#!/bin/bash

# Bash script to download geoscrub_input table dump (created by AMK) from the

# vegbien database, and load it into the geoscrub database (i.e., the

# postgis database prepped with geonames.org data, GADM2 data, and

# associated mapping tables).

#

# Won't be necessary if we end up injecting all of the geoscrubbing and

# geovalidation functionality directly into vegbien itself. And if we

# end up implementing this stuff as a standalone service instead, we'd

# need to rethink (and generalize) how the input data is handled. But

# for now, this should at least serve as a placeholder that could be

# tweaked manually to load any arbitrary geoscrub input data table.

#

# Jim Regetz

# NCEAS

# Created Nov 2012

# 

# Paul Sarando

# iPlant Collaborative

# Updated Oct 2013

# Note, to force data to download from DATA_URL, ensure the DATAFILE is deleted

# before running this script.

DB_NAME="geoscrub"

DB_USER="bien"

DB_HOST_OPT=""

SCRIPT_DIR="$(dirname $0)"

DATA_URL="http://fs.vegpath.org/exports/geoscrub_input.csv"

DATADIR="${HOME}/geoscrub_input"

: "${DATAFILE:=geoscrub-corpus.csv}"

function usage {

    echo "Usage: $0 [OPTIONS]" >&2

    echo "Valid Options:" >&2

    echo "-d, --dbname=DBNAME      database name psql commands will connect to" >&2

    echo "-h, --host=HOSTNAME      database server host or socket directory" >&2

    echo "-U, --username=USERNAME  database user name" >&2

    echo "-i, --geoscrub-input     Geoscrub input directory (default: ${HOME}/geoscrub_input)" >&2

    echo "                         Delete this directory, or the input CSV in it," >&2

    echo "                         to re-download the data." >&2

    exit 1;

}

while [[ $# -gt 0  ]]; do

    case "$1" in

        -\? | --help)

            usage

            ;;

        -h)

            if [[ -z $2  ]];  then

                echo "Option $1 requires an argument." >&2

                usage

            fi

            DB_HOST_OPT="-h $2"

            shift 2

            ;;

        --host=*)

            DB_HOST_OPT="-h ${1#*=}"

            shift

            ;;

        -U)

            if [[ -z $2  ]];  then

                echo "Option $1 requires an argument." >&2

                usage

            fi

            DB_USER="$2"

            shift 2

            ;;

        --username=*)

            DB_USER="${1#*=}"

            shift

            ;;

        -d)

            if [[ -z $2  ]];  then

                echo "Option $1 requires an argument." >&2

                usage

            fi

            DB_NAME="$2"

            shift 2

            ;;

        --dbname=*)

            DB_NAME="${1#*=}"

            shift

            ;;

        -i)

            if [[ -z $2  ]];  then

                echo "Option $1 requires an argument." >&2

                usage

            fi

            DATADIR="$2"

            shift 2

            ;;

        --geoscrub-input=*)

            DATADIR="${1#*=}"

            shift

            ;;

        *)

            echo "Invalid option: $1" >&2

            usage

            ;;

    esac

done

if [[ ! -d "$DATADIR" ]]; then

    echo "making directory ${DATADIR}"

    mkdir -p "$DATADIR"

    if [[ $? != 0 ]]; then

        echo "Could not create directory ${DATADIR}"

        exit 1

    fi

fi

if [[ ! -r "${DATADIR}/${DATAFILE}" ]]; then

    # download distinct records from vegbien

    wget -O "${DATADIR}/${DATAFILE}" "$DATA_URL"

    if [[ $? != 0 ]]; then

        echo "Could not download input to ${DATADIR}/${DATAFILE}"

        exit 1

    fi

fi

echo "Loading vegbien data from ${DATADIR}/${DATAFILE}"

# clear previous data

psql -e -U "$DB_USER" $DB_HOST_OPT -d "$DB_NAME" --set ON_ERROR_STOP=1 < "${SCRIPT_DIR}/truncate.vegbien_geoscrub.sql"

if [[ $? != 0 ]]; then

    echo "Could not clear data from vegbien_geoscrub tables."

    exit 1

fi

# load vegbien_geoscrub table with input data

echo "Copying vegbien_geoscrub from ${DATADIR}/${DATAFILE}"

psql -U "$DB_USER" $DB_HOST_OPT -d "$DB_NAME" -c "\COPY vegbien_geoscrub FROM '${DATADIR}/${DATAFILE}' WITH CSV HEADER"