\"/p"|tail -1|cut -d'"' -f4)
                    [ ${debug} = 'on' ] && echo "download_url=\""${download_url}"\""
                    export target_file=$(basename $(cat ${wrkdir}/prod_fileid_version | sed -n "/\"file_version\": \"${prod_file_version}\",/,/aws_object_key/p"|tail -1|cut -d'"' -f4))
                    [ ${debug} = 'on' ] && echo "target_file=\""${target_file}"\""
                    break
                else
                    printf "sorry can't find the version, please re-enter the version again from the list above \n"
                fi
            done
            break
        else
            printf "Sorry can't find that file, please re-enter the product file from the list above\n"
        fi
done

Thanks, @faisalpivotal. I've incorporated your edit, added a loop for multiple downloads, and a fix for retrieving the rel_id I encountered while trying to download the latest Ops Metrics release.