BASH: Mock data generator with some basic ETL for simple jobs. (Handles fields with commas.)

genMockData.sh

#!/usr/bin/env bash
# genMockData.sh
# ------------------------------------------------------------------------------
# Gets some free sample data from: https://www.briandunning.com/sample-data/,
# Extracts Transformas and Loads the data. (ETL in BASH...)

# regions: us, ca, uk, au
region=${1:-ca} # ca (default), us, uk, au - supply as arg1
size=500 # 500 sample size is free
file="${region}-${size}"

# Get the data
[[ ! -f "$file".zip ]] && wget -q https://www.briandunning.com/sample-data/"$file".zip

# Extract & normalize for ^M
[[ ! -f "$file".csv ]] && unzip -q "$file".zip
[[ ! -f "$file".normalized.csv ]] && tr '\r' '\n' < "$file".csv > "$file".normalized.csv

# Note: Using AWK as this data set has commas within some fields.
# Extract and transform fields as necessary
#   - zip is getting uppercased and spaces removed
#   - emails are being lowercased
gawk -vFPAT='[^,]*|"[^"]*"' \
  '{first_name=$1; second_name=$2; company_name=$3; address=$4; city=$5; \
     country=$6; state=$7; zip=$8; phone1=$9; phone2=$10; email=tolower($11); web=$10; } \
   {zip=toupper(zip)} {sub(" ","",zip)} \
   {email=tolower(email)} \
   {print first_name "," second_name "," zip "," email} \
  ' ca-500.normalized.csv

# An alternate approach would be defining human-readabel vars outside of awk for each column,
# like 'first_name=1', and using them with awk like:
#   gawk -vFPAT='[^,]*|"[^"]*"' "toupper(sub(/ /,\"\",\$$zip)); BEGIN {print \$$first_name "," \$$last_name "," tolower(\$$email)}" ca-500.normalized.csv