How use GNU parallel and GNU SQL with \copy in PostgreSQL

Both GNU parallel and pg_bulkload are cool, but not available in most default installations. I think this task can be achieved by using '&' (background sub-shell) and 'wait' in a shell script, which invokes multi \COPY operations simultaneously. Here is an example:

#!/bin/bash

PG_USER_NAME=bizusr
PG_DB_NAME=bizdb
BCP_DIR=/data/biz/bcp/input

do_bcp()
{
    TABLE_NAME=$1
    echo "`date` $$ copy $TABLE_NAME begin"
    psql -q -U $PG_USER_NAME -d $PG_DB_NAME << EOF
-- SET DATESTYLE TO 'ISO,YMD'; -- you may need this when dealing with timestamps
\COPY $TABLE_NAME FROM '${BCP_DIR}/${TABLE_NAME}.bcp' WITH (FORMAT CSV, DELIMITER '|');
EOF
    echo "`date` $$ copy $TABLE_NAME done"
}

echo "`date` $$ parallel copy started"
for BCP_FILE in `ls ${BCP_DIR}/*.bcp`; do
    TABLE_NAME=`echo $BCP_FILE|awk -F"/" '{print $NF}'|sed -e s/\.bcp$//`
    do_bcp $TABLE_NAME &
done

wait
echo "`date` $$ parallel copy finished"