python3 virtualenv/venv helper bash script (env init, env bootstrap, pip requirement installation, program exec) with explanation

venv_explained.sh

#!/bin/bash

# Go to the location where your script located wherever your run it,
# and store the path contains the script at DIR.
# ${BASH_SOURCE[0]} is a more precised way to locate compare to $0
# https://stackoverflow.com/questions/35006457/choosing-between-0-and-bash-source
DIR="$( cd "$( dirname "${BASH_SOURCE[0]}" )" >/dev/null 2>&1 && pwd )"


# Set DIR as the working directory.
# If the directory does not have the .venv folder,
# then we create a new one and install wheel inside.
export VIRTUAL_ENV="$DIR/.venv"
if [[ ! -e "$VIRTUAL_ENV" ]]; then
    if ! python3 -m venv "$VIRTUAL_ENV"; then
        echo "can not init venv"
        exit 1
    fi
    "$VIRTUAL_ENV/bin/pip" install wheel
fi


# Set the virtual env to PATH
export PATH="$VIRTUAL_ENV/bin:$PATH"
# Clean the bash cache and from now on for the new command bash need to lookup the PATH
hash -r

# Get the first arg after venv.sh
cmd=$1
# Renamed #n to #n-1 at this case
shift

# If cmd is setup
# then we install requirements in the venv
if [[ "$cmd" = "setup" && "$1" = "" ]]; then
    exec pip install -r "$DIR/requirements.txt"
# If no cmd
# then we place a (dirname) before the command line identifier
# eg. (py3)~
elif [[ "$cmd" == "" ]]; then
    dir_name=$(basename "$DIR")
    exec bash --rcfile <(cat ~/.bashrc 2>/dev/null; echo "PS1=\"($dir_name) \$PS1\"")
fi

# Finally we run the cmd with the rest of args
# $@ means args except $0
exec "$cmd" "$@"

# how to use:
# $ ./venv.sh         -- init and get a bash shell with venv prompt
# $ ./venv.sh setup   -- init and install pip requirements
# $ ./venv.sh pip ... -- run programs in venv