#!/bin/bash
: '
 @vendor    Chorke Academia, Inc.
 @web       https://cdn.chorke.org/docs/academia
 @version   1.0.00
 @since     1.0.00
'

:<<EOF
if [[ -x "$(command -v curl)" ]];then \
sudo dnf -q clean all;\
export SSH_PUB_KEY_TAG="spark@${HOSTNAME}";\
export PYSPARK_PYTHON_DIR='/usr/bin/python3';\
export HADOOP_HOME_DIR='/var/hadoop/hadoop-3.3.4';\
export SPARK_HOME_DIR='/var/spark/spark-3.3.0-bin-hadoop3';\
export JAVA_HOME_DIR='/usr/lib/jvm/java-11-openjdk-11.0.17.0.8-1.fc37.x86_64';\
bash <(curl -s 'https://cdn.chorke.org/exec/cli/bash/install/apache-spark/3.3.0-fedora.sh.txt');\
else printf 'curl \033[0;31mnot found! \033[0m:(\n';fi
EOF


:<<EOF
if [[ -x "$(command -v curl)" ]];then \
sudo rpm-ostree cleanup --base;\
export SSH_PUB_KEY_TAG="spark@${HOSTNAME}";\
export PYSPARK_PYTHON_DIR='/usr/bin/python3';\
export HADOOP_HOME_DIR='/var/hadoop/hadoop-3.3.4';\
export JAVA_HOME_DIR='/usr/lib/jvm/java-11-openjdk';\
export SPARK_HOME_DIR='/var/spark/spark-3.3.0-bin-hadoop3';\
bash <(curl -s 'https://cdn.chorke.org/exec/cli/bash/install/apache-spark/3.3.0-fedora.sh.txt');\
else printf 'curl \033[0;31mnot found! \033[0m:(\n';fi
EOF


function validate(){
    # resolve spark user home if not setted up
    if [[ "${SPARK_USER_HOME}/env/set" == '/env/set' ]];then
        export SPARK_USER_HOME='/home/spark'
    fi

    # resolve hadoop home if not setted up
    if [[ "${HADOOP_HOME_DIR}/env/set" == '/env/set' ]];then
        export HADOOP_HOME_DIR='/var/hadoop/hadoop-3.3.4'
    fi

    # check existence of wget
    if [[ ! -x "$(command -v wget)" ]];then
        printf 'wget \033[0;31mnot found! \033[0m:(\n';exit 1
    fi

    # resolve spark home if not setted up
    if [[ "${SPARK_HOME_DIR}/env/set" == '/env/set' ]];then
        export SPARK_HOME_DIR='/var/spark/spark-3.3.0-bin-hadoop3'
    fi

    # resovle and check spark base from it's home
    export SPARK_BASE="$(dirname "${SPARK_HOME_DIR}")"
    if [[ "${SPARK_BASE}" == '.' ]];then
        printf 'SPARK_BASE \033[0;31mpath is invalid \033[0m:)\n';exit 1
    fi

    # check existence of spark
    if [[ -d "${SPARK_HOME_DIR}" ]];then
        if [[ "$(ls -A "${SPARK_HOME_DIR}")" ]];then
            printf 'SPARK_HOME \033[0;31malready exists \033[0m:)\n';exit 1
        else rm -rf "${SPARK_HOME_DIR}";fi
    fi

    # check existence of hadoop
    if [[ ! -d "${HADOOP_HOME_DIR}/etc/hadoop" ]]||
       [[ ! "$(ls -A "${HADOOP_HOME_DIR}/etc/hadoop")" ]];then
            printf 'HADOOP_CONF_DIR \033[0;31mnot found! \033[0m:(\n';exit 1
    fi

    if [[ ! -f "${PYSPARK_PYTHON_DIR}" ]]||[[ ! -x "$(command -v python3)" ]];then
        printf 'python3 \033[0;31mnot found! \033[0m:(\n';exit 1
    fi

    if [[ ! -f "${JAVA_HOME_DIR}/bin/java" ]]||[[ ! -x "$(command -v java)" ]];then
        printf 'java \033[0;31mnot found! \033[0m:(\n';exit 1
    fi

    if [[ -f '/etc/systemd/system/spark.service' ]];then
        printf 'spark.service \033[0;31malready installed \033[0m:)\n';exit 1
    fi
}

