deterministic-fake-names.rb

require 'digest/md5'

# This demonstrates an approach you can use to deterministically generate fake
# data based on user data to anonymize it.

# An array of substitute names read from a file
substitute_names = File.read('names.txt').split("\n")

# Real customer data, e.g., from a database
real_name = "John Realname"

# Hash the real name
name_hash = Digest::MD5.hexdigest(real_name) # => "ee4743cf5e5e2e90e27bf15f44e4afa9"

# Convert the name to an integer
name_hash_as_integer = name_hash.to_i(16) # => 316726291424644485479127505202124730281

# Fit the integer within the number of substitute names
new_name_index = name_hash_as_integer % substitute_names.count

# New name to use instead
replacement_name = substitute_names[new_name_index] # => "Ava Johnson"

puts replacement_name