function download(){
    if [[ ! -d "${SPARK_HOME_DIR}" ]];then
        if [[ ! -d "${SPARK_BASE}" ]];then sudo mkdir -p "${SPARK_BASE}";fi
        sudo wget -cq https://archive.apache.org/dist/spark/spark-3.3.0/spark-3.3.0-bin-hadoop3.tgz -P "${SPARK_BASE}"
        sudo tar -xzf "${SPARK_BASE}"/spark-3.3.0-bin-hadoop3.tgz -C "${SPARK_BASE}"
        sudo rm  -rf  "${SPARK_BASE}"/spark-3.3.0-bin-hadoop3.tgz
    fi
}

function approve(){
    if [[ "$(grep -c '^spark:' /etc/group)"  -eq 0 ]];then sudo groupadd --system spark;fi
    if [[ "$(grep -c '^spark:' /etc/passwd)" -eq 0 ]];then
        sudo useradd --system spark --gid spark --create-home --home-dir ${SPARK_USER_HOME};
    fi

    sudo mkdir -p ${SPARK_HOME_DIR}/etc
    sudo chown -R spark:spark ${SPARK_HOME_DIR}
}

function setting(){
    # spark environment config
    SPARK_ENV_FILE="${SPARK_HOME_DIR}/conf/spark-env.sh"
    SPARK_ENV_TEMP="${SPARK_HOME_DIR}/conf/spark-env.sh.template"
    if  [[ -f "${SPARK_ENV_TEMP}" ]]&&[[ ! -f "${SPARK_ENV_FILE}" ]];then
        sudo cp ${SPARK_ENV_TEMP} ${SPARK_ENV_FILE}
    fi

    if [[ -f "${SPARK_ENV_FILE}" ]];then
        sudo tee -a ${SPARK_ENV_FILE} >>/dev/null <<EOF


# spark environment config
export JAVA_HOME=${JAVA_HOME_DIR}
export PYSPARK_PYTHON=${PYSPARK_PYTHON_DIR}


# hadoop integration config
export HADOOP_CONF_DIR=${HADOOP_HOME_DIR}/etc/hadoop
export YARN_CONF_DIR=${HADOOP_HOME_DIR}/etc/hadoop

EOF
    fi

    # worker environment config
    WORKERS_FILE="${SPARK_HOME_DIR}/conf/workers"
    WORKERS_TEMP="${SPARK_HOME_DIR}/conf/workers.template"
    if  [[ -f "${WORKERS_TEMP}" ]]&&[[ ! -f "${WORKERS_FILE}" ]];then
        sudo cp ${WORKERS_TEMP} ${WORKERS_FILE}
    fi

    sudo -i -u spark bash << EOF
# geneate ssh key for password less shell access 
ssh-keygen -q -b 4096 -t rsa -f ${SPARK_USER_HOME}/.ssh/id_rsa -C ${SSH_PUB_KEY_TAG} -N ''
cat ${SPARK_USER_HOME}/.ssh/id_rsa.pub >> ${SPARK_USER_HOME}/.ssh/authorized_keys
EOF
}

function install(){
    sudo chown -R spark:spark ${SPARK_HOME_DIR}
    sudo tee -a /etc/systemd/system/spark.service >/dev/null <<EOF
[Unit]
Description=Apache Spark Master
Documentation=https://spark.apache.org/docs/3.3.0
Wants=network-online.target
After=network-online.target

[Service]
User=spark
Group=spark
Type=forking
SuccessExitStatus=143

WorkingDirectory=${SPARK_HOME_DIR}/sbin
ExecStartPost=/bin/bash -c "echo \$MAINPID > ${SPARK_HOME_DIR}/etc/spark.pid"
ExecStart=${SPARK_HOME_DIR}/sbin/start-all.sh
ExecStop=${SPARK_HOME_DIR}/sbin/stop-all.sh

[Install]
WantedBy=multi-user.target
EOF
}

function launch(){
    sudo systemctl daemon-reload
    sudo systemctl enable spark.service
    sudo systemctl start  spark.service
    sudo systemctl status spark.service
}

function init(){
    validate
    download
    approve
    setting
    install
    launch
}

init